<%BANNER%>






Vaccine-Associated Guillain-Barre Syndrome: A Pharmacovigilance Analysis of Data in the United States' Vaccine Adverse E...
www.ispor.org ( Publisher's URL )
CITATION PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/IR00000601/00001
 Material Information
Title: Vaccine-Associated Guillain-Barre Syndrome: A Pharmacovigilance Analysis of Data in the United States' Vaccine Adverse Event Reporting System (1990-2009).
Physical Description: Conference Papers
Creator: Ayad K Ali
unknown
Conference: 16th ISPOR Annual International Meeting.
Publisher: ISPOR
Place of Publication: USA
Publication Date: May 21, 2011
 Notes
Abstract: OBJECTIVES: To describe reporting rates and characteristics of vaccine-associated GBS reports in the Vaccine Adverse Event Reporting System (VAERS) over a two decade period. METHODS: Adverse event reports submitted to VAERS from January 1990 to November 2009 are utilized to conduct a retrospective pharmacovigilance analysis by calculating the proportional reporting ratios (PRR) and corresponding 95% confidence intervals (CI) for vaccine-GBS event pairs. RESULTS: One thousand and 259 reports of vaccine-associated GBS are identified for 37 vaccines. The mean age of patients experienced GBS after vaccination was 31 years; 51% of patients were females. Majority of the reports were for influenza (FLU) (791) and hepatitis B (HEP) virus vaccines (103). Reports for human papillomavirus quadrivalent (HPV4), meningococcal conjugate (MNQ), and measles/mumps/rubella (MMR) virus vaccines accounted for 37, 36, and 34 reports, respectively. Death, disability and hospitalization among the serious GBS outcomes were correspondingly reported in 36 (2.9%), 199 (15.8%), and 963 (76.4%) reports. Most of these outcomes were in FLU reports. The average postvaccination GBS onset delay was 3.5 weeks; ranging from 2.3 weeks for FLU to 8.1 weeks for MNQ. About 79% and 12% of GBS reports included zero, and 2-4 concomitant vaccines, respectively. About 49% of the reports indicated that patients were on concurrent medications at the time of vaccination, yet; only 12% of the reports stated that the patients were having coexisting illnesses at vaccination time. Compared to other vaccines, FLU and MNQ vaccines showed significantly higher PRR values (10.5, 95%CI 9.4-11.8), (4.7, 95%CI 3.4-6.6). CONCLUSIONS: It can be postulated that some vaccines are associated with GBS, especially FLU and MNQ vaccines; underscoring the importance of monitoring patients for postvaccination signs and symptoms of GBS. While healthcare professionals are expected to promote vaccination, they should also continue to report vaccine-associated adverse events to vaccine safety monitoring systems, e.g. VAERS.
Acquisition: Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Ayad Ali.
Publication Status: Published
General Note: Citation: Value in Health. May 2011;14(3):A113-114.
 Record Information
Source Institution: University of Florida Institutional Repository
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution.
System ID: IR00000601:00001

Downloads

This item is only available as the following downloads:

Ali AK_VAERS_2011 ( PDF )


Full Text

PAGE 1

r fnt n rfntn rb r ntn ntn rrfntf

PAGE 3

EDITORIALBOARDCo-Editor-in-Chief MichaelDrummond,PhD UniversityofYork Heslington,York,UK mike.drummond@york.ac.uk Co-Editor-in-Chief C.DanielMullins,PhD UniversityofMaryland Baltimore,MD,USA dmullins@rx.umaryland.eduCO-EDITORSChrisBingefors,PhD,MSc UppsalaUniversity Uppsala,Sweden chris.bingefors@farmaci.uu.se AlanBrennan,PhD UniversityofShefeld Shefeld,UK A.Brennan@shefeld.ac.uk JalpaDoshi,PhD UniversityofPennsylvania Philadelphia,PA,USA jdoshi@mail.med.upenn.edu SheriFehnel,PhD,MA RTIHealthSolutions ResearchTrianglePark,NC,USA sfehnel@rti.org BenjaminP.Geisler,MD,MPH WingTech,Inc. BlueHill,ME,USA ben.geisler@gmail.com DanGreenberg,PhD Ben-GurionUniversityoftheNegev Beer-Sheva,Israel dangr@bgu.ac.il JohnHornberger,MD,MS CedarAssociates,LLC MenloPark,CA,USA jhornberger@cedarecon.com TeresaKauf,PhD UniversityofFlorida Gainesville,FL,USA kauf@cop.u.edu GordonG.Liu,PhD PekingUniversity,GuanghuaSchoolof Management PRChina ggliu@unc.edu AndrewLloyd,DPhil,BSc OxfordOutcomes,Ltd. Oxford,UK andrew.lloyd@oxfordoutcomes.com AndreaManca,PhD,MSc UniversityofYork York,UK andrea.manca@york.ac.uk MichelleNaughton,PhD WakeForestUniversitySchoolofMedicine Winston-Salem,NC,USA naughton@wfubmc.edu PaulScuffham,PhD,BA GrifthUniversity-SchoolofMedicine Queensland,Australia p.scuffham@grifth.edu.au JohanL.(Hans)Severens,PhD ErasmusUniversity Rotterdam,TheNetherlands severens@bmg.eur.nl Ya-Chen(Tina)Shih,PhD,MS UniversityofChicago Chicago,IL,USA tshih@medicine.bsb.uchicago.edu UllaS.Skjoldborg,PhD,MA EliLillyDenmarkA/S Copenhagen,Denmark ulla@skjoldborg.bizEDITORIALADVISORYBOARDMarcL.Berger,MD IngenixLifeSciences NewYork,NY,USA marc.berger@i3innovus.com AndrewBriggs,DPhil UniversityofGlasgow Glasgow,UK a.briggs@clinmed.gla.ac.uk DSPeteFullerton,PhD StrategicPharmacyInnovations Seattle,WA,USA peteful@msn.com JeanPaulGagnon,PhD Pittstown,NJ,USA japagon08@gmail.com HenryGlick,PhD UniversityofPennsylvania Philadelphia,PA,USA hlthsvrs@mail.med.upenn.edu DonHusereau,MSc,BSc UniversityofOttawa Ottawa,ON,Canada don.husereau@gmail.com PaulKind UniversityofYork Heslington,York,UK pk1@york.ac.uk PaulC.Langley,PhD UniversityofMinnesota Woodbury,MN,USA P8366@MSN.com PabloLapuerta,MD Bristol-MyersSquibb Princeton,NJ,USA pablo.lapuerta@bms.com AdrianLevy,PhD DalhousieUniversity Halifax,NS,Canada adrian.levy@dal.ca StevenE.Marx,PharmD,MS AbbottLaboratories AbbottPark,IL,USA steve.marx@abbott.com KarlA.Matuszewski,MS,PharmD FirstDataBank,Inc. SouthSanFrancisco,CA,USA karl_matuszewski@rstdatabank.com WilliamF.McGhan,PharmD,PhD UniversityoftheSciencesinPhiladelphia Philadelphia,PA,USA w.mcghan@usip.edu RichardJ.Milne,PhD UniversityofAuckland Auckland,NewZealand rj.milne@auckland.ac.nz LouisA.Morris,PhD LouisA.Morris&Associates DixHills,NY,USA lmorris@optonline.net PeterJ.Neumann,ScD TuftsNewEnglandMedicalCenter Boston,MA,USA pneumann@tuftsmedicalcenter.org MarkNuijten,PhD,MD,MBA ArsAccessusMedica Dorpsstraat,TheNetherlands marknuijten@arsaccessusmedica.com ChrisL.Pashos,PhD UnitedBioSourceCorporation Lexington,MA,USA Chris.Pashos@unitedbiosource.com DennisRevicki,PhD UBC Bethesda,MD,USA dennis.revicki@unitedbiosource.com KevinA.Schulman,MD DukeClinicalResearchInstitute Durham,NC,USA schul012@mc.duke.edu UweSeibert,MD,MPH,MSc,ScD UniversityofHealthSciences, MedicalInformatics&Technology Halli.T.,Austria uwe.siebert@umit.at SeanSullivan,PhD UniversityofWashington Seattle,WA,USA sdsull@u.washington.edu MiltonC.Weinstein,PhD HarvardSchoolofPublicHealth Boston,MA,USA mcw@hsph.harvard.eduMANAGEMENTADVISORYBOARDShelbyReed,PhD,RPh(Chair) DukeUniversity Durham,NC,USA shelby.reed@duke.edu PaulKind UniversityofYork Heslington,York,UK pk1@york.ac.uk Bong-MinYang,PhD SeoulUniversity Seoul,SouthKorea bmyang@sun.ac.krEDITORIALOFFICEManagingEditorStephenL.Priori ISPOR Lawrenceville,NJ,USA spriori@ispor.orgEditorialAssistantDanielleMroz ISPOR Lawrenceville,NJ,USA dmroz@ispor.org

PAGE 4

PodiumSessionI A1 CancerOutcomesResearch: CN1–CN4 A1 ComparativeEffectivenessResearch: CO1–CO4 A2 EffectsofDrugManagementProgramsonPatients: DM1–DM4 A3 EmployeeHealth&ProductivityOutcomesResearch: OR1–OR4 A4 CaseStudiesinAddressingSelectionBias: SB1–SB4 PodiumSessionII A5 ResearchonMethods:Cost-EffectivenessAnalysis: CE1–CE4 A6 ResearchonMethods:DatabaseAnalysis: DS1–DS4 A6 DrugUseandPatientSafety: DU1–DU4 A7 InfectionOutcomesResearch: IN1–IN4 A8 ImpactofMedicationCompliance: MC1–MC4 PodiumSessionIII A9 ResearchonMethods:EconomicEvaluations: EE1–EE4 A10 EvolvingConceptsinOutcomesResearch: EV1–EV4 A10 AnalysisofMedicarePolicyandResourceUse: MD1–MD4 A11 ExaminingtheQALY: QA1–QA4 PosterSessionI HealthCareUse&PolicyStudies A12 HealthCareUse&PolicyStudies–ConsumerRoleinHealthCare: PHP1–PHP5 A13 HealthCareUse&PolicyStudies–DiagnosisRelatedGroup: PHP6 A13 HealthCareUse&PolicyStudies–Drug/Device/DiagnosticUse&Policy: PHP7–11,PHP13–PHP20,PHP22–PHP28 A17 HealthCareUse&PolicyStudies–EquityandAccess: PHP29–PHP31 A17 HealthCareUse&PolicyStudies–FormularyDevelopment: PHP32–PHP33,PHP35 A18 HealthCareUse&PolicyStudies–HealthCareCosts&Management: PHP36–PHP38,PHP40–PHP57, PHP59–PHP68 A23 HealthCareUse&PolicyStudies–HealthCareResearch&Education: PHP69–PHP82 A26 HealthCareUse&PolicyStudies–HealthTechnologyAssessmentPrograms: PHP83–PHP85 A27 HealthCareUse&PolicyStudies–PopulationHealth: PHP86 A27 HealthCareUse&PolicyStudies–PrescribingBehavior&TreatmentGuidelines: PHP87–PHP90 A28 HealthCareUse&PolicyStudies–QualityofCare: PHP91–PHP93 A28 HealthCareUse&PolicyStudies–RegulationofHealthCareSector: PHP94–PHP100 A30 HealthCareUse&PolicyStudies–RiskSharing/Performance-BasedAgreements: PHP101–PHP105 A30 HealthCareUse&PolicyStudies–ConceputalPapers: PHP106–PHP114 VOLUME14NUMBER3MAY2011 ISPOR16thANNUALINTERNATIONALMEETINGRESEARCHABSTRACTS

PAGE 5

Disease-SpecicStudies A32 CardiovascularDisorders–ClinicalOutcomesStudies: PCV1–PCV8,PCV10–PCV31 A38 CardiovascularDisorders–CostStudies: PCV32–PCV36,PCV38–PCV58 A43 CardiovascularDisorders–Patient-ReportedOutcomes&Preference-BasedStudies: PCV59–PCV78 A46 CardiovascularDisorders–HealthCareUse&PolicyStudies: PCV79–PCV108 A52 CardiovascularDisorders–ResearchonMethods: PCV109–PCV112 A53 SensorySystemsDisorders–ClinicalOutcomesStudies: PSS1–PSS6 A54 SensorySystemsDisorders–CostStudies: PSS7–PSS14 A56 SensorySystemsDisorders–Patient-ReportedOutcomes&Preference-BasedStudies: PSS15–PSS22 A57 SensorySystemsDisorders–HealthCareUse&PolicyStudies: PSS23–PSS26 A58 SensorySystemsDisorders–ResearchonMethods: PSS27–PSS30 A59 SystemicDisorders/Conditions–ClinicalOutcomesStudies: PSY1–PSY7 A60 SystemicDisorders/Conditions–CostStudies: PSY8–PSY29 A64 SystemicDisorders/Conditions–Patient-ReportedOutcomes&Preference-BasedStudies: PSY30–PSY47 A68 SystemicDisorders/Conditions–HealthCareUse&PolicyStudies: PSY48–PSY52,PSY54–PSY74 A72 SystemicDisorders/Conditions–ResearchonMethods: PSY75–PSY81 A74 Urinary/KidneyDisorders–ClinicalOutcomesStudies: PUK1–PUK6 A75 Urinary/KidneyDisorders–CostStudies: PUK7–PUK17 A77 Urinary/KidneyDisorders–Patient-ReportedOutcomes&Preference-BasedStudies: PUK18 A77 Urinary/KidneyDisorders–HealthCareUse&PolicyStudies: PUK19–PUK21 A78 Urinary/KidneyDisorders–ResearchonMethods: PUK22–PUK24 PosterSessionII SelectedHealthCareTreatmentStudies A79 MedicalDevice/Diagnostics–ClinicalOutcomesStudies: PMD1–PMD10 A80 MedicalDevice/Diagnostics–CostStudies: PMD11–PMD27,PMD29 A84 MedicalDevice/Diagnostics–Patient-ReportedOutcomes&Preference-BasedStudies: PMD30–PMD31,PMD34 A85 MedicalDevice/Diagnostics–HealthCareUse&PolicyStudies: PMD35–PMD39 A85 MedicalDevice/Diagnostics–ResearchonMethods: PMD40–PMD41 A86 Surgery–ClinicalOutcomesStudies: PSU1–PSU5 A87 Surgery–CostStudies: PSU6–PSU11 A88 Surgery–Patient-ReportedOutcomes&Preference-BasedStudies: PSU12–PSU15 A89 Surgery–HeathCareUse&PolicyStudies: PSU16–PSU19 A89 Surgery–ResearchonMethods: PSU20–PSU25 Disease-SpecicStudies A91 Diabetes/EndocrineDisorders–ClinicalOutcomesStudies: PDB1–PDB18 A94 Diabetes/EndocrineDisorders–CostStudies: PDB19–PDB36 A97 Diabetes/EndocrineDisorders–Patient-ReportedOutcomes&Preference-BasedStudies: PDB37–PDB49 A100 Diabetes/EndocrineDisorders–HealthCareUse&PolicyStudies: PDB50–PDB71 A104 Diabetes/EndocrineDisorders–ResearchonMethods: PDB72–PDB76VALUEINHEALTH14(2011)

PAGE 6

A105 Individual’sHealth–ClinicalOutcomesStudies: PIH1–PIH7 A106 Individual’sHealth–CostStudies: PIH8–PIH21 A109 Individual’sHealth–Patient-ReportedOutcomes&Preference-BasedStudies: PIH22–PIH32 A111 Individual’sHealth–HealthCareUse&PolicyStudies: PIH34–PIH40 A112 Individual’sHealth–ResearchonMethods: PIH41–PIH46 A113 Infection–ClinicalOutcomesStudies: PIN1–PIN8 A115 Infection–CostStudies: PIN9–PIN17,PIN19–PIN35 A120 Infection–Patient-ReportedOutcomes&Preference-BasedStudies: PIN36–PIN41 A121 Infection–HealthCareUse&PolicyStudies: PIN43–PIN45,PIN47–PIN52 A123 Infection–ResearchonMethods: PIN53 A123 Muscular-SkeletalDisorders–ClinicalOutcomesStudies: PMS1–PMS7 A124 Muscular-SkeletalDisorders–CostStudies: PMS8–PMS9,PMS11–PMS36 A129 Muscular-SkeletalDisorders–Patient-ReportedOutcomes&Preference-BasedStudies: PMS37–PMS47 A132 Muscular-SkeletalDisorders–HealthCareUse&PolicyStudies: PMS48–PMS54,PMS56–PMS58, PMS60–PMS69 A136 Muscular-SkeletalDisorders–ResearchonMethods: PMS70–PMS74 A136 Respiratory-RelatedDisorders–ClinicalOutcomesStudies: PRS1–PRS8 A138 Respiratory-RelatedDisorders–CostStudies: PRS9–PRS21 A140 Respiratory-RelatedDisorders–Patient-ReportedOutcomes&Preference-BasedStudies: PRS22–PRS31 A142 Respiratory-RelatedDisorders–HealthCareUse&PolicyStudies: PRS32–PRS42 A144 Respiratory-RelatedDisorders–ResearchonMethods: PRS43–PRS46 PosterSessionIII ResearchonMethodsStudies A145 ResearchonMethods–ClinicalOutcomesMethods: PRM1–PRM3 A146 ResearchonMethods–CostMethods: PRM4–PRM9 A147 ResearchonMethods–Databases&ManagementMethods: PRM10–PRM11 A147 ResearchonMethods–ModelingMethods: PRM12–PRM14 A148 ResearchonMethods–Patient-ReportedOutcomesStudies: PRM15–PRM30,PRM32–PRM35 A152 ResearchonMethods–StatisticalMethods: PRM36–PRM38 A152 ResearchonMethods–StudyDesign: PRM39 A152 ResearchonMethods–ConceptualPapers: PRM40–PRM49 Disease-SpecicStudies A154 Cancer–ClinicalOutcomesStudies: PCN1–PCN24 A159 Cancer–CostStudies: PCN25–PCN43,PCN46–PCN50,PCN52–PCN78 A168 Cancer–Patient-ReportedOutcomes&Preference-BasedStudies: PCN79–PCN96 A172 Cancer–HealthCareUse&PolicyStudies: PCN97,PCN99–PCN125 A177 Cancer–ResearchonMethods: PCN126–PCN138 A180 GastrointestinalDisorders–ClinicalOutcomesStudies: PGI1–PGI5 A181 GastrointestinalDisorders–CostStudies: PGI6–PGI16 A183 GastrointestinalDisorders–Patient-ReportedOutcomes&Preference-BasedStudies: PGI17–PGI24 A185 GastrointestinalDisorders–HealthCareUse&PolicyStudies: PGI25–PGI26VALUEINHEALTH14(2011)

PAGE 7

A185 MentalHealth–ClinicalOutcomesStudies: PMH1–PMH18 A188 MentalHealth–CostStudies: PMH19–PMH44 A193 MentalHealth–Patient-ReportedOutcomes&Preference-BasedStudies: PMH45–PMH57 A196 MentalHealth–HealthCareUse&PolicyStudies: PMH58–PMH80 A200 MentalHealth–ResearchonMethods: PMH81–PMH85 A201 NeurologicalDisorders–ClinicalOutcomesStudies: PND1–PND9 A203 NeurologicalDisorders–CostStudies: PND10–PND27 A207 NeurologicalDisorders–Patient-ReportedOutcomes&Preference-BasedStudies: PND28–PND50 A211 NeurologicalDisorders–HealthCareUse&PolicyStudies: PND51–PND59 A213 NeurologicalDisorders–ResearchonMethods: PND60–PND65 A215DISCLOSUREINFORMATION A223AUTHORINDEXVALUEINHEALTH14(2011)

PAGE 8

EditorialOfce. ValueinHealth,ISPOR,3100PrincetonPike,Suite3E, Lawrenceville,NJ08648. ISPOROfce. MarilynDixSmith,RPh,PhD,ExecutiveDirector,3100Princeton Pike,Suite3E,Lawrenceville,NJ08648.Tel:(609)219-0773,Fax:(609)2190774,E-mail:info@ispor.org,Website:http://www.ispor.org/valueinhealth_ index.asp. CUSTOMERSERVICE(orders,claims,online,changeofaddress): Elsevier HealthSciencesDivision,SubscriptionCustomerService,3251RiverportLane, MarylandHeights,MO63043.Tel:(800)654-2452(U.S.andCanada);(314) 447-8871(outsideU.S.andCanada).Fax:(314)447-8029.E-mail: JournalsCustomerService-usa@elsevier.com(forprintsupport); JournalsOnlineSupport-usa@elsevier.com(foronlinesupport).Addresschanges mustbesubmittedfourweeksinadvance. YEARLYSUBSCRIPTIONRATES: UnitedStatesandpossessions:Individual $250.00;Institution$535.00.Allothercountries(pricesincludeairspeed delivery):Individual$250.00;Institution$535.00.SingleIssues$67.00.To receivestudent/residentrate,ordersmustbeaccompaniedbynameof afliatedinstitution,dateoftermandthesignatureofprogram/residency coordinatoroninstitutionletterhead.Orderswillbebilledattheindividualrate untilproofofstatusisreceived.Currentpricesareineffectforbackvolumes andbackissues. FurtherinformationonthisjournalisavailablefromthePublisherorfromthis journal'swebsite(http://www.elsevier.com/locate/jval).Informationonother ElsevierproductsisavailablethroughElsevier'swebsite(http://www. elsevier.com). Authorinquiries Forinquiriesrelatingtothesubmissionofarticles(includingelectronic submissionwhereavailable),visithttp:/www.elsevier.com/authors.Thesitealso providesthefacilitytotrackacceptedarticlesandsetupe-mailalertstoinform youofwhenanarticle'sstatushaschanged,aswellasdetailedartwork guidelines,copyrightinformation,frequentlyaskedquestions,andmore.Please seeInformationforAuthorsforindividualjournal.Contactdetailsforquestions arisingafteracceptanceofanarticle,especiallythoserelatingtoproofs,are providedafterregistrationofanarticleforpublication. Englishlanguagehelpservice: Uponrequest,Elsevierwilldirectauthorsto anagentwhocancheckandimprovetheEnglishoftheirpaper( before submission ).Pleasecontactauthorsupport@elsevier.comforfurtherinformation. Reprints. Forqueriesaboutauthoroffprints,e-mailauthorsupport@ elsevier.com.Toorder100ormorereprintsforeducational,commercial,or promotionaluse,contacttheCommercialReprintsDepartment,ElsevierInc., 360ParkAvenueSouth,NewYork,NY10010-1710.Fax:(212)462-1935;emailreprints@elsevier.com.Reprintsofsinglearticlesavailableonlinemaybe obtainedbypurchasingPay-Per-Viewaccessfor$36perarticleonthejournal websitehttp://www.elsevier.com/locate/jval. Copyright2011,InternationalSocietyforPharmacoeconomicsandOutcomes Research(ISPOR).PublishedbyElsevierInc. Thisjournalandtheindividualcontributionscontainedinitareprotectedunder copyrightbyInternationalSocietyforPharmacoeconomicsandOutcomes Research,andthefollowingtermsandconditionsapplytotheiruse: Photocopying Singlephotocopiesofsinglearticlesmaybemadeforpersonaluseasallowed bynationalcopyrightlaws.PermissionofthePublisherandpaymentofafeeis requiredforallotherphotocopying,includingmultipleorsystematiccopying, copyingforadvertisingorpromotionalpurposes,resale,andallformsof documentdelivery.Specialratesareavailableforeducationalinstitutionsthat wishtomakephotocopiesfornon-proteducationalclassroomuse. PermissionsmaybesoughtdirectlyfromElsevier'sRightsDepartmentin Oxford,UK:phone(215)238-7869or 44(0)1865843830,fax 44(0) 1865853333,e-mailhealthpermissions@elsevier.com.Requestsmayalsobe completedonlineviatheElsevierhomepage(http://www.elsevier.com/locate/ permissions). IntheUSA,usersmayclearpermissionsandmakepaymentsthroughthe CopyrightClearanceCenter,Inc.,222RosewoodDrive,Danvers,MA01923, USA;phone:(978)750-8400,fax:(978)750-4744,andintheUKthroughthe CopyrightLicensingAgencyRapidClearanceService(CLARCS),90Tottenham CourtRoad,LondonW1P0LP,UK;phone:( 44)2076315555;fax:( 44)20 76315500.Othercountriesmayhavealocalreprographicrightsagencyfor payments. DerivativeWorks Subscribersmayreproducetablesofcontentsorpreparelistsofarticles includingabstractsforinternalcirculationwithintheirinstitutions.Permissionof thePublisherisrequiredforresaleordistributionoutsidetheinstitution. PermissionofthePublisherisrequiredforallotherderivativeworks,including compilationsandtranslations. ElectronicStorageorUsage PermissionofthePublisherisrequiredtostoreoruseelectronicallyany materialcontainedinthisjournal,includinganyarticleorpartofanarticle. Exceptasoutlinedabove,nopartofthispublicationmaybereproduced,stored inaretrievalsystemortransmittedinanyformorbyanymeans,electronic, mechanical,photocopying,recordingorotherwise,withoutpriorwritten permissionofthePublisher. Addresspermissionsrequeststo:ElsevierRightsDepartment,atthefaxand e-mailaddressesnotedabove. AbstractingandIndexing ValueinHealth isindexedinIndexMedicus/MEDLINE,CurrentContents/Social& BehavioralSciences,SciSearch/SCIExpanded,SocialSciencesCitationIndex, InternationalPharmaceuticalAbstracts,Embase/ExcerptaMedica,and PsychInfo/PsychologicalAbstracts,andJournalCitationReports/ScienceEdition (ThomsonISI). Notice NoresponsibilityisassumedbythePublisherortheInternationalSocietyfor PharmacoeconomicsandOutcomesResearchforanyinjuryand/ordamageto personsorpropertyasamatterofproductsliability,negligenceorotherwise, orfromanyuseoroperationofanymethods,products,instructionsorideas containedinthematerialherein.Becauseofrapidadvancesinthemedical sciences,inparticular,independentvericationofdiagnosesanddrugdosages shouldbemade. Althoughalladvertisingmaterialisexpectedtoconformtoethical(medical) standards,inclusioninthispublicationdoesnotconstituteaguaranteeor endorsementofthequalityorvalueofsuchproductoroftheclaimsmadeofit byitsmanufacturer.

PAGE 9

ABSTRACTSISPOR16thANNUALINTERNATIONALMEETINGRESEARCHABSTRACTS PODIUMSESSIONI: CANCEROUTCOMESRESEARCH CN1 MONOTHERAPYOFANDROGENDEPRIVATIONTHERAPYVERSUSRADICAL PROSTATECTOMYAMONGVETERANSWITHLOCALIZEDPROSTATECANCER:A COMPARATIVEEFFECTIVENESSANALYSISOFRETROSPECTIVECOHORTSLiuJ,ShiL,SartorO TulaneUniversity,NewOrleans,LA,USAOBJECTIVES: Thereisnoconsensusregardingtheoptimaltreatmentforlocalized prostatecancer.Thisstudyaimedtoexaminethecomparativeeffectivenessof monotherapyofeitherprimaryandrogendeprivationtherapy(PADT)orradical prostatectomy(RP)intermsofoverallsurvivalrate. METHODS: Malepatientswith localizedprostatecancerwereidentiedintheVeteransAffairsVeteransIntegratedServiceNetwork16datawarehouse(January,2003-June,2006),withone yearbaselineandatleast3-yearfollow-up(till06/2009).Eligiblepatients(18-75 yearsold)hadnoothercancerhistoryandusedPADTormonotherapyofRPwithin 6monthsaftertherstdiagnosisofprostatecancer.Theoverallsurvivalfrom initiationofindextreatmentwasanalyzedusingKaplan-MeiermethodandCox proportionalhazardregression,adjustedforage,race,maritalstatus,insurance type,cancerstage,Charlsoncomorbidityindex,alcoholandtobaccouse. RESULTS: Theagewas66.2(6.07)[Mean(SD)]yearsin211PADTpatients,59.9(6.15)in215RP patients.Duringthefollow-upof4.2(0.95)years,thecumulativeincidenceofdeath was29(13.74%)amongPADTpatientsand6(2.79%)amongRPpatients(p 0.001). Theoverall3-yearsurvivalratewas89.57%inPADTand98.60%inRP(p 0.001). PatientswhoreceivedPADThadalmost4timesashighmortalityriskasthose usingRP(HR 3.820,95%CI 1.483to9.845,p 0.006). CONCLUSIONS: OverallsurvivalratefollowingRPamonglocalizedprostatecancerpatientswassignicantly higherthanthatafterPADT,controllingforothercovariates.Moreresearchamong alargerpopulationwithlongerfollow-uparewarrantedtoconrmthisnding. CN2 ESTIMATEDEFFECTSOFTHENATIONALBREASTANDCERVICALCANCER EARLYDETECTIONPROGRAMONCERVICALCANCERMORTALITYEkwuemeDU1,UzunangelovV2,HoergerT2,SaraiyaM1,MillerJ1,HallI1,BenardV1, RoyaltyJ1 ,LiC1 1CentersforDiseaseControlandPrevention,Atlanta,GA,USA,2RTIInternational,Research TrianglePark,NC,USAOBJECTIVES: TheNationalBreastandCervicalCancerEarlyDetectionProgram(NBCCEDP)isthelargestorganizedcancerscreeningprogramforlow-income,un-insured andunder-insuredwomenintheUnitedStates.Theprogram'seffectivenessinincreasingthelifeexpectancyofparticipatingwomenhasneverbeenmeasured.We estimatedthebenetsofNBCCEDP-fundedcervicalcancerscreening(Program)in termsoflife-years(LYs)savedcomparedtoNoProgramandNoScreeningscenarios. METHODS: BasedonanexistingmodeldevelopedbyMyersetal.,weconstructeda cervicalcancersimulationmodelbymodifyingtheageandscreeningscheduleofthe cohorttoreectscreeningfrequencyforNBCCEDPparticipantsfrom1991-2007.We estimatedscreeninghabitsintheabsenceoftheprogrambasedondatafromthe 1990-2005NationalHealthInterviewSurvey.WeperformedMarkovcohortanalysis foreachageinthe18-64rangeandcalculatedanoverallweightedaverageusingthe agedistributionatrstNBCCEDPPaptestforscreening.WeightedaverageswereproducedforthreescenarioswomenreceivingtestingfromtheNBCCEDP(theProgram), womenreceivingtestingfromalternativesourcesintheabsenceoftheprogram(No Program),andwomenreceivingnotestingatall(NoScreening).WecomparedLY estimatesfor69,100womendetectedwithhumanpapillomavirusinfection,low-andhigh-gradesquamousintraepitheliallesionsorcervicalcancerundertheprogramto thecounterfactualofhavingtheirdiseaseundetectedunderNoProgramandNo Screeningscenarios. RESULTS: From1991-2007,weestimatethattheProgramadded 10,369LYstothetotallifespanoftestedwomenwhencomparedtoNoProgram,and 101,509LYswhencomparedtoNoScreening.Furthermore,theProgrampreventedan estimated325cervicalcancerdeathsrelativetoNoProgram,and3,825relativetoNo Screening. CONCLUSIONS: TheseestimatessuggestthatNBCCEDPcervicalcancer screeningmayhavereducedmortalityamongmedicallyunderservedwomenin theUnitedStates. CN3 THEVALUEOFRESEARCHFORERCC1TESTINGINSTAGEINON-SMALLCELL LUNGCANCERRothJ1,CarlsonJJ2,SteutenL3,VeenstraD1 1UniversityofWashington,PharmaceuticalOutcomesResearchandPolicyProgram,Seattle,WA, USA,2UniversityofWashington,Seattle,WA,USA,3UniversityofTwente,Enschede,The NetherlandsOBJECTIVES: ToassessthevalueofadditionalresearchforERCC1expressiontestingtoguideadjuvantchemotherapydecisionsinfullyresectedStageInon-small celllungcancer(NSCLC). METHODS: Werenedapreviously-developeddecisionanalyticmodeltoestimatetheexpectedvalueofperfectinformation(EVPI)and expectedvalueofsampleinformation(EVSI)fortwotreatmentstrategies:1)ERCC1 testingtoinformadjuvantchemotherapydecisions,withERCC1 patientsreceivingnochemotherapyandERCC1-patientsreceivingchemotherapy;2)standard care,withallpatientsreceivingnochemotherapy.ModelparametersanduncertaintyrangeswerederivedfromaretrospectiveanalysisoftheInternationalAdjuvantLungCancerTrial,publishedliterature,andgovernmentsources.TheaffectedpopulationwasderivedfromSEERincidenceestimates,andexaminedover adiscounted10-yeartimehorizon. RESULTS: Atawillingness-to-payof$150,000 perquality-adjustedlifeyear,ERCC1andstandardcarestrategiesresultedinaveragenet-benetsof$630,500and$625,200,respectively.TheERCC1andstandard carestrategiesproducedgreaternet-benetin64%and36%of10,000simulations, respectively.Theaveragenet-benetdifferencewas$14,000insimulationswhere thestandardcarestrategywasoptimal.Withanaffectedpopulationof233,825; EVPIwas$1.2billion.PreliminaryestimatessuggestanEVSIofapproximately$20 millionatplausiblesamplesizes. CONCLUSIONS: Considerablevaluecouldbe realizedthroughadditionalresearchtoreduceuncertaintyaboutthecomparative healthoutcomesofERCC1andstandardcarestrategies.TheEVPIof$1.2billionwas drivenbythelarge10-yearaffectedpopulation,probabilitythatERCC1testingis nottheoptimalstrategy,andconsequencesofselectingthenon-optimalstrategy. Forthcomingresultswillenableestimationoftheexpectednet-benetofsampling,whichcomparestheEVSIofvariousstudydesignsandsamplesizestothe costofconductingsuchstudies.ThesendingscanassiststakeholdersinprioritizingfundingforERCC1researchrelativetoalternativeresearchinvestments. CN4 PALONOSETRONVERSUSOTHER5-HYDROXYTRYPTAMINE3RECEPTOR ANTAGONISTSFORPREVENTIONOFCHEMOTHERAPYINDUCEDNAUSEAAND VOMITINGAMONGMEDICAREPATIENTSWITHCANCERCraverC1,GayleJ1,BaluS2,BuchnerD2 1Premier,Inc.,Charlotte,NC,USA,2Eisai,Inc.,WoodcliffLake,NJ,USAOBJECTIVES: Toassesstherateofuncontrolledchemotherapyinducednauseaand vomiting(CINV)associatedwithpalonosetroninitiationversusother5-hydroxy tryptamine3-receptorantagonists(5-HT3-RAs)amongMedicarepatientswithcancer onchemotherapy(CT)treatmentinahospitaloutpatientsetting. METHODS: MedicarepatientswithacancerdiagnosisinitiatingCTandanti-emeticprophylaxiswith palonosetron(Group1)andother5-HT3-RAs(Group2)forthersttime(indexdate) betweenApril1,2007March31,2009wereidentiedfromthePremierPerspective database.Inclusioncriteriawerenoevidenceofnauseaandvomiting,CT,andantiemeticmedicationinthe6-monthpre-indexdateperiodand36-consecutivemonths ofdatasubmission.Anegativebinomialdistributiongeneralizedlinearmultivariate regressionmodelestimatingtherateofCINVeventsonCTemetogenicityandcycle matchedgroupsinthefollow-upperiod(rstofeightCTcyclesorsixmonthspostindexdate)wasdevelopedafteradjustingforseveraldemographicandclinical variables. RESULTS: Of4799identiedpatients,962initiatedpalonosetron(Group1; 20.1%).Group1patientsweresignicantlyyounger[70.4(SD:9.3)versus71.6(9.0) years;p 0.0001],comprisedmorefemales[52.9%versus48.6%;p 0.0001],lessAfrican Americans[8.7%vs.11.3%]andmoreHispanicpatients[6.3%versus2.5%];all p 0.0001,morehighlyandmoderatelyemetogenicCT[33.6%versus20.7%and;47.3% versus40.3%,respectively;p 0.0001],andmorelungandbreast[30.9%vs.24.9%and 12.3%vs.9.6%,respectively;p 0.0001].Inthefollow-upperiod,theregressionmodel predicteda11.8%decreaseintheCINVeventsperCTcycleforGroup1patientsversus Group2patients;p 0.05. CONCLUSIONS: Inthisretrospectivehospitaloutpatient study,aftermatchingforCTemetogenicityandcycleandadjustingforotherpotential confounders,Medicarepatientswithcancerinitiatedonpalonosetronweremore likelytoexperienceasignicantlylowerrateofCINVeventsperCTcycleversusthose initiatingother5-HT3-RAs. PODIUMSESSIONI: COMPARATIVEEFFECTIVENESSRESEARCH CO1 COMPARATIVEEFFECTIVENESSANALYSISOFTNFBLOCKERSINRHEUMATOID ARTHRITIS(RA)PATIENTSINAREAL-WORLDSETTINGBonafedeRP1,PearsonD2,BabichJ3,ChastekB4,BeckerL4,WatsonC5,ChaudhariS6, HarrisonDJ7 ,GandraSR7VALUEINHEALTH14(2011)A1–A214 availableatwww.sciencedirect.com journalhomepage:www.elsevier.com/locate/jval Copyright2011,InternationalSocietyforPharmacoeconomicsandOutcomesResearch(ISPOR).PublishedbyElsevierInc.

PAGE 10

1ProvidenceArthritisCenter,Portland,OR,USA,2VenturaCountyMedicalCenter,Ventura,CA, USA,3CarolinaBone&Joint,Charlotte,NC,USA,4i3Innovus,EdenPrairie,MN,USA,5Amgen, Inc.,NewburyPark,CA,USA,6KForceClinicalResearch,Tampa,FL,USA,7Amgen,Inc., ThousandOaks,CA,USAOBJECTIVES: ToevaluateeffectsofdoseescalationonclinicaloutcomesofRA patientsinitiatingTNF-blockertreatmentsincommunitypractice. METHODS: TNF-blocker-na•veadultRApatientsinitiatingetanercept,adalimumab,oriniximab(index)betweenJuly1,2005andMay31,2008with 12months'enrollment post-indexwereidentiedfromtheIngenixdatabase.Patientsreceiving 9 monthsTNF-blockertreatmentordiagnosedwithpsoriasis,psoriaticarthritis,ankylosingspondylitis,orCrohn'sdiseasewereexcluded.Ratesofdoseescalation using3differentmethodswerecalculatedusingclaimsdata.Participatingphysiciansprovidedde-identiedcharts.Eachchartwasreviewedby46clinicalrheumatologiststoevaluateandagreeonoverallclinicalchangefrombaselinetothe visitclosestto1yearpost-index(12 3months).Multivariatemodelscompared changeinclinicaloutcomesanddoseescalationrates,controllingfordifferences amongetanercept,adalimumab,andiniximabpatientsatindex. RESULTS: Overall,715etanercept,501adalimumab,and393iniximabpatientswereidentied fromclaims;141etanercept,115adalimumab,and104iniximabpatientshad evaluablecharts.Patientcharacteristicsweresimilaramongtheclaimsandcharts. Regardlessofdoseescalationmethodused,signicantlyfeweretanercept-treated patientshaddoseescalations(1.8%,5.2%,6.7%)thanpatientstreatedwithadalimumab(9.8%,8.6%,10.4%respectively)oriniximab(50%,31%,34%respectively) (p 0.05forallcomparisons).Aftertreatmentinitiation,86%ofetanercept-treated patientshad"muchbetter"or"better"clinicaloutcomesat12 3months,versus 82%ofadalimumabpatientsand78%ofiniximabpatients.Multivariateanalyses showedsignicantlyfewerdoseescalationsinetanerceptpatients(p 0.05),with nosignicantdifferenceinclinicalchangescorebetweenetanerceptpatientsand adalimumab(p 0.22)oriniximab(p 0.07)patients. CONCLUSIONS: Thisstudy showeddoseescalationinfeweretanerceptthanadalimumaboriniximabpatients, butsimilarimprovementsinclinicaloutcomesforall3treatments,indicatingthat higherdoseescalationratesmaynotbeassociatedwithbetterclinicaloutcomes. CO2 REAL-WORLDCOST-EFFECTIVENESSANALYSISOFCANCERDRUGS: COMPARATIVEEFFECTIVENESSRESEARCHUSINGRETROSPECTIVECANADIAN REGISTRYDATABEFOREANDAFTERDRUGAPPROVALKhorS1,KrahnM2,HodgsonD3,BremnerK4,LuoJ5,HochJ1 1CancerCareOntario,Toronto,ON,Canada,2TorontoHealthEconomicsandTechnology Assessment(THETA)Collaborative,Toronto,ON,Canada,3PrincessMargaretHospital,Toronto, ON,Canada,4UniversityHealthNetwork,Toronto,ON,Canada,5TheInstituteforClinical EvaluativeSciences,Toronto,ON,CanadaOBJECTIVES: UsinglinkedadministrativedatabasesfromOntario,ourstudyexaminedthe"realworld"cost,effectivenessandcost-effectivenessofRituximabindiffuselarge-B-celllymphoma. METHODS: Patientsweredenedasthosewhohadadiagnosisofdiffuse-large-B-celllymphomaaccordingtoICD-Ohistologyclassication betweenJanuary1997andDecember2007andreceivedeitherCHOP(cyclophosphamide,doxorubicin,vincristineandprednisone)orR-CHOP(CHOPplusRituximab)as rstlinetreatment.Weusedahistoricalcohortdesigntocomparetheoverallsurvival, toxicityproles,directcosts,andcost-effectivenessofCHOPbeforeRituximabwas approved(pre-eraCHOP)withR-CHOPafterRituximabapproval(post-eraRCHOP). R-CHOPandCHOPpatientswerehard-matchedonage,andthensubsequently matchedonpropensityscoresbyuseofa1:1matchingalgorithm.Propensityscores werecalculatedfromdemographicandclinicalhistoryinformation.Weestimated resourceuseanddirectmedicalcostsusingthelinkedadministrativedata.Toanalyze censoredcostdata,weemployedandcompareddifferentmethods,includingthe simplenon-adjustedaverage,theKaplan-Meiersampleaverageestimator,inverse probabilityweightingestimator,PfeiferandBang'sestimator(2005)andBasu'stwopartestimator(2010). RESULTS: Atotalof1131matchedpairsofpatientswereevaluated.3-yearoverallsurvivalwassignicantlyimprovedinthepost-eraRCHOPgroup comparedtopre-eraCHOP(69%[95%CI66-71]vs59%[95%CI56-62];Kleintest p 0.001).Groupsdidnotdifferinthefrequencyofadverseevents,but3-yeardirect costwassignicantlyhigherinthepost-eraRCHOPgroup.Theincrementalcosteffectivenessratiovarieddependingonthemethodemployed. CONCLUSIONS: This studyillustratedhowdifferentmethodscanbeappliedtoobservationaldatatoestimatecostsandcost-effectiveness.Theresultsfromthisstudycanbecomparedto thosefromclinicaltrialsandeconomicmodels.Thiswillhelpdrugdecision-makers calibratehealthcarepolicieswhilehelpingresearchersevaluateassumptionsmade andmethodsusedineconomicmodels. CO3 PROJECTLIBRA:ANEWANALYTICTOOLFORCOMPARATIVEEFFECTIVENESS ANALYSESOFMULTIPAYERCLAIMSDATABASESMarkT1,PepitoneA2,HatzmannM1,NavatheA3,GoodrichK3,ChangS1 1ThomsonReuters,Washington,DC,USA,2ThomsonReuters,SantaBarbara,CA,USA,3AssistantSecretaryforPlanningandEvaluation,Washington,DC,USAOBJECTIVES: Theprojectaimedtodevelopasecure,interactivetooltoenable researcherstoperformcomparativeeffectivenessstudiesandothertypesofresearchonamultipayerclaimsdatabasewithreducedneedforcomplicated programming. METHODS: Acommondatamodel,throughwhichmultipledata sourcesarestandardizedandlinkedviacommondatastructuresandvocabularies, wasestablished.Itwasusedtoformatveadministrativedatabases:theMedicare ChronicConditionWarehouse,theThomsonReutersMarketScan¨MedicaidMultistate,MedicareSupplemental,andCommercialdatabases,andtheHealthcare CostandUtilizationProjectNationalInpatientSampledatabase.Aweb-basedUser-InterfacewasdevelopedthatcapturesthelogictypicallyrequiredbyCERmethodsandcapitalizesonthelongitudinalityofadministrativedata.Toolsweredevelopedtoallowuserstosearchtaxonomiestoselectparticulardrugs,diagnoses, orproceduresbytypinginsubstringsofthenumericcodesortextualdescriptions. Thetoolallowsresearcherstoapplyenrollmentanddemographicconstraintsand createvariables.CERstudieswereconductedincludingacomparisonofatrialbrillationtreatmentwithrateorrhythmcontrolmedications. RESULTS: Thetool alloweduserstoquicklydeneastudysample.Flowdiagramsgraphicallyillustratedtheattritionofthesamplesizeandvisualizationoftreatmentandoutcomes. EmbeddedSASproceduresenabledreportingandanalysisofcomparisonpopulations.Theanalysesrevealedahigherrateofcoronaryarterydiseaseandheart failurepriortodruginitiationamongtheamiodoroneversusthecalciumchannel blockerpopulationandahigherrateofpost-druginitiationacquiredhypothyroidism,andpulmonarydiseaseamongtheamiodoroneversusthecalciumchannel blockerpopulation. CONCLUSIONS: Newdatadesignsandsoftwareanalytictools mayallowclaimsdatabasestobemoreefcientlyleveraged.Thetooldevelopedfor thisprojecthasthefollowingadvantages:1)allowsforasubstantialportionofthe researchexploration,hypothesisgeneration;andstatisticalanalysistobeperformedinreal-timeusingaweb-basedinterface;2)improvesthespeedofresearch; and3)allowsaccesstoamultipayerdatabase. CO4 POTENTIALCOSTSAVINGSFROMCOMPARATIVEEFFECTIVENESSRESEARCH: LESSONSFROMCOURAGESTUDYBonakdartehraniA,HowardD EmoryUniversity,Atlanta,GA,USAOBJECTIVES: Duringthedebateoverhealthreform,comparativeeffectivenessresearchwastoutedasarelativelypainlessapproachtoreducingcosts.Acomparativeeffectivenessstudyoftwotreatmentswillndeitherthatthecostliertreatmentismoreeffectiveorisnotmoreeffectivethanalessexpensivealternative. Studiesthatreportnegativeresultshavethepotentialtoreducecosts,butonlyif ndingsaffectclinicalpractice.Oneconcernisthatthesamefactorsthatpromote rapidadoptionofnewtherapiesintheU.S.mayretardtheabandonmentofexistingtechnologiesfoundtobeineffective. METHODS: TheCOURAGEtrialfoundthat optimalmedicaltherapyisaseffectiveaspercutaneouscoronaryintervention(PCI) forpatientswithstableangina.PCIreferstostentingandangioplasty.Thetrialwas publishedandwidelypublicizedinearly2007.WeevaluatetrendsinPCIvolume pre-andpost-COURAGEbyindicationusing1)100%samplesofoutpatientand inpatientdischargedataforCalifornia,Florida,NewJersey,andMaryland,2)a100% sampleofdischargedataforVeteran'sAdministrationhospitalsand3)datafroma proprietarycardiaccatheterizationlaboratoryregistryat15hospitals. RESULTS: Betweenthefourthquarterof2006andthefourthquarterof2007,PCIvolumein California,Florida,NewJersey,andMarylandamongpatientswithoutseriouscoronarydiseasedeclinedfromapproximately17,000to13,000procedures(an18% decline).Therewasonlya5%declineamongpatientswithunstableangina,who werenotincludedinCOURAGE.Wefoundsimilarpatternsintheotherdatasets. CONCLUSIONS: PublicationoftheCOURAGEtrialhadanimpactonPCIvolume. ManypatientswithstableanginacontinuetoreceivePCI.Theresultsareconsistentwiththeviewthataslongasthehealthsystemisconguredaroundprocedural-basedmedicine,theimpactoftrialswhichndthatmedicaltherapyisas effectiveasinvasiveprocedureswillbemodest. PODIUMSESSIONI: EFFECTSOFDRUGMANAGEMENTPROGRAMSONPATIENTS DM1 IMPACTOFAPHARMACYREFILLMANAGEMENTSYSTEMONOUTCOMESIN ENDSTAGERENALDISEASE(ESRD)PATIENTSRubinJL1,WilsonSM1,GolombJ2 1DaVitaClinicalResearch,Minneapolis,MN,USA,2DaVitaRx,SanBruno,CA,USAOBJECTIVES: Indialysispatients,boneandmineral(phosphorous,calcium)and regulatoryhormones(parathyroidhormone(PTH))becomedysregulated,increasingriskoffractures,cardiaceventsanddeath.Firstlinetreatmentisalowphosphorusdietandprescriptionphosphatebinders.Weexaminedtheimpactofarell managementsystem(RMS)-whichhelpspatientsproactivelymanagetheirrells usingpredictivealgorithmsandrellremindersforprescriptions-onserumphosphorus,calciumandPTHinpatientsprescribedphosphatebindermonotherapy. METHODS: Datafromalargedialysisorganizationwereusedtoidentifydialysis patientsprescribedmonotherapyphosphatebinderbetween1/1/2008-9/30/2010 withatleast6monthsoffollow-up.PatientsenrolledintheRMSwere1:1propensityscorematchedtopatientsnotenrolledutilizingage,race,gender,dialysis vintage,bodymassindex,baselinelaboratoryvalues(albumin,calcium,Kt/V, phosphorus,PTH,normalizedproteincatabolicrate),Charlsoncomorbidityscore, andothercomorbidconditionscommonlyassociatedwithESRD.Thematched cohortswerecomparedonthepercentmeetingguidelinerangesforphosphorus (3.5-5.5mg/dL),correctedcalcium(8.4-9.5mg/dL)andPTH(150-600pg/mL).Values wereassessedoverthe6-monthsfollowingtherstphosphatebinderprescription. Differencesbetweengroupsweretestedusingchi-squareforproportions. RESULTS: 3,247RMSpatientsmettheinclusioncriteriaandwerematched1:1toa cohortofnon-RMSpatients.Therewerenosignicantdifferencesbetweenthe groupsonanybaselinevariables.PatientsenrolledintheRMSweremorelikelyto beintargetrangeoverthe6monthperiodonall3measures:phosphorus(58.0%vs 55.1%);correctedcalcium(74.3%vs.69.1%)andPTH(80.5%vs.77.2%),comparedto propensitymatchedcontrols.Alldifferencesweresignicantatthep 0.05level. CONCLUSIONS: ResultsindicatethatparticipationinapharmacyRMSisassociatedbetterlaboratoryoutcomesfordialysispatients.A2VALUEINHEALTH14(2011)A1–A214

PAGE 11

DM2 12-MONTHOUTCOMESOFAPHARMACIST-PROVIDEDTELEPHONE MEDICATIONTHERAPYMANAGEMENT(MTM)PROGRAMMoczygembaL1,BarnerJC2,GabrilloE3 1VirginiaCommonwealthUniversity,Richmond,VA,USA,2UniversityofTexasatAustin, Austin,TX,USA,3Scott&WhiteHealthPlanPrescriptionServices,Temple,TX,USAOBJECTIVES: DetermineifMedicarePartDbeneciarieswhoreceivedtelephone MTMserviceshad:1)Decreasedmedication/health-relatedproblems(MHRPs);2) Improvedmedicationadherence;and3)DecreasedtotalPartDdrugcostswhen comparedtoacontrolgroup. METHODS: PartDbeneciariesfromaTexashealth planparticipated.TheAndersenModelwasthetheoreticalframework.Independentvariableswere:predisposingfactors(age,gender,andrace);andneedfactors (numberofmedicationsandchronicdiseasesandmedicationregimencomplexity (MRC).Thehealthbehavior(intervention)wasMTMutilization.Outcomeswere change(frombaselineto12-monthfollow-up)in:1)NumberofMHRPs;2)Medicationadherencemeasuredbythemedicationpossessionratio(MPR);and3)Total drugcosts.Descriptiveandinferentialstatisticalanalyseswereconducted. RESULTS: Theintervention(n 60)andcontrol(n 60)groupswerenotstatistically differentregardingage(71.2 7.5vs.73.9 8.0),medications(13.0 3.2vs.13.2 3.4), chronicdiseases(6.5 2.3vs.7.0 2.1)orMRC[(21.5(range843)vs.22.8(range 942.5)],respectively.Themajority(51%)weremaleintheinterventiongroupbut only28%weremaleinthecontrolgroup(p 0.009).Atbaseline,4.8 2.7(interventiongroup)and9.1 2.9(controlgroup)MHRPswereidentiedand2.2 2.0and 7.3 3.0MHRPsremainedatthe12-monthfollow-up,respectively.Multivariate regressionrevealedthatMHRPsdecreasedsignicantly(p 0.0120)amongtheinterventiongroupwhencomparedtothecontrolgroup.Therewerenosignicant predictorsofchangeinMPR.Totaldrugcosts(changefrombaselinetofollow-up) decreasedby$588 $2,086intheinterventiongroupandincreasedby$207 $1,752 inthecontrolgroup.At-testindicatedthecostdifferencebetweenthe2groupswas signicant(p 0.03),butthemultivariateregressiondidnotindicatesignicant predictors. CONCLUSIONS: AtelephoneMTMprogrampositivelyimpactedMHRPs. Unadjustedcostcomparisonsalsoshowedcostsavingsamongtheintervention group.Futureresearchshouldfocusonunderstandingpredictorsthatimpactadherenceandcost-relatedMTMoutcomes. DM3 IMPACTOFMONTHLYPRESCRIPTIONCAPONMEDICATIONPERSISTENCY AMONGPATIENTSWITHDIABETES,HYPERTENSION,ORHYPERLIPIDEMIAWangCC,WeiD,FarleyJ UniversityofNorthCarolinaatChapelHill,ChapelHill,NC,USAOBJECTIVES: ToevaluatetheeffectofapolicyimplementedintheLouisianaMedicaidprogramwhichcappedreimbursementtoeightprescriptionspermemberper monthonmedicationpersistencyinpatientswithdiabetes,hypertension,or hyperlipidemia. METHODS: Apre-postnon-equivalentcomparatorgroupdesign wasappliedusingMedicaidclaimsdatafrom2001-2003forLouisiana(LA)and Indiana(IN)toidentifypatientswiththespeciedconditionsandtheirmedication persistency.Persistencywasdenedasthenumberofdaystodiscontinuation whichwasidentiedasagapintreatment30daysorlonger.Tocapturepreinterventiontrendsinmedicationpersistency,wecompared"pre-policy"cohorts inLAandINfollowedfortenmonthspriortopolicyadoption(March3,2002to December31,2002)topolicycohortsfollowedtenmonthsafterpolicyadoption (March3,2003toDecember31,2003).Allincidentcohortswereidentiedusinga six-monthwashoutperiod.WeusedCox-proportionalhazardmodelstocompared discontinuationratesinLAandINacrossthepre-policyandpolicyperiodcohorts. RESULTS: Afteradjustingforpatientcharacteristicsandcomorbidconditions,no signicantdifferencesinpersistencywerefoundpriortopolicyimplementation betweenLAandINforanyofthethreechronicconditions.Inthepost-policy period,allcohortshadsignicantlylowerpersistencyinLAthaninIN.Patientsin LAwithdiabetesandhypertensionwere1.38(p 0.03)and2.00(p 0.01)times morelikelytodiscontinuetheirmedicationsatday30ofthefollow-up,respectively.Thehazardratiosdeclinedto1.21and1.67fordiabetesandhypertension patientsrespectivelyafter260days.ThehazardratioofdiscontinuationforpatientswithhyperlipidemiainLAwasconstantly1.31(p 0.01)acrossthefollow-up period. CONCLUSIONS: Patientswithchronicconditionssubjecttomedication capsmaybevulnerabletomedicationdiscontinuation.Policymakersneedtoconsidercarefullywhenimplementingsuchpoliciesonpatientswithchronicconditions. DM4 EVALUATIONOFCLINICALLABORATORY-PHARMACYLINKAGEDECISION SUPPORTINTHEUSEOFPOTASSIUMSUPPLEMENTSYuS,GalanterW,LinFJ,LambertB UniversityofIllinoisatChicago,Chicago,IL,USAOBJECTIVES: Clinicaldecisionsupport(CDS)hasbeenutilizedtolinklaboratory andpharmacydatatooptimizemedicationtherapy.Thisstudyaimedtoevaluate theeffectofsynchronousandasynchronousCDSininpatientsreceivingpotassium supplements. METHODS: Thesynchronousandreal-timeasynchronouslab-pharmacyCDSwasimplementedinour450-bedacademicteachinghospitalinJune 2003.Non-hemolyzedserumpotassium([K ])valuesandmedicationordersfor potassiumsupplements,12monthspriortoandafterCDSimplementation,were analyzed.ACoxproportionalhazardsmodelwasconstructedtoassesstheeffectof CDSinimprovingclinicians'responsetimeatthepresenceofhyperkalemia.Potassium 5.0mEq/Land 5.4mEq/Lwereusedtodenehighnormalandelevated hyperkalemia.Responsetimewasmeasuredfromthetimeoftherstinstanceof hyperkalemiatocancelationofthemedicationorder.Responsetimewascensored atthetimeofaconsequentnormalpotassiumorpatientdischarge. RESULTS: In thepre-CDSperiod,12.5%(1439/11512)ofthepotassiumsupplementorderswere followedbyatleastoneabnormalserumpotassiumvalue;whereasinthepost-CDS period,10.96%(1206/11005)oftheorderswerealertedbyCDS.Whilecontrollingfor patientcharacteristics,postingtimeofthehyperkalemicresults,andseverityof hyperkalemia(5.0mEq/L [K ] 5.4mEq/Lvs.[K ] 5.4mEq/L),potassiumsupplementorderswerecancelledmorerapidlyafterCDSimplementation(hazardratio[HR] 1.14,p 0.007).Moreover,cliniciansrespondedsooneriftheresultwith hyperkalemiawaspostedduring6am-noonvs.midnight-6am(HR 1.36,p 0.001), whiletheresponsetimewaslongeriftheresultwaspostedduringnoon-6pm (HR 0.73,p 0.004)or6pm-Midnight(HR 0.57,p 0.001).Patientage,sex,race,and theseverityofhyperkalemiahadnosignicanteffectonclinicians'responsetime. CONCLUSIONS: Thesynchronousandasynchronousreal-timelab-pharmacylinkagedecisionsupporthelpedclinicianstomanagepotassiumsupplementsina moretimelymannerinpatientswithhighnormalorelevatedpotassium. PODIUMSESSIONI: EMPLOYEEHEALTH&PRODUCTIVITYOUTCOMESRESEARCH OR1 THEASSOCIATIONBETWEENSELF-PERCEIVEDCOGNITIVEDIFFICULTIESAND LEVELOFDEPRESSIONAMONGEMPLOYEESWITHCURRENTDEPRESSIONLawrenceC1,RoyA2,HarikrishnanV2,YuS2,DabbousOH2 1XcendaAmerisourceBerganConsultingServices,PalmHarbor,FL,USA,2Takeda PharmaceuticalsInternational,Inc.,Deereld,IL,USAOBJECTIVES: Manyfacetsofjobperformancemaybeimpairedbydepression.Impairedperformancebydepressedemployeesmaybeattributedtoself-perceived cognitivedifculties.Thegoalofthecurrentstudywastoassessself-perceived decitsincognitionexperiencedbyemployeeswithdepression. METHODS: Individuals( 18yearsofage)employedfull-timewithdiagnoseddepression(excludingbipolardisorder)completedaWeb-basedcomputer-generated25-minutesurveyinFebruary2010(studypopulationidentiedbyHarrisInteractiveTM).The patientsurveyusedthePerceivedDecitsQuestionnaire(PDQ)toassessself-perceiveddifcultiesinmemory,attention,planningandorganization,andconcentrationusinga0-20scale,wherehigherscoresindicategreaterimpairment.The PatientHealthQuestionnaire(PHQ-9)wasusedtoassessdepressionseverity.The impactofdepressiononthePDQscoreswasassessedusingatrendtestbasedonan analysisofcovariancewithage,gender,andPHQ-9scoreasindependentvariables. RESULTS: Atotalof1051employeeswereincludedintheanalysis(58%female, meanage47yrs,and38%heldprofessionalemployment).PHQ-9scoresindicated 423(40.25%)employeeswithnodepressionsymptoms,319(30.35%)withmilddepression,166(15.79%)withmoderatedepression,82(7.80%)withmoderatelyseveredepression,and61(5.80%)withseveredepression.PDQscoresshowedthat perceivedcognitivefunctioningworsenedprogressivelywithincreasingseverityof depressionsymptoms(p 0.0001).PDQscoresshowedthemostimpairmentinthe attention/concentrationandplanning/organizationscalesintheseverelydepressed(12.26and12.25,respectively)comparedwithnon-depressedsubjects(4.45 and3.75,respectively). CONCLUSIONS: Infull-timeemployeesexperiencingdepression,self-perceivedcognitivedifcultiesworsenedwithincreasingseverityof depressivesymptoms. OR2 ASSESSINGTHERELATIONSHIPBETWEENMEDICATIONADHERENCEAND EMPLOYEEPRODUCTIVITYLoeppkeR1,HaueV2,JinnettK3 1U.S.PreventiveMedicine,Inc.,Brentwood,TN,USA,2Alere,Rosemont,IL,USA,3Integrated BenetsInstitute,SanFrancisco,CA,USAOBJECTIVES: Thisstudyexaminedhowadherencetoprescribedmedicationsfor specicchronicconditionsaffectslostworktimeandon-the-jobperformanceinan employedpopulation. METHODS: Patientsaged18to64,withatleastoneoffour conditionsCAD,hypertension,TypeIIdiabetes,ordepressionwereidentied fromclaimsdataforemployeesofveemployers.Patientshadalsocompleteda healthandproductivityassessment(HPA).Weconducteddescriptiveandregressionanalysesofthemergedclaims(12-monthspre-HPA)andHPAdatatoassess whethermedicationadherence,comorbidity,healthrisksandotherfactorswere associatedwithworkplaceproductivity.AdherencemeasuresincludedMedication PossessionRatio(MPR)andadichotomousmeasurethatassigned"adherent"patientswithanMPRgreaterthan80%. RESULTS: Thestudypopulationshowedhigh ratesofadherence:meanMPRrangedfrom0.70to0.79acrossconditions.Better adherencewasassociatedwithfewerlosthoursduetoabsenceforCADpatientson statintherapyanddiabetespatientsonstatinorACE/ARBtherapy.ForDiabetes patientsoninsulinorliketreatment,thosewithMPRsof0.80orhigherhadhigher jobperformanceandfewertotalworkhoursmissedthanindividualswithMPRs lowerthan0.80.Unexpectedly,betteradherencebyCADpatientsonanti-platelet therapywasassociatedwithpooreron-the-jobperformance.Thepresenceofcomorbiddiseasesandhigherhealthriskstatuswasassociatedwithincreasedabsenceandpoorerjobperformanceacrossthefourconditions.Thoseintheprofessional/executivejobclasstypicallyhadlessabsencefromworkbutlowerjob performance. CONCLUSIONS: Despitehighratesofadherenceinthispopulation, signicantmedicationtreatmenteffectswereobserved,aswereconsistenteffects ofcomorbidities,healthrisksandjobclassonlostworktime.Bettertargetingof treatmentbyjobclassforindividualswithhighhealthrisksandcomorbiditiesmay helpreducehealthrisksandimproveproductivityoutcomes.A3VALUEINHEALTH14(2011)A1–A214

PAGE 12

OR3 THEDIRECTANDINDIRECTCOSTSASSOCIATEDWITHHYPOGONADISM AMONGPRIVATELY-INSUREDEMPLOYEESINTHEUNITEDSTATESKaltenboeckA1,FosterS2,ThomasN2,IvanovaJ1,DienerM1,BergmanR1,BirnbaumH3, SwindleR2 1AnalysisGroup,Inc.,NewYork,NY,USA,2EliLillyandCompany,Indianapolis,IN,USA,3AnalysisGroup,Inc.,Boston,MA,USAOBJECTIVES: Comparedirectandindirect(workloss)costsbetweenprivately-insuredU.S.employeeswithhypogonadism(HG)anddemographicallymatchedcontrolswithoutHG. METHODS: Maleemployees,ages35-64,with 2HGdiagnoses (ICD-9-CM:257.2x)or 1HGdiagnosisand 1claimfortestosteronetherapybetween1/1/2005-3/31/2009wereidentiedfromaprivately-insuredclaimsdatabase (N 12,000,000).TheindexdatewasdenedasthemostrecentHGdiagnosiswith continuouseligibility 1yearbefore(baselineperiod)and1yearafter(studyperiod).EmployeeswithHGwerematched1:1onage,region,employmentstatus,and indexyeartocontrolswithoutHG.Descriptiveanalysescompareddemographic characteristics,comorbidities,resourceutilization,directcosts(reimbursements toprovidersformedicalservicesandprescriptiondrugs)andindirectcosts(disabilityandmedically-relatedabsenteeism)inatedto$2009.Multivariateanalyses adjustingforbaselinepatientdifferenceswereusedtoestimaterisk-adjusted costs. RESULTS: 4,269HGemployees,meanage51,withmatchedcontrolsmet inclusioncriteria.Comparedwithcontrols,HGemployeeshadhigherbaselinecomorbidityrates:hyperlipidemia(50.2%vs.25.3%),hypertension(37.7%vs.21.1%), back/neckpain(32.0%vs.15.7%),andHIV/AIDS(7.1%vs.0.3%)(allp 0.0001).HG employeeshadhigherstudyperiodratesofinpatientstays(10.8%vs.5.2%),EmergencyDepartmentvisits(27.5%vs.16.3%),outpatientvisits(100.0%vs.76.7%),prescriptionmedicationuse(95.7%vs.68.3%),andhighermeanworklossdays(19.3vs. 8.8)(allp 0.0001).HGemployeescomparedwithcontrolshadhighermeanstudy perioddirect($10,914vs.$3,823)andindirectcosts($3,204vs.$1,450);HG-related directcostswere$832.HGemployees'costsremainedhigherafteradjustingfor baselinedifferences(direct:$9,291vs.$5,248;indirect:$2,729vs.$1,840)(all p 0.0001). CONCLUSIONS: EmployeeswithHGhadhighercomorbidityratesand costscomparedwithdemographicallymatchedcontrols.GiventhelowHG-related costs,themaindriverofoverallcostsamongHGpatientsmaybetheircomorbidity burden. OR4 ASSOCIATIONSBETWEENJOBLESSNESSANDALL-CAUSEHEALTHSERVICES UTILIZATIONINDIABETICWORKINGAGEADULTSINTHEUNITEDSTATESDavis-AjamiML1,NahataM1,SeiberE1,ReardonG2,BalkrishnanR3 1OhioStateUniversity,Columbus,OH,USA,2Informagenics,LLC,Worthington,OH,USA,3UniversityofMichigan,AnnArbor,MI,USAOBJECTIVES: Toassessassociationsbetweenjoblessnessandall-causeemergency department(ED),hospitalization,outpatientandofce-basedhealthservicesutilizationinUSdiabeticworking-ageadults. METHODS: Thisretrospectivelongitudinalpaneldesignusednationally-representative20012007pooledpublicdomaincompletepaneldatafromtheMedicalExpenditurePanelSurvey(MEPS). EligibleMEPSrespondentsaged2459yearswithanICD-9-CMdiabetesdiagnostic code"250",wereincluded.Thosewithpregnancydiagnosticcodes,seasonaljob status,orprescribedinsulinwereexcluded.Subjectsreportinganemployment statusas"notemployedwithnojobtoreturnto"wereclassiedasjobless.Using MEPSweightstoaccountforthecomplexsurveydesign,logisticregressionmodels estimatedassociationsbetweenjoblessnessandthelikelihoodofutilization.Negativebinomialregressionmodelsassessednumberofutilizations.TheTaylorlinearizationmethodestimatedvariance. RESULTS: 2,678subjects(means:age48.7 years[SD0.28],BMI31.5[SD0.30],CharlsonComorbidityIndex0.369[SD0.12])met eligibilitycriteria.Comparedtothoseemployed,joblessnesssignicantlyincreasedtheoddsforall-causeEDutilization64%(OR1.64,p 0.007)outpatient visits46%(OR1.46,p 0.011)andofce-basedvisit45%,(OR1.45,p 0.009).JoblessnesswasassociatedwithhigherlogsofexpectedcountsforEDvisits( 0.43, p 0.005)outpatientvisits( 0.49,p 0.002),andofce-basedvisits41%( 0.41, p 0.000).Thefollowingcovariatesshowedsignicant(p 0.05)associationsacross ED,hospitalization,outpatientandofce-basedutilizationsectors:familysize,age, theCharlsonComorbidityIndex,andthepresenceofoneormorediabetesrelated complication.HispanicethnicitywasassociatedwithfewerEDvisits( 0.51,p 0.001),andfewerhospitalizations( 0.41,p 0.026)thanotherethnicities. Thoughnon-signicant,comparedtouninsuredindividuals,privateandpublic healthinsurancecoveragealsoshowedincreasedoddsandnumberofutilizations. CONCLUSIONS: Joblessnesswassignicantlyassociatedwithincreasedall-cause healthservicesutilization. PODIUMSESSIONI: CASESTUDIESINADDRESSINGSELECTIONBIAS SB1 COMPARISONOFDIFFERENCE-IN-DIFFERENCE,PROPENSITYSCOREMATCHING ANDINSTRUMENTALVARIABLESINESTIMATINGCOSTDIFFERENCES BETWEENTWOCOHORTSCaoZ,SongX ThomsonReuters,Cambridge,MA,USAOBJECTIVES: Endogeneityisacommonprobleminretrospectiveclaimsdatastudiesbecausepatientsinclaimsdatawerenotassignedtotreatmentbyrandomization.Propensityscorematching(PSM),instrumentalvariables(IV),anddifferencein-difference(DID)havebeenusedtocontrolforselectionbiasinevaluatingthe impactoftreatmentonoutcomemeasures.Thisstudycomparestheestimated incrementalcostsbetweentypicalandatypicalantipsychoticmedicationusersin patientswithschizophreniausingthesethreemethods. METHODS: Patientsages 18-64yearsoldwithatleastoneprescriptionoftypicaloratypicalantipsychotic medicationandatleastonediagnosisclaimofschizophrenia(ICD-9-CMdiagnosis 295.xx)within90daysoftheantipsychoticmedicationwereidentiedinMarketScan ¨ Multi-StateMedicaidDatabase20022009.Theindexdatewastherstprescriptiondateofantipsychoticmedication.Allpatientshad 12-monthcontinuousenrollmentpriortoandposttheindexdate.Outcomesweretotalall-cause expendituresandpsychiatric-relatedexpendituresduringthe12-monthfollowup. Theincrementalcostsassociatedwiththeuseoftypicalantipsychoticmedication versusatypicalmedicationwasestimatedusingsixregressionmodelsthreeestimationmethods(GLM,IV,andDID)ontwosamples(matchedandnon-matched sample).IVwasprescribingphysicians'characteristics. RESULTS: Atotalof447 atypicaland4544typicalantipsychoticusersmetallstudycriteria,withameanage of38.9.Fiveoutofthesixmodels(withtheexceptionoftheDIDmodelonthe unmatchedsample)didnotndasignicantrelationshipbetweentypesofantipsychoticmedicationsandtotalcosts,andveoutofthesixmodels(withthe exceptionoftheIVmodelonthefullunmatchedsample)foundatypicalmedicationusershadsignicantlyhigherpsychiatric-relatedcoststhantypicalmedicationusers. CONCLUSIONS: ThePS-basedapproachcombinedwiththeDIDorIV methodsmaybebetterthaneachapproachalone. SB2 ZEROSANDNON-REPORTEDHEALTHCAREANDWORKPLACEPRODUCTIVITY DATA:ANAPPLICATIONOFTWO-STAGEESTIMATIONTECHNIQUES MEASURINGINPATIENTCOSTSANDABSENTEEISMASSOCIATEDWITHLOW BACKANDNECKPAINSimonsWR1,ChowW2,BiondiD2,BensonC2,KimM2 1GlobalHealthEconomics&OutcomesResearchInc.,Summit,NJ,USA,2Ortho-McNeilJanssen ScienticAffairs,LLC,Raritan,NJ,USAOBJECTIVES: Abundanceofzerovaluesiscommonlyobservedincostdataresultinginskeweddistribution.Thisanalysismeasuredtheinpatientcostandworkplaceabsenteeismassociatedwithlowbackandneckpainanddemonstratedthe consequencesofignoringzerosininpatientcostandunreportedabsenteeism. METHODS: Weusedemployer-basedclaimsfromtheThomsonMarketscanRe searchDatabase(2007),adatabaserepresentingapproximately100payersofinsured employeescontaininghealthandproductivitymanagement(HPM)andhealthcare utilizationdata.Adultinsuredemployeeswithcontinuouseligibilityin2007were included.TheICD-9codesidentiedmedicalconditionsincludinglowbackandneck painwithout(nociceptivepain,NOCI)orwithaneuropathiccomponent(mixedpain, MIXED).Ordinaryleastsquares(OLS)andtwo-stageTobitanalysesevaluatedthemarginalinpatientcostswhileOLSandHeckman'sSelectionBias(HSB)wereappliedto absenteeismdata.EstimatedinpatientcostsandabsenteeismusingOLSversustwostagetechniqueswerecompared. RESULTS: Atotalof2,046,332employees (male 59.2%;meanage40.2 11.6years)wereanalyzed.Hypertension(9.8%),NOCI (9.5%),diabetes(3.7%),MIXED(3.0%)anddepression(1.1%)werethemostprevalent medicalconditionsamongtheseemployees.1,976,952(96.6%)employeeshadnoinpatientepisodes,thus,withnoinpatientcosts.Meaninpatientcostfortheentirestudy populationwas$537.45(median $0)versus$15,851.93(median $8,302.20)among thosewithinpatientepisodes.TheincrementalinpatientcostsassociatedwithMIXED andNOCIwere$1,333.02 26.67and$328.36 15.63usingOLSversus$2,478.97[95%CI: 2,148.502,811.16]and$1,242.41[95%CI:1,020.101,469.18]usingthetwo-stageTobit. Unreportedabsenteeismoccurredin80%oftheemployees.AnnualabsenteeismassociatedwithMIXEDandNOCIusingOLSwere5.25 0.21and4.06 0.35comparedto 45.92 1.06and16.33 2.01hoursusingtheHSBtechnique. CONCLUSIONS: Ignoring zerosincostdataandunreportedabsenteeismmayresultinsubstantialunderestimationofinpatientcostandworkplaceabsenteeismassociatedwithlowback andneckpain. SB3 INNOVATIVEDESIGNFORACOMPARATIVEEFFECTIVENESSSTUDYOF SCHIZOPHRENIATREATMENTS:ANALYSISOFRECORDREVIEWDATA INCORPORATINGRANDOMIZATIONANDPROPENSITYSCOREMATCHINGMcCarrierKP1,,DurkinMB2,DiraniR2,MarkowitzM2,SlabaughSL2,MartinML1 1HealthResearchAssociates,Inc.,Seattle,WA,USA,2Ortho-McNeilJanssenScienticAffairs, LLC,Titusville,NJ,USAOBJECTIVES: Abstractionofhospitalrecordsiscurrentlyunderwayatinpatient psychiatricfacilitiesacrosstheUnitedStatestofacilitatealargecomparativeeffectivenessstudywiththefollowinggoals:1)toobserveanddescribere-hospitalizationpatternsamongpatientswithschizophrenia,and2)tocomparere-hospitalizatonoutcomesbetweenpatientsreceivingpaliperidonepalmitateandthose receivingoralatypicalantipsychotics.Thisabstractisintendedtodescribethe innovativedesignofthisstudy. DESIGN/METHODS: Thisnaturalisticrecordreview studyincorporatesseveralnoveldesignelementsandauniquetwo-phaseabstractionprocess.Intherstphase,allpatientswithaqualifyinginpatienthospitalizationforschizophreniaareidentiedandbasicdemographic,clinical,andtreatmentdataisabstracted.Fromthispoolofpotentially-eligiblepatients,twogroups areidentied;1)patientsdischargedonpaliperidonepalmitate,and2)patients dischargedonoralatypicalantipsychotics.Randomsamplesofpatientsaredrawn fromeachofthesegroupsanddesignatedforfulldatacollectioninphasetwo.In thesecondreviewphase,thesedesignatedrecordsarefurtherabstractedtocollect detaileddemographicvariables,hospitalizationandtreatmenthistory,condition severity,comorbidconditions,anddischargecharacteristics.Thesevariablesare usedtomodelpropensityscoresforreceiptofthetargetdrug,andidentifytwo propensity-matchedcohortsforthesubsequentcomparativeeffectivenessanalysis.Pilottestingatthreehospitalshasconrmedtheavailabilityofkeydataele-A4VALUEINHEALTH14(2011)A1–A214

PAGE 13

mentsandsupportsthefeasibilityofthisapproach. CONCLUSIONS: Thisnovel designseekstoaddressthebiasinherentinobservationalresearchthroughthe impositionofrandomization.Byseparatingdatacollectionintoapreliminary phasecollectingonlyvariablesneededfortreatmentidenticationandrandomizationandaseparatefullreviewofonlytheserandomly-selectedpatientrecords, chartabstractionburdenisminimized.Furthermore,theuseofpropensityscore matchingtocreatetwomatchedcohortsforcomparisonallowsgreatercontrolof potentialconfoundinginanalysesoftreatmenteffect. SB4 AMETHODOLOGYFORASSESSINGTREATMENTEFFECTINTHEPRESENCEOF DISEASESEVERITYANDCOMORBIDITYINRETROSPECTIVEOBSERVATIONAL STUDIESKiriVA PAREXELInternational,Uxbridge,London,UKOBJECTIVES: Therearemanyexamplesinhealthoutcomesresearchwhereinadequatecontrolforcomorbidityinuencehasresultedineffectestimatesconfoundedbydiseaseseverity.Selectionbiasisacommonfeatureofdatafromroutinehealthcaresettingwherethedecisiontogiveaparticulardrugtoapatientwith agivendiseaseisgenerallybasedonpatientcharacteristics,includingdisease condition.Thus,failuretoproperlycontrolforthebiascouldresultinfalseassociations.Propensityscoresmethodologyiscommonlyuseddespiteitslimitations becauseofitspotentialforminimisingtheassociationbetweenexposureandconfoundingfactors.Wedescribeamethodologyforassessingdrugeffectinlongitudinaldatathatminimisesconfoundingbydiseaseseveritygenerallyassociated withobservationalstudies. METHODS: Foraparticularoutcomeofinterest,we obtaintheprolesofratesratiosfromtwosetsofmatchedcohorts.InsetA,patientswithdiseaseXarecomparedwithothersfreeofXintheperiodspriortoand postdiagnosisofX.InsetBwhichinvolvesonlypatientswithdiseaseX,those exposedtotreatmentYarecomparedwiththoseunexposedtothedruginthe periodspriortoandpostexposure.Thetwosetsofprolesarethenassessedusing simpleregressionovertherespectiveperiods.Ineffect,weattempttodisentangle thediseaseandtreatmenteffects.DatafromtheUKGPRDareusedtoassess possibleassociationbetweenaparticularoutcomeandtreatmentinCOPD RESULTS: WefoundevidenceofassociationbetweentheoutcomeandCOPDbut noneforthedrug. CONCLUSIONS: Theproleapproachutilizesthedatacollected overthediseasenaturalhistoryandexposurehistorytoassesstherelationships betweentheoutcomeandboththediseaseandtreatment.Thisisakeystrength oftenignoredwhenresultsarereportedaspointestimates.Bydesign,italsominimisestheeffectofselectionbias. PODIUMSESSIONII: RESEARCHONMETHODS:COST-EFFECTIVENESSANALYSIS CE1 COST-EFFECTIVENESSSENSITIVITYANALYSISMETHODS:ACOMPARISONOF ONE-WAYSENSITIVITY,ANALYSISOFCOVARIANCE,ANDEXPECTEDVALUE OFPARTIALPERFECTINFORMATIONCampbellJ1,McQueenRB1,LibbyA1,BriggsA2 1UniversityofColorado,Aurora,CO,USA,2UniversityofGlasgow,Glasgow,UKOBJECTIVES: Advancedsensitivitymethodsincludingvalueofinformationwere developedtoquantifyoveralldecisionuncertaintyandtoassessthecost-effectivenessofadditionalresearchthatwouldreducethatuncertainty.Ourobjectivewas tocomparetheinformationgainedbyutilizingthreealternativesensitivitymethodswithincreasingcomplexity:asimpleone-waysensitivityanalysis;probabilisticanalysisofcovariance(ANCOVA);andexpectedvalueofpartialperfectinformation(EVPPI)ofinputparameters. METHODS: Wereplicatedandexpandeda publishedHIV/AIDScost-effectivenessMarkovmodel(zidovudinevs.zidovudine pluslamivudineintheUK)usingTreeAge¨.HealthstatesincludedthreeHIV/AIDS statesanddeath.Ouroutcomeofinterestwastheincrementalnetmonetarybenet(INMB)assumingawillingness-to-payof£20,000/QALY.Wegeneratedone-way andprobabilisticsensitivityanalysesoftheINMBusingpublishedinputparameter uncertainties.One-waysensitivityanalysisidentiedthe10mostinuentialparameters.Atotalof10,000MonteCarlodrawswereusedtoestimatetheANCOVA resultsfromthesametenparameters.EVPPIforeachofthesametenparameters wasestimatedspecifying1000innerand1000outerMonteCarlodraws.Weranked theparametersbasedontheirinuenceonvariationforeachsensitivitymethod andcomparedthemusingSpearman'srankcorrelation. RESULTS: MeanINMBwas £9694infavorofcombinationtherapy.Thetwomostinuentialinputswerethe sameacrossallmethods,contributed78%ofvariationinoutcome(ANCOVA),and weretheonlyinputswithnon-zeroEVPPIvalues.Therankorderforthetopten inputsfromallmethodswassimilar(correlation 0.99forone-wayvs.ANCOVA, 0.70forone-wayvs.EVPPIand0.70forANCOVAvs.EVPPI,allp-values 0.05). CONCLUSIONS: Thecorrelationwassignicantbetweenone-wayandmoreadvancedsensitivityanalyses.Althougheachmethodprovidesuniqueinformation, theadditionalresourcesneededtogenerateadvancedanalysesshouldbeweighed, especiallywhentheoutcomedecisionuncertaintyandthereforevalueofinformationislow. CE2 ANOVELWAYOFESTIMATINGCOST-EFFECTIVENESSRATIOSFROMCLINICAL TRIALSWITHMISSINGDATA:ASIMULATIONSTUDYGagnonDD1,EngelhartL2 1ThomsonReuters,SantaBarbara,CA,USA,2DePuy,Inc.,Raynham,MA,USAOBJECTIVES: Inasimulateddataset,evaluateincrementalcost-effectivenessratios (ICERs)adjustedforcovariatesandmissingdatausingthreedifferentregression models.Theregressionparameterofinterestistheincrementalnetmonetary benet(INMB).ModelsareANCOVA,mixedeffects(ME),andjointmixedeffects andlogtime-to-dropout(jointME),aselectionmodel. METHODS: Traditionalcosteffectivenessanalysis(CEA)usestheincrementalcost-effectivenessratio(ICER),a measurewithstatisticalissuesandlimitationswithmissingdata.Regressionanalysiscanestimateincrementalnetmonetarybenet(INMB)andavoidthesestatisticalissueswhileadjustingforcovariatesaswellasmissingdata.Thecost-effectivenessacceptabilitycurve(CEAC)generatedfromafamilyoftheseregressions canidentifyanICERadjustedforthefactorsincludedintheINMBregressions(the ICERisthepointontheCEACwheretheprobabilityofbeingcost-effectiveis50%). Dataweresimulatedtoincludemissingatrandom(MAR)andmissingnotatrandom(MNAR).Simulatedtreatmenteffectprovideda"true"INMBformodelevaluationsthatincludedbias(absolutedifferencefrom"true"),precision(ratioofvariances),andCEACswithwillingness-to-pay()valuesfrom$0to$100K. RESULTS: TheANCOVAandMEmodelsproducedtheleastbiasedestimates.At $50K,bias was$1.3K,$1.4K,and2.3K,andprecisionwas1.27,0.90,and1.24forME,ANCOVA, andjointME,respectively.ThejointMEmodelperformedbestwhenmissingness washigh. CONCLUSIONS: OncetheCEACshadbeengenerated,derivingICERs adjustedforcovariatesandmissingdatafromthoseCEACsbaseduponINMBregressionsprovedeasyandfeasible.Themodelsusedinthissimulationanalysis performeddifferentlyunderalternativemissingnessconditionsandweresensitive tononresponsemechanisms.Allestimateswerepoorwhenmissingnesswashigh; suggestingpreventionofmissingdatashouldbeagoalofresearch. CE3 COST-EFFECTIVENESSANALYSISANDBUDGETIMPACTASSESSMENT:A GRAPHICALWAYTOCOMBINETHETWOFORTHEAIDOFDECISION-MAKERSPauldenM,PhamB UniversityofToronto,Toronto,ON,CanadaOBJECTIVES: Cost-effectivenessanalysis(CEA)hastraditionallybeenseenasa meansofsatisfyingaspecicandexplicitsocialobjectivesubjecttoaxedbudget constraint.Asaresult,existingCEAmethodslargelyignorebudgetimpactconsiderationsinhealthsystemswherebudgetsarenotxed.Inparticular,noneofthe traditionalmethodsofpresentingresults(suchasthecost-effectivenessplane, ICERtablesandCEACgraphs)canbeusedtosummarizetheresultsofaCEAand budgetimpactassessmentsimultaneously.Ourobjectivewastodevelopsucha methodinamannerwhichismeaningfultodecisionmakers. METHODS: We presentanovelwayofcombiningcost-effectivenessandbudgetimpactconsiderationsintoasinglegraph.Todothis,wedisaggregatetheincrementalcostsofthe newtechnologyintothosewhichfallonthehealthbudgetanddisplaceother technologies(resultinginforgonehealth)andthosewhichleadtoanexpansionof thehealthbudget(resultinginanetbudgetimpact).Theincrementalhealthbenetofthetechnologyandanyforgonehealtharecombinedtogivethenethealth benetofthetechnology,whichisplottedagainstthenetbudgetimpact. RESULTS: Ourmethodclearlyrevealsthetrade-offbetweenthecost-effectivenessand budgetimpactofthetechnologyinquestion.Thistrade-offissimultaneouslyrevealedacrossarangeofplausiblevaluesofthecost-effectivenessthreshold. CONCLUSIONS: Decisionmakerswhoareconcernedwithboththecost-effectivenessandbudgetimpactofnewtechnologieshavetendedtoconsidereachofthese separately,withtheinherenttrade-offbetweenthetwoblurredintheprocess.Our proposedmethodmakesthistrade-offexplicitanddoessoacrossarangeof thresholdvalues,enablinganalyststoprovidemeaningfulinformationtodecision makerswhilerespectingdecisionmakers'authorityindeterminingtheappropriatethresholdtouse. CE4 USINGDYNAMICTRANSMISSIONMODELSTOESTIMATETHECOST EFFECTIVENESSOFVACCINES:FOURDIFFERENTMETHODSANDTHEIR RELEVANCEFORDECISIONMAKERSMauskopfJ1,TalbirdSE1,StandaertB2 1RTIHealthSolutions,ResearchTrianglePark,NC,USA,2GlaxoSmithKlineBiologicals,Wavre, BelgiumOBJECTIVES: Tocomparethedifferencesinthemethodsusedtoestimatethe cost-effectivenessofvaccinationprogramsusingtheclinicaloutcomesfromdynamictransmissionmodels. METHODS: Atargetedelectronicliteraturesearchof titlewordsinthePubMeddatabaseswasperformedtoidentifystudiespublished since2000thatincludedadescriptionofthemethodsandpresentationofresultsof cost-effectivenessanalysisofvaccinationprogramsbasedondatafromdynamic transmissionmodelsforanyinfectiousdisease.Furtherstudieswereidentiedin thebibliographiesoftheinitialsetofpapers. RESULTS: Informationwasabstracted from29paperspresentingcost-effectivenessanalysesofvaccinationprogramsfor inuenza,HPV,varicellavirus,pertussis,meningococcalmeningitis,rotavirus,H. pylori,andhepatitisA.Bothcohortandpopulation-basedestimatesofcost-effectivenesswerepresented.Thepopulation-basedestimateshadvariabletimehorizonsfrom1yearforinuenzaorpertussis(thesteadystateyear)upto100yearsfor HPV,varicella,andmeningococcalvaccination.Allcohortanalysesusedalifetime timehorizon.Fourmethodtypesfortheestimationandpresentationofacosteffectivenessratiowereidentied:1)averagepopulationvalues(costsandbenets)overalongtimehorizonassumingacontinuingvaccinationprogram(20papers)2)averagepopulationvaluesoveralongtimehorizonassumingalimited durationvaccinationprogram(1paper)3)populationvaluesforthesteady-state yearonly(1paper)and4)cohortvalueswithalifetimehorizon(7papers). CONCLUSIONS: Thevariabilityoftheestimationframework(populationorcohort) andtimehorizonusedaswellasthevariabilityinotherinputparametersobservedA5VALUEINHEALTH14(2011)A1–A214

PAGE 14

inthereviewillustratetheproblemsthatmaybeencounteredincomparingcosteffectivenessestimatesofdifferentvaccinationprogramsamongthemselvesas wellaswithotherpreventionortreatmentinterventions. PODIUMSESSIONII: RESEARCHONMETHODS:DATABASEANALYSIS DS1 INTEGRATINGDATASOURCESTOCONDUCTCOMPREHENSIVEONCOLOGY BASEDOUTCOMESRESEARCHAlbrightF1,BolluV2,KuoKL3,RaimundoK1,BarneyR3,StenehjemD3,BrixnerD3 1UniversityofUtahCollegeofPharmacy,SaltLakeCity,UT,USA,2NovartisPharmaceuticals Corporation,EastHanover,NJ,USA,3UniversityofUtah,SaltLakeCity,UT,USAOBJECTIVES: Individualdatasourcescontainnon-integrateddatacomponents neededtoassessoutcomes,resourceuse,andcostsincancerpatients.Thiswork describesmethodologytointegratedisparateelectronicdatasourcesinchronic myelogenousleukemia(CML)patientswithacommonidentier(CI). METHODS: A CMLPatientcohortfromtheHuntsmanCancerInstitutewascreatedbyextracting informationacrosstheUtahCancerRegistry;theUtahPopulationDatabase(UPDB); andtheEnterpriseDataWarehouse,includingCernerinpatientandEPICambulatorycareclinicdata.Medicationusewasfrominpatientmedicationorders.A uniquepatientindexidentierlinkeddisparaterecords. RESULTS: Atotalof602 patientswereidentiedbyICD-9diagnosiscodeforCML(250.1,205.10-12)from 1995through2009,medianage 51,42.6%female.Ofthese598(99.3%)werelinked totheUPDBand245hadastatedeathcerticate.CharlsonComorbidityIndex(CCI) analysis( / 90days)identied232(38.5%)subjectswithascoreofzero,199 (33.1%)with1-3,99(16.4%)with4-6,47(7.8%)with7-9and25(4.2%)withascoreof 10-17(median 2,mean 2.6,andSD 3.1).Inpatientadmissiondatawasavailable for380(63.1%)patients,withatotalof267CMLrelateddrugorders.Procedures wereobservedfor531(88.2%)patients.Labresultswereavailablefor564(93.7%) subjects.Ofthose,BCR/ABLbiomarkerresultswereavailablefor210(37.2%ofall labresults)patients. CONCLUSIONS: Integratingdataacrossdifferentdatasources inanacademichealthcarecenterwithaNationalComprehensiveCancerNetwork hospitalcanprovidecomprehensivehealthcaredata.Thismethodologymayinuencetheevolutionofelectronichealthrecords,asadataresourcetoolforoutcomesdata,resourceuseandcostutilizationacrosscomplexdiseasestatessuchas CML.Futureresearchwillexpandondrugdatasourcingandevaluatethemedical recordnotestoevaluateCMLspecicoutcomes. DS2 AVALIDATIONSTUDYOFALGORITHMSFORIDENTIFYINGMETASTATIC BREAST,LUNG,ORCOLORECTALCANCERINADMINISTRATIVECLAIMSDATAWhyteJL1,Engel-NitzNM2,TeitelbaumA2,GomezReyG2,KallichJ1 1Amgen,Inc.,ThousandOaks,CA,USA,2i3Innovus,EdenPrairie,MN,USAOBJECTIVES: Incancerresearchusingclaimsdata,identifyingmetastasesisoften essentialyetdifcult.Theobjectiveofthisstudywastoexaminethevalidityof algorithmsidentifyingmetastaticbreast(BC),lung(LC),orcolorectal(CRC)cancer inhealthcareclaimsdata. METHODS: Aproprietaryclinicalcancerdatabasecontainingphysician-reportedclinicaldataonpatientswithBC,LC,orCRCbetween January1,2007andMarch31,2010,waslinkedtoclaimsdata.Inclusionrequired healthplanenrollment 3monthspriortotheinitialclinicalcancerdiagnosis date.Un-validatedclaimsalgorithmsfrompreviousresearchwereidentied.A genericmetastaticalgorithmwithallmetastaticICD-9codesandtumor-specic variationsofthealgorithmwereassessedforvalidity.Thealgorithms'validity versustheclinicallyreportedmetastaseswastestedusingsensitivity,specicity, positivepredictivevalue(PPV),andnegativepredictivevalue(NPV). RESULTS: Of 14,480patientsinthedatabase,4631BC(meanage53.6yr),2449LC(meanage62.9 yr),and2058CRCpatients(meanage58.3yr)metinclusioncriteria.Metastasesat diagnosiswererecordedin8.0%(371)ofBC,49.2%(1204)ofLC,and25.7%(528)of CRCpatients.ThetumorspecicalgorithmforidentifyingmetastaticBChad53.2% sensitivityand98.6%specicity;PPVandNPVwere77.6and95.8.Thelung-specic algorithmhad55.2%sensitivityand85.3%specicity;PPVandNPVwere81.0and 62.6.Similarly,theCRC-specicalgorithmhad59.4%sensitivity,89.8%specicity, withPPV72.9andNPV82.7.Thegenericalgorithmhadlowerspecicityandhigher sensitivityforall3cancersandasignicantlylowerPPVforbreastcancer. CONCLUSIONS: Specicity,butnotsensitivity,washighforalltumor-specicalgorithms.Althoughnottested,bettersensitivitymightbegainedbyincluding chemotherapyinthealgorithmsforsometumortypes. DS3 AVAILABILITYOFLABORATORYRESULTSDATAINACLAIMSDATABASEIN THEUNITEDSTATESHorneLN,MingEE,Doylec AstraZenecaPharmaceuticalsLP,Wilmington,DE,USAOBJECTIVES: Todescribethefrequencyofavailablelaboratoryresultsdataina commercialhealthcaredatabase,amongpatientswhoarebeingtreatedfordiabetesordyslipidemiaandwhohavehaveatleastonedocumentedlaboratoryresult forahemoglobina1c(HbA1c)orlipidtest. METHODS: Thesourcepopulationwas adultscontinuouslyenrolledinalargeU.S.healthplanduring2009withatleast oneCPTcodeforaHbA1corlipidtest.Laboratoryresultsdatawereconsidered availableifLOINCcodesorfreetextidentiedaresultrecordedwithin / 3days oftheCPTclaimdate.Thenalstudycohortincludedonlypatientswithatleast oneresultavailable.Wecalculatedthe1-yearperson-levelpercentofthenumber oftestsorderedin2009thathadresultsavailable.Resultsforeachtestwerestratiedbywhetherthepatientreceivedanantidiabeticorantidyslipidemiadrugduringthesameyear. RESULTS: Overall,aresultwasavailablefor41%ofHbA1ctests and42%oflipidtests.Personswithatleastoneprescriptionclaimforanantidiabeticorantidyslipidemiadrughadmorefrequenttestsrecordedduringthestudy period(HbA1c:mean4.5withdrug,2.0withoutdrug;Lipid:mean3.9withdrug,2.0 withoutdrug).However,resultswerelesslikelytobeconsistentlyavailableamong treatedpatients:44%ofthosetreated(amongwhom70%oftestshadresults),and 39%ofthosenottreated(amongwhom83%oftestshadresults)hadanyresults available. CONCLUSIONS: Whilelaboratorydatamayenhancestudiesconducted inadministrativeclaimsdatabases,resultsmaybeinconsistentlyavailable.Inthis study,amongtreatedpatients,44%hadanylaboratoryresultsrecorded,forwhom resultsweremissingapproximately30%ofthetime.Anevaluationofthecompletenessoflaboratorydatapriortoanystudyisfeasibleandmayhelpunderstand anypotentialbias. DS4 BURDENOFPROOF ... PROOFOFPRINCIPLE:REPLICATIONQUANTIFICATION, REPLICATIONANDVALIDATION ... STANDARDSOFEVIDENCEINOUTCOMES RESEARCHSURROGATEENDPOINTSFORALL-CAUSEMORTALITYSimonsWR GlobalHealthEconomics&OutcomesResearchInc.,Summit,NJ,USAOBJECTIVES: Todemonstratereplicationofthequanticationofrelationshipsbetweensurrogatesandendpointsaswellasreconciliationwithpreviousepidemiologicalstudies;originalstudiesforheartrateasasurrogateforall-causemortality, painmanagementandgastrointestinaladverseevents,andtreatmentfordiabetes andHbA1candHbA1candcomplications. METHODS: Forheartrate,threeepidemiologicalstudiesfromthreecountriesusingaWeibullsurvivalanalysisandGeneralizedEstimatingEquationswereused;namely,theCoronaryArterySurgery Study(CASS),theCopenhagenCityHeartStudy(CCHS)andtheGeneralPractitionerResearchNetwork(GPRN).Theseequationsreproducedameta-regressionand meta-analysisofallavailableplacebo-controlledclinicaltrialswithheartrateasa prognosticfactorforall-causemortality.Forpain,dataconsistedof2005Health CareUtilizationProject(HCUP)andPremier.Logisticregressionswereusedtoobtainevaluateandcompareodds-ratios.Indiabetes,GeneralizedEstimatingEquations(GEE)allowingseriallycorrelatedbehaviorwithrepeatedHbA1creadingat variablefrequenciesanddurationsbetweentheirmeasurement. RESULTS: Heartis consistentlyprognosticforall-causemortality.Moreover,itsquanticationisconsistent,0.00694(P 0.001)inCASSand0.00683(P 0.001)inCCHS(1981-1983)and 0.00717inCCHS(1991-1993)withtheWeibull.WiththeGEE,thecoefcientis0.0268 (P 0.006)inGPRN,0.0249(P 0.008)inthemeta-regressionofcontrolledclinical trials,and0.01595intheGEEwithCCHSdata.Allthreeequationsreproducedthe publishedclinicaltrialswithodd-ratioswithin1/100ths.Conditionalodds-ratios werereplicatedinmeasurebetweenthetwodatasetsforfecalimpaction,postoperativeilleus,otherbowelobstruction,vomitingandabdominalpain.Thediabeticequationswerereplicatedexactlyin3countries,treatmentandHbA1cand complicationswithcoefcientswithin1/100thinpatientswithnewlydiagnosed T2DM. CONCLUSIONS: Thesearethreestudieswherethequanticationofthe relationshipbetweenasurrogateandandendpointhavebeedreplicatedwith precisionandsubsequentlyappliedtoclinicaltrials. PODIUMSESSIONII: DRUGUSEANDPATIENTSAFETY DU1 COMPARATIVESAFETYOFSTIMULANTANDATOMOXETINEASSOCIATED WITHTHERISKOFSUBSTANCEUSEDISORDERAMONGADOLESCENTSWITH ATTENTION-DEFICIT/HYPERACTIVITYDISORDERBhattacharjeeS1,ChenH2,BhataraV3,AparasuRR2 1WestVirginiaUniversity,Morgantown,WV,USA,2UniversityofHouston,Houston,TX,USA,3UniversityofSouthDakotaSanfordSchoolofMedicine,SiouxFalls,SD,USAOBJECTIVES: Thisstudycomparedtheriskofdevelopingsubstanceusedisorderin childrenwithAttention-Decit/HyperactivityDisorder(ADHD)utilizingstimulant andatomoxetine. METHODS: Thisstudyinvolvedretrospective,propensityscore matchedcohortassessingtheriskofdevelopingsubstanceusedisorderamong stimulantandatomoxetineuserswithADHDusingtheIMSLifeLinkHealthPlan ClaimsDatabase.Adolescentsbetween12and18yearsofagestartingonstimulant oratomoxetinetherapyfromJuly1,2004toDecember31,2005wereincludedinthe studysample.Substanceusedisorderwasclassiedinto(i)tobaccouse(ICD-9-CM code-305.1);(ii)drugabuse(ICD-9-CMcodes-305.2-305.9,304.x,292.x,304.00304.03,304.70-304.73);and(iii)alcoholabuse(ICD-9-CMcodes-265.2,303.x,305.0, 357.5,425.5,291.0-291.5,291.9,571.0,571.2,571.3,535.3,and790.3).Patientswith stimulantandatomoxetinewerematchedonpropensityscorescalculatedonthe basisofbaselinecharacteristics.Conditionallogisticmodelwasdevelopedusing theSTRATAoptiontoaccountforthematchedpairdesigntoassesstheriskof substanceusedisorderdevelopment.Persistencymeasuredasdurationoftherapy wasincludedasacovariateinthenalmodelalongwithothercovariateswhich weresignicantaftermatching. RESULTS: Thepropensityscorematchedcohort consistedof2,030adolescentswithADHDinboththestimulantandatomoxetine usergroups(totalof4,060adolescents).Conditionallogisticregressionanalysisdid notshowanystatisticallysignicantdifferencebetweenstimulantoratomoxetine useandtheriskofsubstanceusedevelopment(OddsRatio[OR]-0.86;95%CI0.29-2.50).Agewastheonlycovariatethatwassignicantlyassociatedwiththe substanceusedisorder(OR-3.55;95%CI-1.25-10.13). CONCLUSIONS: Utilizationof stimulantwasnotsignicantlyassociatedwithhigherriskofsubstanceusedisordercomparedtoatomoxetineinadolescents.Moreresearchisneededtoevaluate thelong-termeffectsofuseofmedicationsinADHD.A6VALUEINHEALTH14(2011)A1–A214

PAGE 15

DU2 PREVALENCEANDPREDICTORSOFPOTENTIALLYSIGNIFICANTDRUG-DRUG INTERACTIONSINTHEELDERLYDudashK1,NegriG2,BaccariniS3,RabinowitzC1,MaioV1 1ThomasJeffersonUniversity,Philadelphia,PA,USA,2LocalHealthAuthorityParma,Parma, Italy,3OspedalediFidenza-SanSecondo,Fidenza,Parma,ItalyOBJECTIVES: Toestimatetheprevalenceofpotentiallysignicantdrug-druginteractions(DDIs)amongambulatoryelderlyresidentsoftheLocalHealthAuthority (LHU)ofParma,Italy,andtoidentifypossiblepredictorsofDDIs. METHODS: We conductedaretrospectivecohortstudyusingthe2009ParmaLHUoutpatientpharmacydatabaseof86,571elderlysubjectsaged 65yearswhohadatleastone prescriptionlledduringthestudyyear(92.4%ofthetotalelderly).Potentially signicantDDIsweredeterminedusingapreviouslypublishedlist,whichwas updatedtoincludeanallistof17drugpairsaccordingtothedrugsreimbursedby the2009ItalianNationalFormulary.ApotentiallysignicantDDIwasdenedas thepresenceofaminimum5-dayoverlapindayssupplyfordrugsinaninteracting pair.Theprevalenceestimatewasmeasuredovera1-yearperiodandreportedas thetotalnumberofexposuresper10,000subjects.Logisticregressionwasusedto identifypatientcharacteristicsaspossiblepredictorsofDDIs. RESULTS: Anoverall 1,218elderlywereexposedtoatotalof1,279potentiallysignicantDDIs,yieldinga 1-yearperiodprevalenceof148exposuresper10,000subjects.Themostcommon DDIswerewarfarinandNSAIDS(47per10,000),simvastatin/lovastatinandamiodarone(45per10,000),andsimvastatin/lovastatinandmacrolides(30per10,000). Oddsofexposureweresignicantlyhigheramongthoseaged75-84yearsand thosewithmorechronicconditions,whileoddsofexposurewerelowerinfemales. Riskofexposureincreased1.23timesforeachadditionalprescribeddrug. CONCLUSIONS: OurresultsrevealedthatasubstantialnumberofDDIsoccurredin theelderlyoutpatientsofParmaLHU,especiallyinthechronicallyillsubjects. FindingsshouldbedisseminatedtogeneralpractitionerstoincreasetheawarenessofpotentiallysignicantDDIsandcharacteristicsassociatedwithanincreasedriskofexposuretoensuresafe,effective,andappropriateuseofmedicationfortheirelderlypatients. DU3 IMPACTOFTHEFDA’SANTIPSYCHOTICBLACKBOXWARNINGON PSYCHOTROPICDRUGPRESCRIBINGINELDERLYPATIENTSWITHDEMENTIA INOUTPATIENTANDOFFICE-BASEDSETTINGSDesaiVC,HeatonPC,KeltonC UniversityofCincinnati,Cincinnati,OH,USAOBJECTIVES: MostpatientswithdementiaalsosufferfromBehavioralandPsychologicalSymptomsofDementia(BPSD).BPSDisamajorreasonforacceleratedcognitivedeclineandpoorfunctioningindementiapatients,institutionalization,and increasedcaregiverdistress,andhenceposesasignicantquality-of-lifeandeconomicburdenonpatientsandtheirfamilies.ThereiscurrentlynoFDA-approved treatmentforBPSD.Ourobjectiveswere : Todetermine(1)whethertheFDA'sBlack BoxwarninginApril2005hasledtoadeclineinoff-labelprescriptionsofatypical antipsychoticsforBPSDaswellas(2)whetherprescriptionsforotherpsychotropic drugs,includingacetylcholinesteraseinhibitors(ACIs),antidepressants,anxiolytics,andantiepilepticshaveincreased. METHODS: Datawereobtainedandcombinedfromtwolargenationalsurveys:theNationalAmbulatoryMedicalCareSurveyandtheNationalHospitalAmbulatoryMedicalCareSurvey.Visitsofelderly( 65)patientswithdementiawereidentiedusingtheInternationalClassicationof Disease,9thRevision,ClinicalModication(ICD-9CM)codesfordementia.Ex cludedwerethevisitsforwhichthepatienthadadiagnosisofschizophreniaor bipolardisorder.Percentagedrugutilizationstatisticswerecalculated. RESULTS: Thepercentageofvisitsmentioninganatypicalantipsychoticdecreasedfrom 12.5%pre-warningto11.5%post-warning.ThepercentutilizationofACIs,antidepressants,andanxiolyticshasfallen,whiletheuseofantiepilepticshasremained thesamesincetheFDA'swarning.Post-warning,34.4%ofpatientsweretaking noneofthestudymedications,asopposedto26.1%pre-warning.In2005,theyear ofthewarning,44.6%ofvisitshadnostudy-drugmention. CONCLUSIONS: Our ndingssuggestthattheFDA'sApril2005BlackBoxwarninghadjustasmall negativeimpactontheuseofatypicalantipsychotics.Thereisnoevidenceof substitutionofotherpsychotropicmedicationsfortheatypicals.Whatareneeded areFDA-approveddrugsforBPSDsothatprescriberchoicescanexpandbeyond off-labeloptions. DU4 THELONGTERMUTILIZATIONOFSTIMULANTSINCHILDRENAND ADOLESCENTS:AMEDICAIDSTUDYSudharshanL,ChenH UniversityofHouston,Houston,TX,USAOBJECTIVES: Toassessthelongtermstimulantuseandotherpsychotropicmedicationsfora4-17agegroupthatreceivedastimulantprescriptionin1995.The secondaryobjectivewastoexaminethefactorsassociatedwithlongtermstimulantuse. METHODS: Aretrospectivelongitudinalanalysiswasconductedusingthe 1995-2001Medicaidclaimsdata.Patientsthatreceivedaprescriptionofastimulant in1995werefollowedthrough2001todeterminethedurationofstimulantuse, denedaspercentagethatreceivedaprescriptionofstimulanteveryyearofthe studyperiod.Longtermstimulantuseisdenedasthedurationinyearsofcontinuousstimulantuse,measuredasabeneciaryreceivingatleastoneprescriptionofDextroamphetamine,MethylphenidateorPemolineperyear.Amultivariate linearregressionmodelwasemployedtoidentifythefactorsassociatedwiththe long-termuseofstimulants. RESULTS: Outof5,022subjectswhoreceivedaprescriptionofatleastonestimulantin1995,1,985(39.52%)patientscontinuedtotake stimulantsattheendof7years.Almostonethirdofthecohorthadatleastone psychiatricdisorderduringthestudyperiod.ADHD(Attention-decit/hyperactivitydisorder)wasthemostcommonpsychiatricdisorderseenin1,950(98.23%)of therecipients.Beingwhite,youngageatbaseline,havingADHDandcomorbid substanceusedisorder,useofantianxiety,antidepressants,anticonvulsants,alphaagonistsandantipsychoticmedicationsaswellasphysicianspecialty(psychiatrist/psychologistvs.otherspecialties),countyofrecipient(ruralvs.metropolitan)andhospitalizationduetopsychoticdisordersweresignicantly(p 0.05) associatedwiththelongtermuseofstimulants. CONCLUSIONS: Morethanone thirdofpatientswhoreceivestimulantsaschildrencontinuereceivingstimulants attheendof7yearfollow-up.Patientswhocontinuouslyusedstimulantsduring the7yearfollow-uphadhigherpsychiatricco-morbiditiesandtookmorepsychotropicco-medications. PODIUMSESSIONII: INFECTIONOUTCOMESRESEARCH IN1 MEASUREMENTOFSYMPTOMSANDIMPACTOFINFLUENZA:DEVELOPMENT OFTHESYMPTOMINTENSITYANDIMPACTOFINFLUENZAQUESTIONNAIRE (FLU-IIQ)OsborneRH1,NorquistJ2,ElsworthG1,BusijaL3,MehtaV4,HerringT4,GuptaS4 1DeakinUniversity,Melbourne,Victoria,Australia,2MerckSharp&DohmeCorp.,NorthWales, PA,USA,3TheUniversityofMelbourne,Melbourne,Victoria,Australia,4MerckResearch Laboratories,NorthWales,PA,USAOBJECTIVES: Inuenzaclinicaltrialsusuallyfocusonoutcomessuchlaboratory conrmedinfectionusingviralcultureand/orriseinantibodytiterstoinuenza antigens.However,newagentsarenotonlyexpectedtoreduceinfectionbutinfectionintensity.Symptomseverityanddurationsymptomssuchasfever,cough, malaise,functionalimpact,andtimetoreturntodailyactivitieshaveincreased importance.TheaimwastodevelopandvalidateaPatientReportedOutcomes (PRO)measureoftheintensityandimpactofinuenzainfection. METHODS: ConsistentwithUSFoodandDrugAdministrationguidelines,constructanditemgenerationwereguidedbytheliteratureandconsultationwithpatientsandexpert physiciansthroughconceptmapping,focusgroups,andinterviews.Itemswere administeredto311peoplewithinuenzalikeillness(ILI)across25USAsiteswith randomassignmenttopaperorelectronicinteractivevoiceresponsesystem (IVRS).Analysesincludedclassicalpsychometrics,structuralequationmodeling (SEM)andRaschanalyses. RESULTS: Atotalof149conceptscoveringtheinuenza experiencewerederivedfromconceptmappingandtheseclusteredintoSymptomsandImpactonDailyactivities,EmotionsandOthers.Impactscaleswere unidimensionalwhereasthesymptomsdomainfactoredintoSystemicandRespiratorysymptoms.Alldomainsdisplayedgoodinternalconsistency(Cronbach 0.8)exceptthe3-itemRespiratorydomain( 0.48).Asix-factorSEMindicated excellenttwhereSystemic,RespiratoryandDailyactivitiesdomainsdifferentiatedpatientswithILIorconrmedinuenza.Allscaleswereresponsiveovertime. SEMindicatedpaperandIVRSdatacouldbepooled. CONCLUSIONS: Theuseof FDA-recommendedproceduresforpatientandclinicianconsultation,robustpsychometrics,andevaluationoftheitemsanddomainsinthetargetpopulation,hasgeneratedrigorousdatatosupportthevalidityoftheFluiiQ.TheFluiiQshouldimprove evaluationofexistingandfutureagentsdesignedtopreventorcontrolinuenzainfectionbyincreasingthebreadthanddepthofmeasurementinthiseld. IN2 ASSOCIATIONBETWEENDRUGTOLERABILITYANDECONOMICIMPACTFOR PROPHYLAXISOFINVASIVEFUNGALINFECTIONAFTERALLOGENEIC HEMATOPOIETICSTEMCELLTRANSPLANTGaoX1,JiX2,StephensJM1,SchlammH3,TaralloM3 1PharmeritInternational,Bethesda,MD,USA,2PharmeritNorthAmerica,LLC,Bethesda,MD, USA,3Pzer,Inc.,NewYork,NY,USAOBJECTIVES: Toassesstheassociationbetweendrugtolerabilityandeconomic impactforprophylaxisofinvasivefungalinfection(IFI)afterallogeneichematopoieticstemcelltransplant(HSCT). METHODS: Aprospectiveopen-labelmulticenterstudy(IMPROVIT)forprimaryIFIprophylaxisafterHSCTincludedpatients 12yearswhowererandomizedtoreceiveoralvoriconazole(VOR)ororalitraconazole(ITR)fromHSCTdayforatleast100andupto180days.Datafortherst 100dayswereanalyzed.Medicalresourceuse(MRU)endpointswerevaluedwith 2010UKcosts.Univariate,multivariate,andsubgroupanalyseswereconducted. RESULTS: 224patientswereintheVORand241intheITRgroup,withsimilar demographicdistributions(averageage43-year,59%male,92%Caucasian).AllcauseandstudydrugintolerancediscontinuationswerelessfrequentwithVOR thanITR(73%vs82%,p 0.02;7%vs22%,p 0.0001).VORpatientshadlongerstudy drugexposure(median:96vs68days,p 0.0001;mean:68vs60days,p 0.0162). MultipleregressionshowedthatITRpatientswere2timesmorelikely(P 0.0032)to useconcomitantantifungals(AF)versusVORpatients.Controllingfortreatment andkeybaselinevariables,prophylaxisdurationwasassociatedwithallkeyMRU: longerIFIprophylaxiswasassociatedwithfewerhospitaldays(p 0.0001),special unitdays(p 0.0001),andlessconcomitantAFuse(p 0.0001).Consequently,treatmentexposurewasnegativelyassociatedwithtotalcost(p 0.0001)andconcomitantAFcost(p 0.0035).Patientswhohadstudydiscontinuationorconcomitant AFuseincurred74%(£26,972vs£15,484,p 0.0001)and56%(£41,775vs.£26,782, p 0.0001)highertotalcost,comparedtotheircounterpartsrespectively.Eight moreprophylaxisdayswereassociatedwithapproximately1lesshospitaldayand 5%totalcostreduction.Subgroupanalysesshowedsimilarresults. CONCLUSIONS: TheabilityofpatientstotolerateandcontinuetheirAFprophylaxisafterHSCT appearstobeassociatedwithsignicanteconomicconsequences.A7VALUEINHEALTH14(2011)A1–A214

PAGE 16

IN3 COSTSASSOCIATEDWITHHCVANDRELATEDCOMPLICATIONSINTHE UNITEDSTATESFROMAMANAGEDCAREPAYER’SPERSPECTIVEMcAdamMarxC1,HaneCA2,BiskupiakJ1,DenizB3,McGarryL4,BrixnerD1 1UniversityofUtah,SaltLakeCity,UT,USA,2Ingenix,EdenPrairie,MN,USA,3Vertex PharmaceuticalsIncorporated,Cambridge,MA,USA,4i3Innovus,Medford,MA,USAOBJECTIVES: EstimatecurrenthealthcarecostsforHCVanditsconsequencesina large,USmanagedcareorganization(MCO). METHODS: PatientswithICD-9diagnosiscodesforHepatitisCviral(HCV)infection(1stdiagnosis indexdate),age18 yearswith6 monthsofcontinuousenrollmentwereidentiedinalarge,MCO claimsdatabasefrom1/1/2002to3/31/2010.HCVpatientswerematched1: 10to patientswithoutanHCVdiagnosisoradvancedliverdisease(ALD),basedongender,age,indexyear(wheresyntheticindexdate enrollment medianpost-enrollmentdaystocaseindexdate),hospitalreferralregion(HRR)state,pre-index healthcarecosts,alcoholism,HIV/AIDS,andmodiedCharlsonComorbidityIndex. Caseswerestratiedbydiseasestate:chronicHCVwithoutliverinvolvement(CHCV),compensatedcirrhosis(CC),decompensatedcirrhosis(DC),hepatocellular carcinoma(HCC),orlivertransplant.Meanper-memberper-year(PMPY)costs wereestimatedpost-index(totalpost-indexcostofallpatients/[sumofallpostindexdays/365]).IncrementalPMPYcostsforHCVpatientsvs.matchedcontrols wereestimatedoverallandbyeachdiseasestate. RESULTS: 34,597HCVpatients werematchedto330,435controls.Meanageofcaseswas49.9( 8.5)years;62% werefemale;78%hadC-HCV,4%CC,12%DC,3%HCC,and3%transplant.Incrementalcostsvs.controlsoverallwere$9,681PMPY.IncrementalPMPYcostsfor patientswithALDwereDC:$27,845,HCC:$43,671,andtransplant:$93,609.For patientswithoutALD,incrementalPMPYswereC-HCV:$5,870,andCC:$5,330. IncrementaldrugcostsforHCVtreatmentwere$2,739overall,rangingfrom$1,893$8,736fordifferentstates.C-HCVandCCdrugcostswere$2,659and$3,102, respectively. CONCLUSIONS: CurrentestimatesofHCVcostburdentoMCOswere higherthanpreviouslyreportedandincreasedsubstantiallywithprogressionto ALD.ThehigherestimatedcostsofmanagingchronicHCVwerelikelyduetohigh non-liverrelatedcostsamongHCVpatientsorimprecisecodingofCC. IN4 CARDIOVASCULARDISEASESCREENINGINHIV-INFECTEDPATIENTS–A COST-EFFECTIVENESSANALYSISNolteJEH1,NeumannT2,NeumannA3,ManneJ4,MostardtS3,AbbaraS5,BradyTJ5, HoffmannU5 ,GazelleGS1,WasemJ3,GoehlerA1 1InstituteforTechnologyAssessment,MassachusettsGeneralHospital,HarvardMedicalSchool, Boston,MA,USA,2UniversityHospitalEssen,Essen,Germany,3UniversityDuisburg-Essen, Essen,Germany,4HarvardSchoolofPublicHealth,Boston,MA,USA,5MassachusettsGeneral Hospital,HarvardMedicalSchool,Boston,MA,USAOBJECTIVES: HIV-infectedpatientsareatanincreasedriskofcardiovasculardiseases(CVD),resultingintheneedforintegrationofCVDscreeningintoHIVtreatmentguidelines.WeevaluateddifferentCVDscreeningstrategiesinHIV-infected patientswithregardtoeffectiveness,costs,andcost-effectiveness. METHODS: Cost-effectivenessanalysisusingamicrosimulationmodelreectingcoronaryarterydisease(CAD),myocardialdysfunction,andheartfailureinHIV-positivemen. Datasources:Patient-leveldatafromHIV-HEARTstudy,literature,Germanreimbursementdata.Timehorizon:Diagnosticphase,lifetime.Perspective:Societal. Interventions:Noscreening(SOC),ECG BNPwithechocardiography stress-ECGif pathologicECGorBNP("Outpatient"),ECG BNP echocardiography stress-ECG ("Cardiologist"),"Cardiologist"withcoronarycomputedtomographyangiography ifpathologicechocardiographyorstress-ECG("Cardiologist ").Outcomes:Diagnosticresults,discountedquality-adjustedlifeexpectancy(QALE)andlifetime costs,incrementalcost-effectivenessratio(ICER). RESULTS: TheinitialCADprevalenceinHIV-infectedmenaged40yearswasestimatedat5.9%.One-time"Outpatient","Cardiologist",and"Cardiologist "screeningscorrectlydiagnosed7,43, and48outof59CADpatientsper1,000patientsat 46/person, 109/person,and 429/person,respectively.TheexpectedQALEwasestimatedat16.54,16.56,16.71, and16.75yearsforSOC,andone-time"Outpatient","Cardiologist",and"Cardiologist "screeningsatmeanlifetimecostsperpatientof 321,348, 322,279, 327,670, and 328,864,respectively."Outpatient"and"Cardiologist"wereextendedlydominatedby"Cardiologist ".TheICERof"Cardiologist "vs.SOCwas 35,791per quality-adjustedlifeyear(QALY).Whenscreeningfrequencywasvariedbetween oneandveyearsatone-yearintervals,"Cardiologist"wasextendedlydominated by"Cardiologist ".ThepredictedICERsforthenon-dominatedstrategiesascomparedtoSOCwere 34,508, 40,489,and 49,373perQALYforannual"Outpatient" and"Cardiologist "at5-and4-yearintervals,respectively. CONCLUSIONS: Our preliminaryanalysessuggestthatintegratingroutineCVDscreeningintoHIV treatmentguidelinescouldbeclinicallybenecialandcost-effective. PODIUMSESSIONII: IMPACTOFMEDICATIONCOMPLIANCE MC1 COMPARISONOFREFILLGAPANALYSISMETHODOLOGIESINAPOPULATION OFPATIENTSWITHRHEUMATOIDARTHRITISRECEIVINGADALIMUMABOR ETANERCEPTCarterC,BolgeS,InghamM,Schmeichel-MuellerC CentocorOrthoBiotechServices,LLC,Horsham,PA,USAOBJECTIVES: Adalimumabandetanerceptarepatientself-administeredbiologic agentsusedinrheumatoidarthritis(RA).Prescriptionrellgapanalysesmaybe conductedtounderstandpatientrellbehaviorasaproxyforcompliance.The objectiveofthisstudywastoevaluategapsbetweenprescriptionrellsofadalimumaboretanerceptacrossmultipledatabasesandcompareresultsofdifferent methodologies. METHODS: Datawereobtainedfromthreeretrospectiveclaims databases(HealthCoreIntegratedResearchDatabase-HIRD[adalimumaboretanerceptinitiation07/2004-10/2008],MarketScanCommercialandMedicareSupplementalClaimsdatabases[01/2003-06/2008],andWoltersKluwerPharmaSolutions-WKPSdatabase[01/2004-12/2008]).Inclusioncriteria: 18yearsold, 2RA diagnosiscodes(ICD-9code714.xx),nobiologictherapy6or12monthspriorto currenttherapy,oneyearoftreatmentpersistence(HIRDandMarketScanonly), andnoselectedinammatoryconditions.Meangaps(days)inrellswerecalculatedamongallRApatientsinMarketScanandWKPS,andcalculatedamongonly thosepatientswith 1dayofgapinHIRD.Analysesincludedtherst15rell periods. RESULTS: Datafrom15,818adalimumaband25,175etanerceptpatients wereanalyzed.Meanadalimumabgapsspanned1.3to8.4and6.8to11.4daysin MarketScanandWKPS,respectively.Meanadalimumabgapsspanned17.2to33.4 daysinHIRDwhencalculatedamongonlythosepatientswith 1dayofgapin relling.Meanetanerceptgapsspanned1.3to10.9,6.6to14.2,and16.3to37.4days forMarketScan,WKPS,andHIRD,respectively. CONCLUSIONS: Resultsacrossmultipledatabasessubstantiatethatrellgapsexistamongadalimumabandetanerceptusersduringeachrellperiod.Thespecicmethodologyforcalculatingrell gapsshouldbeconsidered.Theextentofdelayedrellsmaybeunderestimated whencalculatedacrossanentirepopulation,insteadofonlyamongthosewith evidenceofagap.Furtherresearchisneededexaminingclinicalandeconomic consequencesassociatedwiththerapygaps. MC2 ASSESSINGPREDICTORSOFMEDICATIONADHERENCEINUNCONDITIONAL QUANTILEREGRESSIONFRAMEWORKBorahB MayoClinic,Rochester,MN,USAOBJECTIVES: Medicationadherencehasbeenlinkedtobetterhealthoutcomes. Therefore,acomprehensiveunderstandingofthepredictorsofadherenceisessentialforformulatingadherence-improvingstrategies.Existingmethodshavenot consideredevaluationofheterogeneousimpactsofpredictorsatdifferentparts (quantiles)oftheadherencedistributionasdenedbymedicationpossessionratio. Usingthenoveleconometricframeworkofunconditionalquantileregression (UQR),thisstudyassessestheheterogeneityofimpactsofadherencepredictorsfor anAlzheimer'sdisease(AD)population. METHODS: ThisretrospectiveclaimsanalysisidentiedADpatientsfromalargeUShealthplanthatinitiatedoralADtherapy(rivastigmine,donepezil,galantamine,ormemantine)between1/1/2006and 12/31/2007.Baselinecharacteristicswereassessedduringthe6-monthpre-index period;medicationadherencewasassessedduringthe1-yearpost-indexperiod. UQRwasestimatedat10th,20th,...,90thquantiles.Predictorsofadherenceiden tiedfromthedataincludedage,gender,indicatorofmentalhealthinsurance coverage,region,commercialvs.Medicareinsurance,logcost,comorbidity,and formularytierfortheADmedication. RESULTS: Baselinemedicationcountwas positivelyassociatedwithadherence(p 0.05)intheupperhalfoftheadherence distribution.Havingmentalhealthcoverageisnegativelyassociatedwithadherenceinallbutthe10thand20thquantilesbuttheimpactwassubstantiallyhigher inthersthalfoftheadherencedistribution.Baseline(log)costwaspositively associatedwithadherenceinthe40thandupperquantilesoftheadherencedistri bution.Forpatientsinthe80thand90thquantiles,thenumberofbaselineofce visitspredictedloweradherence.ComparedtopatientsfromtheEast,patients fromtheSouthwerelesslikelytobeadherentinthe60thand70thquantiles. CONCLUSIONS: Thestudyresultshighlighttheheterogeneityofimpactsofvarious adherencepredictor sapredictormaybestatisticallysignicantonlyinspecic quantilesoftheadherencedistribution,andtheimpactsmayvarysubstantially betweenquantiles. MC3 IMPACTOFMULTIPLEMEDICATIONCOMPLIANCEONCARDIOVASCULAR OUTCOMESINPATIENTSWITHTYPEIIDIABETESANDCOMORBID HYPERTENSIONCONTROLLINGFORENDOGENEITYBIASAnJJ,NicholMB UniversityofSouthernCalifornia,LosAngeles,CA,USAOBJECTIVES: Toinvestigatetheimpactofmultiplemedicationcomplianceonthe occurrenceofcardiovasculardisease(CVD)outcomesusinginstrumentalvariables (IV)tocontrolforendogeneitybias. METHODS: Weidentiedindividualswho newlystartoraldiabetesorhypertensionmedicationtherapybetweenJuly2006 andJune2007withthepre-existingcomorbidhypertensionordiabetesprescriptionhistoryduring6monthsofpre-indexperiodusingadministrativeclaimsfrom amanagedcareorganizationinsouthernCalifornia(N 1565).Multiplemedication compliancewasdenedasaproportionofdayscoveredforbothdiabetesand hypertensionmedicationsduringthreeyearsoffollow-up.Cardiovascularoutcomesincludedmyocardialinfarction,stroke,andperipheralvasculardisease. Instrumentalvariablesestimationusingphysicianrelatedvariablesincludinga dummyforthesameprescribersforbothmedications,thepercentageoffollow-up visitperphysician,andthepercentageofstatinprescriptionperphysicianwas implemented.ParameterestimateswerecomparedusingprobitandIV-probit models. RESULTS: Meancomplianceswere0.636( 0.008)fordiabetesmedications, 0.686( 0.008)forhypertensionmedications,and0.527( 0.008)forbothmedications.Afteradjustingforage,gender,baselineclinicalmeasures(HemoglobinA1C, bloodpressure,andlipid),pre-existingcondition(eitherdiabetesorhypertension), andElixhauser-comorbidity,adherencetobothmedicationswasnotsignicantly associatedwithdecreasedCVDrate( 0.070 0.118,p 0.554)basedonprobit model.Aftercontrollingforendogeneity,however,theimpactofmultiple medicationadherencebecamestatisticallysignicantusingIV-probitmodelA8VALUEINHEALTH14(2011)A1–A214

PAGE 17

( 2.270 0.345,p 0.0001).IVssatisedidenticationconditionsindividuallyand asagroup.Theincreasingmultiplemedicationcompliancefrom50%to80%was associatedwiththereductionoftheaverageprobabilityofpredictedCVDoccurrencefrom 0.46( 0.091,p 0.0001)to 1.14( 0.042,p 0.0001)atthemeanlevel ofcovariates. CONCLUSIONS: Aftercontrollingforendogeneitybias,compliance withmultiplemedicationswassignicantlyassociatedwithadecreaseintheCVD eventratewithtypeIIdiabetesandcomorbidhypertension. MC4 LONGTERMPERSISTENCEWITHACEI/ARBTHERAPYAFTERACUTE MYOCARDIALINFARCTION:ANANALYSISOFTHE2006-2007MEDICARE5% NATIONALSAMPLEDATALokhandwalaT,YangY,ThumulaV,BentleyJP,StrumM,BanahanBF,NullKD UniversityofMississippi,University,MS,USAOBJECTIVES: TostudythepatternsofdiscontinuationofACEI/ARBtherapyandto identifyfactorsassociatedwithdiscontinuationamongpostmyocardialinfarction (MI)patientsenrolledinMedicare. METHODS: Thisisaretrospectivecohortstudy utilizingtheMedicare5%nationalsampleclaimsdatafor2006-2007.Medicare beneciarieswithcontinuousenrollmentinPartA,B,andDin2006-2007,andwho werehospitalizedforanacuteMIbetweenJanuary1andJune30of2006,were identiedusingavalidatedalgorithm,requiringahospitalizationepisode 3and 180dayswithanICD-9-CMof410.x1asprimaryorsecondarydiagnosis.Post-MI patientswithanACEI/ARBprescriptionwithin90daysofdischargewerefollowed tostudypatternsofdiscontinuationuntilDecember31,2007.Timetodiscontinuationwasdenedasthedaysfrominitiationoftherapytoatherapygapof 90 days.SurvivalcurveswereconstructedusingtheKaplan-Meiertechnique,and potentialpredictorsoftherapydiscontinuation,includingdemographiccharacteristics,comorbiditiesandconcomitantmedicationswereestimatedusingCoxproportionalhazardsregression. RESULTS: Ofthe1,949subjectsinthecohort,66.1% werefemales,82.9%wereCaucasianwithameanageof78.6( 8.2)years.Approximately,20%ofthepatientsdiscontinuedtherapywithinsixmonthsand45% discontinuedwithinayear.Caucasianswerelesslikelytodiscontinuetherapyas comparedtoblacks(HR 0.774;[0.638-0.939]; p 0.0094).Amongthecomorbidconditions,dyslipidemia(HR 0.734;[0.612-0.880]; p 0.0008),cerebrovasculardisease (HR 1.124;[0.986-1.281]; p 0.0806)andCOPD(HR 1.154;[1.013-1.316]; p 0.0319) weresignicantpredictorsoftimetodiscontinuation.Patientsonconcomitant beta-blocker(HR 0.771;[0.627-0.949]; p 0.0141)andstatin(HR 0.857;[0.7360.999]; p 0.0491)therapywerelesslikelytodiscontinueACEI/ARBtherapy. CONCLUSIONS: ManypatientsinitiatingACEI/ARBtreatmentfollowingMIfailto consistentlyremainontherapyasisevidentbythehighratesofdiscontinuation withinayear.Severalfactorsincludingraceandcomorbiditiesarepotentialpredictorsofthisbehavior. PODIUMSESSIONIII: RESEARCHONMETHODS:ECONOMICEVALUATIONS EE1 SYSTEMATICREVIEWOFGUIDELINESFORHEALTHECONOMICEVALUATIONSMelnykP,WagnerM,DourdinN,RindressD BioMedComConsultantsInc.,Dorval,QC,CanadaOBJECTIVES: Numerousguidelinesonconductandreportingofhealtheconomic (HE)evaluationshavebeendevelopedoverseveraldecadesdifferingsomewhatin theirobjectives,scopeandspecicrecommendations.Wesystematicallyreviewed allaccessibleHEguidelinestoidentifycommonalitiesanddifferenceswithrespect totypeandcontentofrecommendations. METHODS: Asystematicsearchof PubMed/Medline,theCentreforReviewsandDissemination,theEQUATORnetwork,andwebsitesofhealthtechnologyassessmentagenciesandhealthcare coveragedecisionmakingbodiescomprehensivelyidentiedpubliclyaccessibleHE guidelines.Thesewerecategorizedasjurisdictionalmandatory,jurisdictionalnonmandatoryorgeneralguidelines.Datawasextractedintoatemplatedevisedto capture30eldsofcontent.Bothqualitativeandquantitativeanalyseswere performed. RESULTS: Seventy-fourHEguidelineswereobtained:23jurisdictional mandatory,11jurisdictionalnon-mandatoryand40general.Themostcommon topicsaddressedbyguidelineswere:sensitivityanalysis(by91%),dataidentication(89%),perspective(84%),choiceofhealthoutcomes(82%),assumptions(78%), valuationofresources(77%),typesofcosts(76%)/resourcestoinclude(74%),comparators(74%)andmodeling(74%).Comparedtojurisdictionalguidelines,general guidelinesfocusedonanarrowerrangeoftopics(e.g.,modeling)andwereless prescriptive.Althoughtherewasalevelofagreementonmanyissues,areasof divergencewereidentied:studyperspective(emphasisonsociety:32%ofguidelines;healthcarepayer:19%),unrelatedcostsinlife-yearsgained(inclusionin base-caseorsensitivityanalysisrecommendedby8%,exclusionby7%),methods tovalueproductivitylosses(frictioncostapproach:12%;humancapitalapproach: 15%;other:3%),andrecommendedtypeofsensitivityanalysis(probabilistic:28%; deterministicsufcient:28%). CONCLUSIONS: AvailableguidelinesforHEevaluationsarediverseandcoverabroadrangeofmethodologicalandreportingissues. Exceptforspecicjurisdictionalneedsandafewcontroversialissues,however,there appearstobeconvergenceofHEguidanceovertimeandstabilityofbasiccontent. EE2 VALIDATIONOFTHEUPDATEDCHARLSONCOMORBIDITYINDEX(CCI)TO PREDICTHEALTHCAREUTILIZATIONFORDIABETICPATIENTSUSING ADMINISTRATIVEDATAChengLI1,RascatiKL2,TriceS3,LawsonK1,BarnerJC2 1TheUniversityofTexasatAustin,Austin,TX,USA,2UniversityofTexasatAustin,Austin,TX, USA,3DepartmentofDefense,FortSamHouston,TX,USAOBJECTIVES: TovalidatetherecentlyupdatedCharlsonComorbidityIndex(CCI) forthepredictionoffuturehealthcareutilizationfordiabeticpatients. METHODS: AdministrativeclaimsdatawereobtainedfordiabeticpatientsenrolledcontinuouslyforthreeyearsintheDepartmentofDefenseTRICAREprogramforretrospectiveanalysis.TheupdatedandtheoriginalCCIscoreswerecalculatedusingbaselineyeardata.Linearregressionmodelswereusedtoestimatelog-transformed healthcareexpenditures(COST)forone-andtwo-yearpost-indexperiods.Zeroinatednegativebinomialregressionmodelswereusedtoestimatethenumberof hospitalizations(HOS)andthenumberofemergencydepartmentvisits(ED)for one-andtwo-yearpost-indexperiods.Theoutcomevariableswerethendichotomized(aboveorbelowthe90thpercentileofCOST; 1HOSornone; 1EDornone) andestimatedusinglogisticregressionmodels.Adjusted R¨,Akaikeinformation criteria(AIC),and c statisticswereassessedtocomparethetwoCCIversions. RESULTS: Atotalof8,704patientswereincludedinthestudy.Thestudypopulation hadameanageof51.0years(SD:10.5),and46.3percentweremale.Inthelinear regressionmodels,theupdatedCCIexplainedmorevariancethantheoriginalCCI inone-yearCOST(adjusted R¨ 13.0%vs.11.3%)andtwo-yearCOST(adjusted R¨ 15.7%vs.13.9%),adjustingforageandsex.TheupdatedCCIwasabetterpredictor ofone-andtwo-yearHOS(AIC 8581,13821)aswellasone-andtwo-yearED(AIC 17009,25200).Inthelogisticregressionmodels,theupdatedCCIperformedbetter inpredictingallstudyoutcomes( c 0.619to0.754)thantheoriginalCCI( c 0.611 to0.737). CONCLUSIONS: Inapopulationofdiabeticpatients,theupdatedCCI showedimprovedpredictiveperformancecomparedtotheoriginalCCI.TheupdatedCCIshouldbevalidatedinotherpatientpopulations. EE3 PERFORMANCEOFDIFFERENTCOMORBIDITYMEASURESINPREDICTING HEALTHCAREEXPENDITUREINPATIENTSWITHDEMENTIAJohnsonML,MehtaS,ChitnisAS,BhowmikD,DwibediN,KambleP UniversityofHouston,Houston,TX,USAOBJECTIVES: Toevaluatetheperformanceofdiagnosisbasedandprescription basedcomorbiditymeasuresinpredictingtotalhealthcareexpenditureinpatients withdementia. METHODS: Thiscross-sectionalstudywasconductedusingthe householdandmedicalprovidercomponentlesofMedicalExpenditurePanel Survey(MEPS)datafrom2000to2003(panel5,6,and7).Dementiapatientswere identiedusingcccodexandInternationalClassicationofDiseases,9thRevision, ClinicalModication(ICD-9CM)codes.Totalhealthcareexpenditurewasdened asthesumofinpatient,outpatientandpharmacycosts.Thecomorbiditymeasures evaluatedwerediagnosisbasedscales(identiedusingICD-9CMcodes):D'Hoore's adaptationofCharlsoncomorbidityindex(CCI),andElixhausercomorbidityalgorithm;andprescriptionbasedscales(identiedusingnationaldrugcodes):Chronic DiseaseScore(CDS-1andCDS-2).Theperformanceofcomorbiditymeasureswas comparedusingtheadjustedR¨derivedfrommultiplelinearregressionmodels thatincludedbaselinedemographic(age,race,sex)socio-economic(income,insurance,geographicregion),andperceivedhealthstatusvariables.Theperformanceofcombineddiagnosesandprescription-basedscoreswasalsocompared. RESULTS: TheadjustedR2ofthebaselinemodelwas0.1464.Modelperformance aftertheadditionofcomorbiditymeasureswas:Elixhauser(0.3326),CDS-2(0.3427), Charlson/D'Hoore(0.2184),andCDS-1(0.2388).Theperformanceofcombineddiagnosisandprescriptionbasedmeasureswas:CDS-2/Elixhauser(0.4356),CDS 1/Elixhauser(0.3979),CDS-2/D'Hoore(0.3493)andCDS-1/D'Hoore(0.2573). CONCLUSIONS: TheElixhauserandCDS-2comorbiditymeasurescombinedhad thehighestpredictionofhealthcareexpenditureinpatientswithdementia.The improvementinriskadjustmentfromthecombinationofdiagnosisandprescriptionbasedmeasuresindicatesthatthesedifferenttypesofmeasuresmaybecapturingdifferentaspectsoftheconstructofcomorbidity,constructsotherthan comorbidity,orboth. EE4 BOOTSTRAPPINGUSEDTOPROVIDEROBUSTMEANANDVARIANCE ESTIMATESFORCOMPARINGPATIENTSTREATEDWITHLIRAGLUTIDETOA LARGECOMPARISONCOHORTMcAdamMarxC1,YeX1,BouchardJ2,AagrenM2,ConnerC3,BrixnerD1 1UniversityofUtah,SaltLakeCity,UT,USA,2NovoNordisk,Inc.,Princeton,NJ,USA,3Novo Nordisk,Inc.,Redmond,WA,USAOBJECTIVES: Inananalysesofpatientswithtype2diabetes(T2DM)inalarge electronicmedicalrecord(EMR)database,smalldifferenceswerefoundtobestatisticallysignicantbetweenN 1162patientswithliraglutideversusacomparison cohortduetothecomparisongroupsamplesize(n 274,922).Thepurposeofthis studyistoevaluateabootstrappingtechniquetoproviderobustmeanandvarianceestimatesforcomparisonpatients,therebyhelpingtoaddresstheissueof beingover-powered. METHODS: Studypatientswereage 18yearswithT2DM, prescribedliraglutideorotherantidiabeticdrugJanuary1,2010toJuly16,2010and with 13monthsofEMRactivity.Bootstrappingwasusedtoprovidecohortmean, standarddeviation,and95%CIestimatesforthecomparisoncohortandwere calculatedasthemeanofthemeanvaluesidentiedin1000randomdrawswith replacementof1162comparisonpatients.Means(95%CI)werecomparedforcontinuousvariables(age,HbA1candbloodpressure[BP])forliraglutideversusthe overallcomparisongroupandtobootstrapestimates. RESULTS: Of1162liraglutide patients,mean(95%CI)agewas55.5(54.9,56.2)yearsversus60.9(60.8,60.9)years forallcomparisonpatientsand60.9(60.1,61.6)yearsforcomparisonpatientsper bootstrapestimates(bothp 0.05).HbA1cwas8.12%(8.00,8.24)forliraglutideversus7.62%(7.61,7.63)and7.63%(7.49,7.76)forallcomparisonpatientsandper bootstrap,respectively(bothp 0.05).BPwas127.0(126.1,127.8)/75.8(75.3,76.4)A9VALUEINHEALTH14(2011)A1–A214

PAGE 18

mmHgforliraglutideversus130.1(130.0,130.1)/75.4(75.4,75.5)mmHgand130.1 (129.0,131.2)/75.5(74.8,76.1)mmHgforallcomparisonpatientsandperbootstrap, respectively(p 0.05exceptbootstrapdiastolicBPp 0.05). CONCLUSIONS: Abootstrapanalysisprovidedmorerobustvarianceand95%CIestimatesforalarge comparisongroup.Thistechniquecanhelpresearchersavoididentifyingstatisticalsignicancewhendifferencesarenotclinicallymeaningfulwhenevaluatinga largepatientcohort. PODIUMSESSIONIII: EVOLVINGCONCEPTSINOUTCOMESRESEARCH EV1 DEVELOPMENTOFAGUIDANCEFORINCLUDINGPATIENT-REPORTED OUTCOMES(PROS)INPOST-APPROVALCLINICALTRIALSOFONCOLOGY DRUGSFORCOMPARATIVEEFFECTIVENESSRESEARCH(CER)BaschEM1,AbernethyA2,MullinsCD3,TiglaoMR4,TunisSR4 1MemorialSloan-KetteringCancerCenter,NewYork,NY,USA,2DukeUniversityMedicalCenter, Durham,NC,USA,3UniversityofMarylandSchoolofPharmacy,Baltimore,MD,USA,4Center forMedicalTechnologyandPolicy,Baltimore,MD,USAOBJECTIVES: AlthoughanFDAGuidanceisavailabletodirectthedevelopmentand inclusionofPROsinpreapprovalclinicalresearchtowardsdrugapprovalandlabeling,nosuchguidanceexistsinthepost-marketCERcontext.BecausetheclinicalsettingandstakeholdersinCERdifferfromthepreapprovalspace,different methodologicalandpracticalconsiderationsarenecessary.Theaimofthisproject istodevelopspecicrecommendationsforcapturingthepatientperspectivein post-marketoncologytrialsandotherstudiesthatfallintothegeneralcategory termedCER. METHODS: Asemi-structuredquestionnairepertainingtotheuseof PROsinCERwasdevelopedbasedonareviewofscienticliteratureandconsultationwithPROstudymethodologists.In-depthinterviewswereconductedusing thesequestionswith15individualsfromtheclinicalresearch,clinicalpractice, regulatory,payer,andpatientcommunities.Basedoninterviewresponses,alistof potentialguidancetopicswasdeveloped.TheCenterforMedicalTechnologyPolicy convenedamultidisciplinaryworkinggroupofleadingmedicalresearchersin PROs,ePROconsultants,andpatientadvocatestodiscussthesetopicsanddevelop anonoutlineofbestpracticerecommendationsforintegratingPROsinCER. RESULTS: Aseriesofrecommendationsfocusedonestablishingstandardsforthe useofPROsinCERweredeveloped.Theserecommendationsaddressselectionof appropriatemeasures;inclusionofa"coreminimumitemset"acrossalloncology CERtrials;andinclusionofitemswhichassesssymptoms,globalQOL,perceived valueofcare,andtreatmentcompliance.TherecommendationsalsoprovideguidanceforhowtoimplementPROsinreal-worldstudiesandminimizemissingdata. CONCLUSIONS: ThepatientperspectiveisanessentialcomponentofCER.PRO methodsusedinCERdifferfromthepreapprovalcontext.Forimprovedconsistencyacrosstrials,aguidanceisbeingdevelopedwhichidentiesastandardcore itemsettobeconsideredforuseacrossalloncologyCERstudies. EV2 PAYERSANDPROS:BEYONDQOLMillerKL,StevensCA PAREXELConsulting,Waltham,MA,USAAtthe2010ISPORconferencemedicaldirectorsfromUShealthandpharmacy plansprovidedtheirperspectiveonthevalueofhealthoutcomesresearchindecidingdrugcoverage.ThemessagetotheISPORmemberswasthatqualityoflife holdslittleswayamongthesedecision-makers.Neverthelesshealthcareresearchers,policymakers,legislators,andotherstakeholderscontinuetoemphasizethe inclusionofthepatientperspectiveincomparativeeffectivenessresearch(CER) andinhealthcaredecisionmaking.Recently,inarguingtheimportanceofpatient reportedoutcomes(PROs)inCER,Wu,etal.emphasizethatinterventionsmust "improv[e]outcomes,reduc[e]costs,orboth".Indeed,astheseauthorsstate"many relevantoutcomesrequirepatient-reportedinformation".SoifQOLmeasures aren'ttheanswer,whichPROswillprovideboththepatientperspectiveandinformationcriticaltodecisionmakersintheprivateandpublicsectors?Collaboration betweenpayersandresearcherswillbeessentialtoanswerthisquestion.However,theonuswillfallsquarelyuponthehealthoutcomesresearchers,toidentify andquantifytherelationshipbetweenoutcomesandvalue.Furthermore,these researcherswillhavetodemonstratetheimportancetotheindividualpayers,as theyhavelittleincentivetofurthercomplicatedrugcoveragedecisionmaking. Therefore,evenifimprovedoutcomes-asmeasuredbyPROs-provetoprovide meaningfulbenettopatients,payerswillrequirethattheinformationbeapplicablespecicallytotheirpopulations.ForinstanceWu,etal.suggestfunctional status90daysafterdischargeasanexampleofarelevantpatientoutcome.We,as healthoutcomesresearchersunderstandtherelevancetopatients;butforpayers wemustanswerhowisimprovedfunctionalstatusavaluablemeasuretoindividualpayers?Withouttheanswertothistypeofquestion,payerswillbereluctantto includetheseoutcomesendpointsintheirdecisionmaking. EV3 ACOMPARATIVEEFFECTIVENESSINDEXTOINFORMCLINICALDECISIONSHorowicz-MehlerN1,DoyleJ2,ArikianS3,HaganM4 1QuintilesGlobalConsulting,Hawthorne,NY,USA,2Quintiles,Hawthorne,NY,USA,3Genesis BioPharma,Hoboken,NJ,USA,4TBD,NewYork,NY,USAOBJECTIVES: Presentaframeworktosystematicallymeasureandcombinekey attributesofhealthinterventionsintoasingleComparativeEffectivenessIndex (CEI). METHODS: Relevantattributesincludingefcacy/effectiveness,compliance/ persistency,safety,health-relatedqualityoflife(HRQoL)/utilityarelinkedthrough atransparentformula.Efcacyandeffectivenessaregatheredfromameta-analysisofrandomizedcontrolledtrialsandobservationalstudies,accordingly.Complianceandpersistencyarebederivedfrompharmacyclaimsdatausingmetrics suchasthemedicationpossessionratio.Weproposea5-pointratingscaleforthe safetymeasurefromlowrisk"suitableforwidespreaduse"tohighrisk"careful considerationofriskversusbenet".Wealsoconsideracombinedmeasureof risk-benetassessmentborrowingfromexistingquantitativemethodssuchasa NumberNeededtoTreatandNumberNeededtoHarm.Genericordisease-specic patient-reportedoutcomemeasuresprovideinformationontheimpactoftheinterventiononthepatient'sHRQoLandindicatethevalueofaspecictreatmentfor agivenhealthstate.Toensurerelevantcomparisons,weaccountfortherapeutic areawhenassessingtheCEIofmultipleinterventions.Finally,ameasureofquality ofevidence(Jadadscore)ensurethatonlyevidencefromrobustdesignsfedtheCEI. CONCLUSION: TheCEIaimstoprovideacomprehensiveandbalancedpictureof disparateyetlinkedperformanceattributes.Currently,acompletepictureof healthinterventionsisnotcentralizedandnoteasilyaccessiblethuslimitinginformedclinicaldecision-makingbyvariousstakeholders(physicians,patients, payersandpolicymakers).Thisindexisintendedtoguidehealthcareparticipants inmakingevidence-basedclinicaldecisionsatthepopulationandpatientlevels suchthataprovidermightwanttotrade-offeffectivenessforincreasesafetywhen treatingapatientwithseverecomorbidities.Forfulltransparencyanduse,we proposethecreationofaUSpublicallyavailableweb-baseddatabaseoftheinformationunderlyingtheindex. EV4 MOREBANGFORYOURBUCK:TAKEARISKWHENANALYSINGINTERVIEW DATARobertsG DoubleHelixConsulting,London,UKISSUE: Dataobtainedbyintervieworquestionnaireareroutinelyreportedwith summarystatistics,frequency,meanandrange.Canweobtainmoreinformative resultsbyinterpretingthedatausingriskanalysistechniques? OVERVIEW: Probabilisticmethodscanbeusedwhendataissparseandaddressareasofuncertainty. Quantitativeanalysisofintervieworquestionnairedatatypicallyinvolvespresentingfrequenciesofresponseandsummariesofaggregatedatawithdescriptive statistics.Theseareusedoftentoaddressresearchquestionsincludingassessing marketaccessopportunities,pricingandreimbursementscenarios,llinggapsin healtheconomicdatasuchasresourceuse.Butdowealwaysgetthemostfromthe datawehave?Howcanwemakebetterinformeddecisions?Letusconsidera hypotheticalquestionwhererespondentsareaskedtorateaseriesofattributes(A, B,C)usingascaleof1to10.Weendupwithadistributionofanswersthatare summarisedasaveragesforeachattribute.Oftenwearethenfacedwithinterpretingaseriesofaveragescoresthatdonotdiffermarkedlybetweenattributes.Uncertaintyintheratingsprovidedbyrespondentscanbeusedtoimproveourinterpretation.Onemethodwouldbetousebootstraptechniques,tosamplewith replacementtherawdata,andrunningasimulationtoobtainthebootstrapuncertaintydistributionforthemean.Fromthiswecandeterminetheprobability thatattributeAisbetterorworsethanBorC.Withthisinformationatourdisposal wearebetterpositionedtomakeaninformeddecision. CONCLUSION: Decision analyticmethodsaddvalue,improvecommunicationaboutrisk,supportdecision making,andidentifyresearchopportunitiesforreducinguncertaintywheninterpretinginterviewandquestionnairedata. PODIUMSESSIONIII: ANALYSISOFMEDICAREPOLICYANDRESOURCEUSE MD1 USINGCOST-EFFECTIVENESSINFORMATIONTOALLOCATEMEDICARE RESOURCES–HOWMUCHMOREHEALTHFORTHEMONEY?ChambersJD1,CohenJT1,NeumannPJ1,LordJ2,BuxtonM2 1TuftsMedicalCenter,Boston,MA,USA,2HERG,BrunelUniversity,Uxbridge,UKOBJECTIVES: TheCentersforMedicareandMedicaidServices(CMS)donotexplicitlyusecost-effectivenessinformationinnationalcoveragedeterminations (NCDs).Theobjectiveofthisstudywastoestimatepotentialgainsinhealthattainablefromreallocatingresourcesusingcost-effectivenessinformationwithinthe existinglevelofspending,andtodeterminetheconsequencesforthedistribution ofresources. METHODS: NCDsfrom1999through2007wereincluded.Estimatesof thecost-effectivenessofincludedcoveragedecisions,bothpositiveandnon-coverage,wereidentiedfromaliteraturereview.Forcoveragedecisionswithan associatedestimateofcost-effectiveness(n 66),anestimateofutilizationandsize oftheunserved'eligiblepatientpopulationwasestablishedfromaMedicare5% claimsdatabase(2007)usingacombinationofdiagnostic(ICD-9)andreimbursement(CPT)codes.Annualandnetcostsweretakenfromthecost-effectiveness studyorestimatedusingreimbursementcodes. RESULTS: Completeinformation wasavailablefor36(55%)coveragedecisionswithanassociatedestimateofcosteffectiveness.Areallocationofresourcesyieldedanincreaseinaggregatequality adjustedlifeyears(QALYs)of1.8million,approximately0.15QALYsperMedicare beneciaryaffectedbythereallocation.Theanalysisprovidesaquanticationof howthereallocationaffecteddifferentpatientpopulationsandtheutilizationof differenttypesoftechnology.Insummary,followingreallocation,agreaterproportionofresourcesweredirectedtooncology,diagnosticimagingandtests,and diseasesaffecting 1millionMedicarebeneciaries,thanpriortoreallocation. Conversely,followingreallocation,adecreasedproportionofresourcesweredirectedtocardiology,interventionsotherthandiagnostictests,anddiseasesaffecting50k-1millionMedicarebeneciaries,thanpriortoreallocation. CONCLUSIONS: Usingcost-effectivenessinformationtoinformCMSNCDshasthepotentialtosub-A10VALUEINHEALTH14(2011)A1–A214

PAGE 19

stantiallyincreasetheaggregatehealthofMedicarebeneciarieswhilenotincreasing Medicarespending.Thereallocationresultedinagreaterproportionofresourcesdirectedtowardsdiagnosticstests,moreprevalentdiseases,andoncology. MD2 PRESCRIPTIONDRUGCOSTANDUSEINTHEMEDICAREPARTD POPULATION-USESOFANEWLIMITEDDATASETPowersCA1,VargheseA1,HsuVD2,O'DonnellJ3,SchneiderK3 1CentersforMedicare&MedicaidServices(CMS),Baltimore,MD,USA,2Buccaneer,AVangent Company,OwingsMills,MD,USA,3Buccaneer,AVangentCompany,WestDesMoines,IA,USAOBJECTIVES: WhileMedicarePartDresearchidentiableleshavebeenavailable toacademicresearchersforsometime,anewlimiteddataset(LDS)containingPart Ddrugutilizationandcostinformationisnowavailabletobothacademicand non-academicresearchers.ThisstudywasundertakentodemonstratetheversatilityofLDSinanalyzingdrugutilizationandcost. METHODS: AgedMedicarebeneciariesaliveforallof2008with12monthsofPartA,B,andDcoveragewere selected.Theaveragenumberofprescriptiondruglls(30-dayadjusted)andaveragecostspermemberpermonth(PMPM)werecalculatedoverall,bydemographics, andforselectedchronicconditions.Specicdrugusewasalsoexaminedfora chronicconditionofinterest. RESULTS: Overall,theaveragePMPMnumberoflls was4.3andthecostwas$212,beneciariestook8.9distinctmedications,and meancostperllwas$49.75.Olderbeneciarieslledmoreprescriptionsper month(4.9for 85yearsvs.3.9for65-74years),buthadlowermeancostsperll ($46.95for 85yearsvs.$51.61for65-74years).FemaleshadhigherPMPMlls(4.5 vs.3.9)andcosts($220vs.$198)comparedtomales.Dual-eligibleMedicareand Medicaidbeneciarieshadapproximately1.5timeshigherPMPMllsandcosts comparedtobeneciarieswithoutanyPartDsubsidies.PMPMllsandcostsalso varywithrace.ComparedtotheoverallMedicarecohort,beneciarieswiththe selectedchronicconditionshadhigherPMPMllsandcosts:diabetes(6.0,$303), Alzheimer's(6.0,$358),depression(6.4,$357),andosteoporosis(4.8,$261).Patterns ofuseandcostbydemographicsdifferedbyconditioncomparedtotheoverall cohort,exceptbydualeligiblestatus. CONCLUSIONS: Asdemonstrated,thenew PartDLDSdataallowsresearcherstoconductutilizationandcoststudiesusingall andsubsetsofMedicarebeneciarieswithselectedchronicconditions. MD3 EVALUATINGTHEWILLINGNESS-TO-PAYOFMEDICAREBENEFICIARIESFOR PARTDPLANASSISTANCEPatelRA,WalbergMP,NaJ,HsiouD,PanchalV,WoelfelJA,GalalSM,Carr-LopezSM, ChanEK UniversityofthePacic,Stockton,CA,USAOBJECTIVES: MedicarePartDallowseachbeneciarytheabilitytochooseandenroll inaprivatelysponsoredMedicare-approvedprescriptiondrugplan(PDP).However, with33differentstand-alonePDPstochoosefrominCaliforniain2011alone,sucha choicecanbeoverwhelming.WesoughttoassistbeneciarieswithPartDplanevaluationandquantifytheirwillingness-to-pay(WTP)forsuchservicesduringthe2011 openenrollmentperiod. METHODS: Nineoutreacheventswereheldincitiesacross central/northernCaliforniaduringwhich395beneciarieswereassistedwiththeir MedicarePartDplan.Duringeachsession,beneciary-specicinformation(e.g., prescriptionmedications)wasenteredintotheMedicarePlanFinderTool( www. medicare.gov )tohelpfacilitatetheintervention.Demographicandplan-specicdata, alongwiththeresultsoftheintervention,werecollectedfromeachassistedbeneciary.Attheconclusionofthesession,eachbeneciary'sWTPwaselicited. RESULTS: Ofthe348(88.1%)beneciarieswhoansweredthequestion,themedian(mean)WTP forPartDplanhelpwas$20($40.04).Asignicantdifference(p 0.001)wasfoundin theWTPofbeneciariesasafunctionofwhetherornottheyreceivedadditional governmentalassistance(e.g.,Medicaid).Themedian(mean)WTPof96subsidy-recipientswas$0($12.43)versus$25($51.38)for247non-subsidyrecipients.WTPwas alsodependent(p 0.01)onwhetherornotthebeneciarywasenrolledintoanew planduringtheinterventionalsession.Themedian(mean)WTPofthe126beneciariesthatwereenrolledintoaPDPplanonsitewas$25($57.79)versus$10($30.00)for the220beneciarieswhowerenot.Finally,beneciaries'WTPwassignicantlycorrelated(rs 0.235;p 0.001)withtheestimatedannualcostsavingsidentiedduring theintervention. CONCLUSIONS: BeneciariesvalueMedicarePartDplanassistance andtheperceivedvaluevariesasafunctionofcertaindemographicandinterventionalcharacteristics. MD4 EVALUATIONOFANINTERVENTIONTOREDUCEPOLY-PHARMACYIN MEDICAREPARTDLivengoodKB,HarrellT,AbarcaJ WellPoint,Indianapolis,IN,USAOBJECTIVES: Theobjectiveofthisstudywastoevaluatetherelationshipbetween aphysician-basedinterventionandthereductioninpoly-pharmacyforasegment ofanationalmanagedcareMedicarePartDcohort. METHODS: ProvidersofMedicarePartDmembersin2009weresentaletterifoneormoreoftheirpatientsmet thecriteriaforbothmedicationtherapymanagement(MTM)andpoly-pharmacy. Theproviderswereaskedtoreviewthemember'sprescriptionproleandmake appropriatechangesbasedsolelyontheprovider'sclinicaljudgement.Twofollow-upquartersweredenedforeachmemberin2009,andtwobaselinequarters weredenedin2008.Fromprescriptionclaims,eldswerecollectedallowingthe creationofchangevariables.Changesinclaimcount,drug(GPI)count,andmedicationcostwereevaluatedusingfeasiblegeneralizedleastsquaresandalinear probabilitymodel(changeinclaimanddrugcount).Theinterventioncohortwas comparedtoacontrolcohort,consistingofMedicarePartDmemberswhomet poly-pharmacycriteriabutnotMTMprogramcriteria. RESULTS: Theanalysisrevealedasignicantlygreaterreductionof3.72claims(secondbaselinequarter2008 versussecondfollow-upquarter2009.p 0.001)fortheintervention.Therewasa signicantlygreaterreductionof1.67drugs(GPI8)fortheintervention(p 0.001). Medicationcostsfortheinterventionweredecreased$204(ingredientcost,$149in amountpaidformedications,p 0.001forbothmedicationcosts).Thelinear probabilitymodelshowedtheinterventionwasassociatedwitha4.4%,6.5%,and 12.9%signicantlygreaterprobabilityofathreetove,sixtoeight,andagreater thaneight,respectively,claimcountreduction(p 0.001).Theinterventionwas alsoassociatedwithasignicantlygreaterprobabilityofdrugcountreduction(p 0.001). CONCLUSIONS: Theinterventionwasassociatedwithastatisticallysignificantmarginalreductioninclaimcount,drugcount,andmedicationcostscomparedtoacontrolMedicarePartDcohort. PODIUMSESSIONIII: EXAMININGTHEQALY QA1 COST-EFFECTIVENESSOFDUTASTERIDEASACHEMOPREVENTIONIN PROSTATECANCER:REDUCEWITHIN-TRIALANALYSISEarnshawSR1,ChirilaC1,McDadeC1,BlackL2,AndrioleGL3 1RTIHealthSolutions,ResearchTrianglePark,NC,USA,2GlaxoSmithKline,ResearchTriangle Park,NC,USA,3WashingtonUniversitySchoolofMedicineinSt.Louis,St.Louis,MO,USAOBJECTIVES: Awithin-trialanalysisexaminedthecost-effectivenessofdutasterideasachemopreventioncomparedwithplacebo,byanalyzingtheresourceuse incurredbymeninReductionbyDutasterideofProstateCancerEvents(REDUCE) trial. METHODS: REDUCEwasa4-year,randomized,double-blind,placebo-controlled,parallel-grouptrialtocomparetheefcacyandsafetyofdutasteride0.5mg dailywithplaceboforchemopreventioninmenatincreasedriskforprostatecancer.Resourceusedatawasprospectivelycollectedandincludeddrugs,procedures, outpatientvisits,andinpatientvisitsingeneralandthoseassociatedwithrelated conditionssuchashematuria,hematospermia,acuteurinaryretention,urinary tractinfection,benignprostatichyperplasia,andprostatecancer.Unitcostsfrom standardcostingsourceswereappliedtoresourceuseincurredduringthetrial. Quality-adjustedlife-years(QALYs)werecalculatedonthebasisofmenbeing healthyandhavingBPH,prostatecancer,andadverseevents.Utilitieswereobtainedfromthepublishedliterature.IncrementalcostsperQALYwerecalculated. Bootstrapanalyses,toderivecost-effectivenessacceptabilitycurves,wereperformed.TheanalysiswasperformedfromtheperspectiveofaUShealthcare payer. RESULTS: Menusingdutasterideincurredameancostof$6,610(including costofdutasteride)whereasmenintheplacebogroupincurredameancostof $3,177overthe4-yearperiodofthetrial.Menusingdutasterideaccrued0.13more QALYsthanmenonplacebo(3.26vs.3.13).Asaresult,theuseofdutasterideasa chemopreventionwascost-effective(incrementalcostperQALY $50,000),with anincrementalcostperQALYof$26,516.Theincrementalcostperprostatecancer avoidedwas$19. CONCLUSIONS: Usingdutasterideasachemopreventionfor prostatecancerinmenatincreasedrisk,asseenintheREDUCEtrial,iscosteffective.Thesecost-effectivebenetsofchemopreventionarerealizableinaperiodoftimeasshortas4years. QA2 CROSS-WALKINGCANCER-SPECIFICINSTRUMENTSTOTHEEQ-5DANDSF-6DTeckleP1,PeacockS1,vanderHoekK1,ChiaS2,MeloskyB2,GelmonK2 1CanadianCentreforAppliedResearchinCancerControl(ARCC),Vancouver,BC,Canada,2MedicalOncology,BritishColumbiaCancerAgency,Vancouver,BC,CanadaOBJECTIVES: Tohelpfacilitateeconomicevaluationsofinterventionsfortreating cancer,weestimatedutilityindicesforthetwofrequentlyusedcancer-specic (FACT-GandEORTC-QLQ-C30)instrumentsofquality-of-life,bymappingthem ontoeachoftheEQ-5DandSF-6Dpreference-based-indices. METHODS: Asample of367cancerpatientsfromtheVancouverCancerCentrecompletedfourhealthrelated-quality-of-lifequestionnaires(EORTC-QLQ-C30/FACT-G/EQ-5D/SF-6D). Thesamplewasrandomlydividedtoprovidedevelopment(n 184)andcross-validation(n 183)samples.ModelsoftherelationshipsbetweentheEORTC-QLQ-C30/ FACT-Gandeachofthepreference-based-indiceswereestimatedusingregression analyses.Weexaminedthreealternativemodelingapproaches:ordinary-leastsquares(OLS);generalized-linear-modeling(GLM)usingaGaussiandistributionand log-link;andcensored-least-absolutedeviations(CLAD).Theperformanceofthe modelswasassessedintermsofhowwelltheresponsestothecancer-specic instrumentspredictedutilitiesfromeachoftheEQ-5D/SF-6Dinstruments. RESULTS: TheCLADapproachconsidersthenon-normal(left-skew)distributionof theutilityscoresandtheirapparenttruncationatone.Resultsfromthenalmodelsofthethreeapproachesdidnotdiffersignicantly.Physical,functionaland emotionalwell-beingdomain-scoresofFACT-GsignicantlypredictEQ-5D/SF-6D utilityscores.Physicalandemotionalfunctioningandpainsubscalesofthe EORTC-QLQ-C30weresignicantpredictorsoftheutilityscores.CognitivefunctioningandinsomniasubscalesoftheEORTC-QLQ-C-30weresignicantlyassociatedwiththeEQ-5D,whilethesocialandrolefunctioning,andfatiguewereonly signicantpredictorsoftheSF-6Dutility-index.Theadditionofage,gender,stage ofdisease,andethnicitydidnotleadtosignicantimprovementinthemodel.The rootmeansquareerror(RMSE)fortheSF-6Dwaslower(0.064),suggestingbetter predictionsthanfortheEQ-5D(0.098). CONCLUSIONS: ThereispotentialtoestimateEQ-5D/SF-6Dutilitiesusingresponsesfromthecancer-specicFACT-G/ EORTC-QLQ-C-30measuresofquality-of-life,eventhoughthelatterwerenotdesignedasutilityinstruments.Ourresultssuggestthatitispossibletoestimate Quality-Adjusted-Life-Years(QALYs)fromstudieswhereonlycancer-specicinstrumentshavebeenadministered.A11VALUEINHEALTH14(2011)A1–A214

PAGE 20

QA3 DOESCOST-EFFECTIVENESSANALYSISDISCRIMINATEAGAINSTPATIENTS WITHSHORTERLIFEEXPECTANCY?PauldenM,CulyerAJ UniversityofToronto,Toronto,ON,CanadaOBJECTIVES: TheUK'sNationalInstituteforHealthandClinicalExcellence(NICE) hasbeenaccusedofdiscriminatingagainstpatientswithshorterlifeexpectancyby embracingtheuseofquality-adjustedlifeyears(QALYs)withincost-effectiveness analysis(CEA).Thebasisofthisclaimisthatpatientswithshorterlifeexpectancy havefewerQALYstogainfromtreatmentandsoNICE'sguidanceinherentlydiscriminatesagainstsuchpatients.Suchcriticismsmayalsobedirectedatsimilar decisionmakingagencies.OurobjectivewastoformallyidentifythecircumstancesunderwhichCEA-baseddecision-makingdiscriminatesonthebasisoflife expectancy. METHODS: WedevelopedasimplemodelofaCEA-baseddecision makingprocessinwhichatechnologyisconsideredcost-effectiveforaparticular patientcohortonlyiftheICERforthatcohortliesbelowaxedcost-effectiveness threshold.Forsuchdecisionmakingtodiscriminateonthebasisoflifeexpectancy, theICERsfortwohypotheticalcohortsofpatientsidenticalinallwaysexceptlife expectancymustlieoneithersideofthethreshold. RESULTS: WendthatCEA doesnot inherently discriminateonthebasisoflifeexpectancybutthatscopefor discriminationarisesinthecaseofspecictechnologieshavingidentiablecharacteristics.Suchdiscriminationmayinfactfavourthosepatientswithshorterlife expectancy.Inallcasestheuseofdiscountingisshowntoreducethelikelihoodof discriminationonthebasisoflifeexpectancythisisparticularlyrelevantinlight oftherecentdiscussionaroundNICE'sdiscountingpractices. CONCLUSIONS: Itis recommendedthatagenciessuchasNICEconsiderthepossibilityofdiscriminationarisingfromtheiruseofCEA.Accusationsofinherentdiscrimination,however,appeartobemisplaced.Itisarguedthattheseclaimsarefoundedupona fundamentalmisunderstandingoftheroleofQALYsinCEA,particularlywithinthe decisionrulesadoptedbypolicymakers. QA4 ISTHEAIMOFTHEHEALTHCARESYSTEMTOMAXIMISEQALYS?AN INVESTIGATIONOF‘WHATELSEMATTERS’INTHENHSPraetC1,ShahK2,DevlinN3,SussexJ2,ParkinD4,ApplebyJ5 1Merck,London,UK,2OfceofHealthEconomics,UniversityofShefeld,London,UK,3Ofceof HealthEconomics,London,UK,4NHSSouthEastCoast,London,UK,5King'sFund,London,UKOBJECTIVES: Itisoftenassumedthattheobjectiveofhealthcareissolelytomaximisehealthusingavailableresources.ThisistheprincipleunderpinningNICE's useofcost-effectivenessanalysisbasedonincrementalcostperQALYgained.Yet researchonlocalNHSdecision-makingshowsthatcostperQALYisfarfromthe onlyconsideration.Similarly,manynationalNHSpolicyinitiativesaredrivenprimarilynotbyQALYgain,butbyprocess-of-care'andotherconsiderations.TheDH isrequiredtoundertakeandpublishImpactAssessments(IAs)identifyingthe costsandbenetsexpectedfromallnewpolicyimplementation.WeanalyseallIAs carriedoutin2008-2009toidentifythebenetsconsideredbytheDHasrelevantto itsdecisionmaking. METHODS: Thestatedbenetsofeachpolicywereextracted fromtherelevantIA.Acombinationofmethodswasusedtocategorisethese. RESULTS: 51IAswereanalysed,8ofwhichmentionedQALYgainsasabenet.162 benetsotherthanQALYgainswereidentied.Apartfromimprovinghealthoutcomes,commontypesofbenetincludedreducingcosts,improvingqualityof care,andenhancingpatientexperienceandempowerment. CONCLUSIONS: Many ofthepoliciesreviewedwereimplementedonthebasisofbenetsunrelatedto healthoutcome.ThemethodsbeingusedtoapplyamonetaryvaluationtoQALY gains(inIAcost-benetcalculations)arenotconsistentacrossIAs,orwithNICE's statedthresholdrange.WeconsidertheimplicationsforNHSdecisionmakingand NICEguidance,andthemeaningofallocativeefciencyintheNHS. POSTERSESSIONI: HEALTHCAREUSE&POLICYSTUDIES HealthCareUse&PolicyStudies–ConsumerRoleinHealthCare PHP1 USEOFSTRUCTURALEQUATIONMODELINGTOEXPLAINCONSUMER BEHAVIORTOWARDSGENERICDRUGDISCOUNTPROGRAMSPatelHK,SansgiryS UniversityofHouston,Houston,TX,USAOBJECTIVES: Thestudyattemptstoassesstherobustnessofconstructswiththe TheoryofPlannedBehavior(TPB)inexplainingintentiontousegenericdrugdiscountprograms(GDDPs). METHODS: Aself-administeredquestionnairewasdistributedtoconsumersllingaprescriptionatpharmaciesinHouston(Texas,USA) thatofferedGDDPs(CVS,Walgreens,Wal-Mart,Kroger,Target,Randalls,andH-E-B stores).ConstructsofTPB,namely,consumer'sattitudetowardsGDDPs,perceived behavioralcontrol(PBC),subjectivenorms(SN),andintentiontouseweremeasuredusingapre-validated5pointlikertscale.Thequestionnairealsomeasured consumerawarenessregardingGDDPs,attitudetowardsgenericdrugsingeneral alongwithdemographicdata.Structuralequationmodeling(SEM)usingAMOSv18 wasusedtotesttheproposedmodel. RESULTS: Responserateof59.46%wasobtained(n 389).Scalesdevelopedtomeasureallthedomainswerereliable( 0.72-0.89).Majority(68.4%)ofrespondentswereawareofGDDPs.Meanattitude towardsGDDPs(4.08 0.83)andintentiontouse(4.33 1.01)scoreswerehigh.The SEMbesttmodelwastheonewhereassociationofawarenesswithintentionwas mediatedbyattitude towardsGDDPsastherewasnodirectrelationship(CMIN/df 3.29;p 0.001;CFI 0.904;RMSEA 0.077).Further,attitudetowardsgenericswas retainedinthemodelandexertedahigherindirecteffectonintentionviaattitude towardsGDDPs.TheeffectofPBConintentionwasverylowandSNwasnot retainedinthemodel. CONCLUSIONS: Aspharmacystoresdevelopprescription drugplansusingGDDPstoincreaseutilizationandincreasestoreloyalty,strategies toimproveconsumerattitudetowardsgenericsandGDDPswillbeuseful.Continuousinformationregardingtheseprogramsmayincreaseawarenessofsuchprogramsleadingtoapositiveattitudeandincreaseduse. PHP2 TRENDSINANDPREDICTORSOFDISCOUNTGENERICMEDICATIONPROGRAM UTILIZATIONGatwoodJ,TungolA,TruongC,EricksonS UniversityofMichigan,AnnArbor,MI,USAOBJECTIVES: Toexaminethetrendsinuseandcharacteristicsofpatientsusing discountgenericprescriptionprograms. METHODS: Across-sectionalsurveyof patientsintheUniversityofMichiganTaubmanGeneralMedicineClinicwastaken overthecourseoftwo,fourweekintervals,rstinthesummerof2008andagainin 2010.Thesurveywasgivenin-person,inpaperformat,andwascompletedwhile patientswerewaitingtoseetheirphysician.Self-reportedinformationwasgatheredonthepatients'mostrecentprescriptionmedicationuseintwoparts:alisting ofcurrentmedications(locationoflling,pricepaid,brand/generic,discountprogramuse)andinquiriesonprogramuse,ifapplicable,aswellaspatient demographics. RESULTS: Theoverallconveniencesampleoverthetwocohorts included414individuals,203in2008and211in2010.Thesamplewasmostly Caucasian(78.7%),mostpatientshadprescriptiondruginsurancecoverage(92.1%), andalargemajorityofmedicationslledwereforchronicuse(84.3%).Patient demographicsbetweenthetwopopulationsweresimilarforallcharacteristics withtwoexceptions:ahighermeannumberofmedicationswasreportedin2010, 4.2versus3.4(t 3.07,p 0.01),and29.9%ofthosesurveyedin2010haduseda discountmedicationprogramversus5.4%in2008(X2 42.10,p 0.001).Factors associatedwithprogramuseincludedpatientage5064years(OR,3.79;95% condenceinterval[CI],1.459.95;p 0.004),anannualhouseholdincomeofless than$24000(OR,2.64;CI,1.126.22;p 0.02),andthellingofmedicationsfor chronicuse(OR,1.86;CI,1.202.88;p 0.005). CONCLUSIONS: Limiteddifferences inpatientsmostlikelytousediscountprescriptionmedicationprogramsexistedin thestudiedpopulation.Moreextendedanalysisofsuchprogramutilizationis neededtobetterunderstandpatientuseofsuchservices. PHP3 FEDERALENFORCEMENTACTIONSAGAINSTFRAUDANDABUSEBY PHARMACEUTICALMANUFACTURERS:1996-2010QureshiZ1,SartorO2,BennettC1 1UniversityofSouthCarolina,Columbia,SC,USA,2TulaneUniversity,NewOrleans,LA,USAOBJECTIVES: Toexpandonthescopeofthelimitedpriorresearchbyreportingon concludedinvestigationsinvolvingpharmaceuticalmanufacturersindependentof whetherornottheyinvolved quitam relatorsbetween1996-2010. METHODS: All casesinvolvedpharmaceuticalmanufacturersandFCAviolations,whichimpose liabilityofuptotripledamagesandcivilpenaltiesof$5,500to$11,000perclaimfor submittingfalseclaimstothegovernment.DatawerefromDepartmentofJustice during1996-2010. RESULTS: Since1996,resolutionof31FCAcasesinvolvingpharmaceuticalmanufacturersaccountedfor$12billioninrecoveries.Totalrecoveries were$8millionduring1996-2000(onecase),$3.9billionduring2001-2005(15 cases),and$8.1billionduring2006-2010(15cases).Billingfraudwasimplicatedin 18cases($3.7billion),off-labelmarketingin12cases($4.4billion),kick-backsin5 cases($1.7billion),andproducingdefectivepharmaceuticalsinonecase($750 million). Quitam relatorsinitiated77%ofthecases(medianreward$31million).Six settlementsincludedcriminalnes. CONCLUSIONS: Withexpansionofgovernmenthealthcare,fraudinvestigationsofpharmaceuticalmanufacturerswillundoubtedlycontinue.Theseinvestigationshavethepotentialtoreducecostsand improvethequalityofpharmaceuticaluse. PHP4 WHENSTAKEHOLDERSDON’TAGREE:DISCREPANCIESINMAB“APPROVALS” OVERTHEPASTTENYEARSMillerKL,StevensCA PAREXELConsulting,Waltham,MA,USAOBJECTIVES: Inthedrugdevelopmentprocesstheprioritiesofthestakeholders (manufacturers,payers,regulators,patients,physicians)oftendonotalign,andat timesareinconict,e.g.intheUSconsumersaren'tconcernedaboutcost,but privatepayersare.Withcostasasignicantdrivertomultipledecisionmakers,we exploremonocolonalantibodies(mAb)asadrugclassandthehandlingoftheir approvalsandauthorizationsbytheUSandUKkeybodies. METHODS: ThefollowingsourcesweresummarizedformAbsapprovedinthepast10years:theregulatoryapprovaldecisionintheUnitedStatesandtheUnitedKingdom;thepayer coveragedecision,ifapplicableandavailable;relevantpatientadvocacygroups' statements;andstatementsonbehalfofmedicalorganizations.Discrepanciesbetweeninitialregulatorydecisionsandthestatementsoftheotherstakeholders werehighlighted. RESULTS: Thesesourcesshowtheuseofclinicaldatatoadvocatedifferingstakeholders'views.Inthecaseofnatalizumab,patientadvocacy groupsandregulatorsdisagreedonwhethertherisksoutweighedthedrugbenets.Injudgingthevalueofbevacizumabforbreastcancerpatients,regulatorsand payersarestillatodds.Furthermore,thedebatebetweenthepayersandthemanufacturerontheuseofoff-labelbevacizumabseemstohingeontheclinicalresults fromtheUSNationalEyeInstitute. CONCLUSIONS: Althoughphysicianscontinue toplayacriticalroleindeterminingdruguseonceaproductisapproved,other stakeholders(e.g.patientsandindependentresearchers)areincreasingtheirswayA12VALUEINHEALTH14(2011)A1–A214

PAGE 21

inthemarket.Giventhediscrepanciesbetweenthedecisionsoftheregulatorsand thepayers,drugmanufacturersareobligedtodemonstrateaproduct'svalueto patientsandpayers,aswellastoshowthatitissafeandefcacioustoregulators. Theincorporationofthepayers'viewpointinthedevelopmentprocesswillease thegapbetweenstakeholders. PHP5 INCENTIVE-BASEDINTERVENTIONS–REWARDINGPATIENTSFORGOOD BEHAVIOURHaynesS,CostelloS,Brooks-RooneyC,HamerN CostelloMedicalConsultingLtd,Cambridge,UKOBJECTIVES: Incentive-basedschemes,inwhichpatientsarerewardedformaking behaviourchangesorreachingtreatmentgoals,arebecomingmorecommonin healthcareworldwide.Issuessurroundingthecosteffectivenessofbehaviour change,aswellastheethicalimplicationsofrewardingmembersofsocietywho exhibitinappropriatebehaviours,aresomeoftheissueswhichmustbetakeninto accountwhenconsideringtheseschemesinreal-life.Theaimofthisstudywasto provideanaccountofincentive-basedschemescurrentlyunderinvestigation,in anefforttodeterminewheretheperceivedbenetsoftheseschemesmayliefor futureimplementation. METHODS: AClinicaltrials.govsearchwasperformedto identifytrialsthatspecicallytestedtheuseofincentivesinencouragingbehaviourmodicationintargetpopulations. RESULTS: Atotalof76clinicaltrials,predominantlybasedinNorthAmerica,wereidentiedasappropriateforanalysis. Themostcommonlytargetedgroupsincludeddrugdependantindividuals,overweightindividuals(withexerciseorweight-lossprogrammes)andsmokers,implyingthattheseschemesaremostcommonlyusedtoinducelifestylechangesthat promotehealth.Mosttrialslookedattheclinical-effectivenessofincentivesat achievingbehaviourchange,withminimalemphasisonthecost-effectivenessof addingtheincentive.Themajorityofthetrials(43/76,57%)usedmonetaryor voucherreimbursementsystemsastheincentiveforpatientbehaviourmodication,with10additionalschemesprovidingtheopportunitytowinaprize.Only1 schemedirectlylinkedreimbursementtotreatmentcostsbyprovidingacycleof treatmentforfreeifthepatientsustainedcomplianceforaspeciedperiod. CONCLUSIONS: Thescopefortheuseofincentive-basedschemesisbroad,with potentialapplicationsinnumerousdifferentdiseasesinwhichgoodcomplianceis required;notjustthosewhichrequirelifestylechange.However,furthereconomic analysisofthecost-effectivenessofsuchschemesisessentialbeforetheyare implementedmorewidely. HealthCareUse&PolicyStudies–DiagnosisRelatedGroup PHP6 EXPLORETHEUSEOFDRUGSINMEDICALINSURANCEBILLINGINCHINATanZ MinistryofLaborandSocialSecurity,Beijing,ChinaOBJECTIVES: WiththeadvancementoftheChinahealthcarereformandtheincreasinghealthcareexpenditure,itisimperativetoexploreotherbillingmethods inadditiontotheexistingitem-basedbilling.Since1980s,USandmanyother countrieshaveadoptedDRGstomanageheathcareinsurancetocontrolhealth careexpenditureandachievedpositiveoutcomes.Theobjectivesofthisstudyare tounderstandexperiencesfromothercountriesintheimplementationofDRGs andmakerecommendationsonhowtoadoptDRGsinChina. METHODS: This studystartedin2003.Itiscomposedofthreephases:phaseoneistostudyhow DRGsareimplementedintheUS,Australia,andGermany;phasetwo,thecorepart ofthisstudy,istoanalyzethegroupingmethod.BasedonlearningsfromUS, Australia,andGermany,werandomlycollected700000patientrecordstaking placebetween2002and2005from12tier-oneBeijinghospitals,developedatheoreticalDRGsgroupingmodel,andcompletedover600DRGsautomatedgrouping programstailoredforChina;phasethreecontainsanalysesonthebasicrequirementofrelevantpolicyandtechnologysupporttoimplementDRGs. RESULTS: DRGshasbeendemonstratedtobeaneffectiveapproachtomanagehealthcare insuranceinmanycountries.ConsideringChina'sspecialsituations(population, healthinsurancesystemandpolicyenvironment),wedevelopedandpilotedDRGs in12tier-onehospitalsinBeijingin2010.Wewillfurtherevaluatetheimpactof thisprograminthefuture. CONCLUSIONS: Thesuccessfulimplementationof DRGsdependsonanappropriatepolicyenvironmentandmaturetechnologysupport.CurrentlyChinaisnotreadytolaunchDRGsnationallyyetduetothelackof policyenvironmentandtechnologysupport. HealthCareUse&PolicyStudies–Drug/Device/DiagnosticUse&Policy PHP7 DEVELOPINGPUBLICHEALTHGUIDANCE-WHATARETHEDATAGAPS? REVIEWOFTHEGAPSINTHEEVIDENCEIDENTIFIEDBYNICEINTHEUNITED KINGDOMKuselJ,CostelloS,HamerN,RoperS,HamerslagL CostelloMedicalConsultingLtd,Cambridge,UKOBJECTIVES: Publichealthpolicyisunderstoodtobeamajordeterminantofoverallpopulationhealth.However,developingpublichealthguidancecanbedifcult duetothelackofsufcientevidenceontheeffectivenessandcost-effectivenessof possibleinterventions.Theobjectiveofthisstudywastoidentifythemajorgapsin theevidencebasethathasbeenusedtodeveloppublichealthguidance. METHODS: ThegapsidentiedbytheUKNationalInstituteforHealthandClinical Excellence(NICE)intheevidencethatwasusedforthedevelopmentoftheir31 publichealthguidancedocumentspublishedasofDecember2010wereassessed andcompared. RESULTS: Themostprevalentdatagap,identiedby25ofthe31 guidancedocuments,wasalackofevidenceontheeffectivenessofpublichealth interventionsinspecicsubgroupsofthepopulation,particularlyethnicminorities,agesubgroupsandthosefromdisadvantagedbackgrounds.Thesecondmajor gapintheevidence,discussedby21documents,wasalackofdataspecictothe countryofinterest(inthiscase,theUK).Nineteendocumentsreportedthelackof cost-effectivenessevidenceasabarriertodevelopingpublichealthguidanceand4 ofthesespecicallynotedthedifcultyofgeneratingQALYsinthepublichealth arena.Furthermajordatagapsincludedalackofwell-designedstudies(14documents),alackoflong-termoutcomes(13),insufcientevaluationofwhichelementsofaninterventionmakeiteffective(9)andalackofevidenceontherelative effectivenessofinterventions(8). CONCLUSIONS: Researchintopublichealthinterventionsandissuesisofparamountimportance.Researchersshouldfocuseffortsuponidentifyingparticularsubgroupsinwhichinterventionsareparticularly effectiveorineffectiveandongeneratingcountry-specicdata.Cost-effectiveness dataisparticularlylacking;onesolutionwouldbetodevisenewmethodsofQALY measurementinpublichealthsituations. PHP8 PHYSICIANSHORTAGEIMPACTONPATIENTRXUSEFORSELECTCHRONIC CONDITIONSDallT1,SeidersM2,PriceP2,ChakrabartiR1,TalS1,KowalS1,LarsonJ1 1IHS,Washington,DC,USA,2AstraZenecaPharmaceuticalsLP,Wilmington,DE,USAOBJECTIVES: Thisstudyinvestigatestheextenttowhichgeographicvariationin adequacyofprimarycareandspecialistsupplyexplainsvariationinpatientuseof prescribedmedicationstotreatchronicconditions. METHODS: GeneralizedLeast Squaresregressionwithperiodrandomeffectswasusedonapooleddatasetof monthly(May2006toOct2010)IMSHealthdatafor360MetropolitanStatistical Areas(MSAs).ThedependentvariablewastotalmonthlyprescriptionsperMSAfor: (1)Statins,(2)PPIs,(3)Anti-psych,and(4)asthma/COPD.Separateregressionswere estimatedbytherapeuticcategoryandbypayertype.Explanatoryvariablesinclude monthly:(1)sizeofthepopulationwiththechroniccondition;(2)economicenvironment;(3)dummyeventvariables;and(4)levelofRxadvertising.Estimatesof theshortfallofprimarycareprovidersandspecialistsineachStatein2010are includedasexplanatoryvariables.Forstate-levelshortfall,providerdemandestimateswerebasedonnationalhealthcareuseanddeliverypatternsappliedtoeach State'spopulationcontrollingfordemographics,ratesofuninsured,andobesity rates. RESULTS: Thereexistsadirectcorrelationbetweenestimatedadequacyof primarycareandspecialist(cardiologist,gastroenterologist,psychiatrist,pulmonologist,andallergist)supplyandvolumeofprescriptions.Thepatternsare relativelyconsistentacrosstherapeuticareas.FortheStatinmarket,each1%shortfallofcardiologistsisassociatedwith0.36%,0.43%,and0.79%decreaseinStatinRx volumeforthecommerciallyinsured,Medicare,andMedicaidpopulations,respectively.Each1%shortfallofprimarycareprovidersisassociatedwithdecreasesin volumeacrossinthecommerciallyinsured(0.15%)andMedicaid(0.83%) populations. CONCLUSIONS: Thisresearchsuggeststhatcontrollingforeconomic andpopulationriskfactors,greaterinadequacyofphysiciansupplyisassociated withloweruseofprescriptionsfortreatingchronicconditions.Physicianshortages disproportionallyaffectaccesstomedicationsfortheMedicaidpopulation,followedbytheMedicareandcommerciallyinsuredpopulations. PHP9 IMPACTOFFREEPRESCRIPTIONSINWALESONPRESCRIBIN G-ATHIN DATABASESTUDYThompsonM,SedaniT,BlakBT,HardsM CegedimStrategicDataMedicalResearchLtd,London,UKOBJECTIVES: TheWelshAssemblyGovernmentintroducedfreeprescriptionsfor allon1stApril2007.IntheotherUKcountries,prescriptionswerearound£6each, althoughchildren,elderlyandotherexemptionsapplied.Thisstudyevaluated whethertheWelshpolicychangeimpactedlevelsofprescribing. METHODS: Total numbersofprescriptionitemswerecalculatedforeight12-monthperiodsbetween 1stJuly2001and30thJune2009fromTheHealthImprovementNetwork(THIN), whichcontainsanonymisedlongitudinalUKprimarycaredata.Thenumberof prescriptionswasdividedbynumberofactivelyregisteredpatientsforeachtime period,countryandagecategory(0-21,22-59and 60yearsold).Percentage changesbetweenperiodswerecalculatedandallresultsdescribed. RESULTS: In Walestherewere19.5prescriptions/patientinthe2001/2002period,whichsteadily increasedto26.9inthe2008/2009period.InEngland,prescriptions/patientincreasedfrom15.1to21.3,Scotland15.4to20.2,NorthernIreland(NI)16.7to21.9. Thepercentagechangebetween2006/2007and2007/2008was4.2%inWales,4.9% England,3.1%Scotland,4.3%NI.Thechangewithinthe0-21yearoldswas4.0%, 1.7%,2.6%and0.8%respectively.For22-59yearoldsitwas6.6%,3.0%,2.5%and 3.7%respectively.For 60yearoldsitwas2.8%,5.1%,2.9%and4.3%respectively. CONCLUSIONS: Thedescriptiveresultssuggestthatoverallthepolicychangedid notseemtoimpactlevelsofprescribinginWalesasthepercentagechangewas similaracrosscountries.However,thepercentagechangeforthe0-21and22-59 agegroupswashigherinWales,whereasthepercentagechangeforthe 60age group,whowerealreadyexemptfromprescriptioncharges,waslower.Thiscould suggestthatprescribingincreasedforpatientswhowerepreviouslymostaffected byprescriptioncharges. Futurestudiescouldevaluatewhetherprescribingofcertain drugclasses,especiallythoseincludingover-the-counterproducts,wasaffected. PHP10 THETEMPORALASSOCIATIONBETWEENPRESCRIBEDOPIOIDSANDTHE NATIONALDEATHRATEDUETOOPIOIDPOISONINGGokhaleMN,MartinB UniversityofArkansasforMedicalSciences,LittleRock,AR,USAA13VALUEINHEALTH14(2011)A1–A214

PAGE 22

OBJECTIVES: Deathsduetoopioidover-dosagehaveincreasedoverthelastdecade andnowrepresentthemostfrequentcauseofdeathduetopoisoning.Thisstudy examinedthetemporalassociationbetweenprescriptionopioidusebetween2000 and2005andthenumberofopioid-poisoningdeaths. METHODS: Thenumberof annualopioid-poisoningdeathsfrom2000-05wasobtainedfromtheNationalVital StatisticsSystemmultiplecausesofinjurymortalityles(Warner2009).Datafor annualprescriptionopioidusewasobtainedfromapublishedstudythatreported opioidusefrom2.7to3.8millionenrolleesannuallyusingtheHealthCoreBlue CrossandBlueShieldnationalcommercialinsuranceplans(Sullivan2008).The temporalrelationshipbetweenseveralopioidusemeasuresandthenumberof opioiddeaths/yearwasstudiedusingordinaryleastsquareregression. RESULTS: Thecumulativeopioiddoseamongallrecipientsincreasedfrom204mgsmorphine equivalents/enrollee/yearto371mgsandamongopioiduserswithanon-cancer paindiagnosisfrom2473mgsto3406mgs.Duringthisperiod,thenumberofyearly opioiddeathsincreasedfrom4,419in2000to10,947in2005.Therewasasignicant linearrelationshipbetweenthecumulativeyearlyopioiddose/enrollee/yearand thenumberofopioid-poisoningdeaths(p 0.0001,R2 0.98).Asignicantpositive linearrelationshipwasalsoobservedbetweenthenumberofannualopioid-related deathsandopioidusemeasuresofcumulativeyearlyopioiddose(p 0.0001, R2 0.98),opioiddose/prescription(p 0.0053,R2 0.88)andmeandays-suppliedof opioids/year(p 0.0001,R2 0.99). CONCLUSIONS: Astronglinearrelationshipbe tweenopioid-poisoningdeathsandseveralprescriptionopioidusemeasuressuggeststhattheacquisitionofprescribedopioidsthroughlegalchannelsisassociated withthenationalopioidmortalityrate.Thiscallsforaddedvigilanceonthepartof prescribersandpolicymakerstoensuresafeuseofopioidanalgesics.Further researchiswarrantedtoconrmthesecorrelations. PHP11 REVIEWOFADJUSTMENTOFESSENTIALMEDICINELISTATPROVINCIAL LEVELINCHINAShiLW1,MaYQ2,XuLP2,ZhaoDH2,ZhangY2 1PekingUniversity,Beijing,China,2AstraZeneca(WUXI)TradingCO.,Ltd,Shanghai,ChinaOBJECTIVES: EssentialMedicinepolicyisaimportantcomponentofChina'sundergoinghealthcarereform,whiletheessentialmedicinelist(the"EML")isthecore oftheessentialmedicinepolicy.OnAugust18,2009,Chinaissuednationalessentialmedicinelist(the"NEML",1stedition)with205westernmedicinesand102 traditionalChinesemedicines(the"TCM")init,andeachprovincecanaddfurther medicinestotheNEMLtoformitsownprovincialessentialmedicinelist(the "PEML")underspeciccircumstances.Thispaperaimsatsupportingforthenext roundadjustmentoftheNEMLthroughanalyzingthesituationofadditiontothe NEMLbyeachprovince. METHODS: TostudytheadditiontotheNEMLbyeach province,thispaperanalyzesthenumberofmedicinesaddedtoNEMLinall31 provincialregionsinchinaduringtheperiodofAugust18,2009toJanuary1,2011 throughsearchingtheofcialwebsites,spotinvestigationandinterviewthe ofcials. RESULTS: (1)Sofarthereare18provincialregionshavingaddedfurther medicinestotheEML;(2)Regardlessoffolkmedicine,theaveragenumberofessentialmedicinesaddedineachprovinceis187,amongwhichwesternmedicines added115whiletheTCMadded72.(3)Shanghaiaddedthelargestnumberof medicinestoitsPEML,i.e.381,whileQinghaiProvinceaddedthesmallestnumber of60.(4)SomeautonomousregionsaddedfolkmedicinestothePEMLaccordingto thelocalconditions. CONCLUSIONS: (1)it'snecessaryforChinatoadjustthe NEML(1stedition)reasonablybasedonthenumberofmedicinesaddedtothePEML anddrugusingcharacteristicsineachprovince;(2)Themoredevelopedtheeconomy,themorethefurtheressentialmedicinesadded,whichleadstounfairand lowertheaccessiblitytoessentialmedicinesinpoorareas,thusthegovernment shouldincreaseinvestmentinpoorareas. PHP13 THEIMPACTOFAQUALITYIMPROVEMENTINITIATIVEONINAPPROPRIATE MEDICATIONUSEINTHEOUTPATIENTELDERLYDudashK1,TemplinM1,KeithSW1,DelCanaleS2,MaioV1 1ThomasJeffersonUniversity,Philadelphia,PA,USA,2LocalHealthAuthorityParma,Parma, ItalyOBJECTIVES: Toassesstheinuenceofaqualityimprovementinitiativeonthe rateofpotentiallyinappropriatemedications(PIMs)inprimarycareelderly patients. METHODS: A3-year,multi-phaseprospectivedemonstrationprojecttargetingall303generalpractitioners(GPs)intheLocalHealthAuthority(LHU)of Parma,Italy,wasestablishedin2007.ApanelofParmaLHUexpertsdevelopedalist ofPIMsdeemedtobealwaysavoidedforolderpatientsaccordingtotheliterature. ThequalityimprovementinterventiondeliveredtoGPshingeduponincreasing physicians'awarenessofprescribingfortheelderlyandincludedthreekeyelements:i)theinitialdisseminationofthedevelopedPIMslist,alongwithalistof alternativedrugstoPIMs;ii)annualreviewsofPIMsprevalencedata;and(iii)a seriesofeducationalsessionsonPIMsviaacademicdetailingandcasestudies reviews.Weusedapre-postdesigntoanalyzethechangeinquarterlyratesofPIMs inelderlypatientsaged365yearsinParmaLHUbetweenthestudystart(preintervention-2007fourthquarter(Q4))andtheendofthestudy(post-intervention 2009Q4),usinganeighboringLHUsimilarinsizeasacomparator. RESULTS: The prescriptionrateofPIMsintheelderlyofParmaLHUdeclinedby19%,from6.3% pre-intervention(2007Q4)to5.1%post-intervention(2009Q4),comparedtoa12% reductioninthesametimeperiodinthecomparatorLHU,from6.8%to5.9%.The degreeofdeclineintherateofPIMswassignicantlygreaterintheLHUofParma (Breslow-Daytest,p 0.004). CONCLUSIONS: AqualityinterventionprogramlookingatimprovingknowledgeinprimarycarephysiciansonprescribingfortheelderlypatientsresultedinasubstantialreductionoftherateofPIMs.Additional approachestargetingspecicmedicationsorphysiciancharacteristicsshouldbe developedtofurtherreducePIMsrates. PHP14 THEEFFECTSOFPHARMACEUTICALCOSTCONTAINMENTONACCESSTO LICENSEDANDSUBSIDIZEDMEDICINESUNDERSINGLEPAYERSYSTEMSIN THEUNITEDSTATES,UNITEDKINGDOM,AUSTRALIAANDNEWZEALANDRagupathyR1,AaltonenK2,TordoffJ1,NorrisP1,ReithD1 1UniversityofOtago,Dunedin,NewZealand,2UniversityofEasternFinland,Kuopio,FinlandOBJECTIVES: Pharmaceuticalcostcontainmentpoliciescanaffectaccesstomedicines.ThepresentstudycomparesmedicineslicensedintheUnitedStates(US), UnitedKingdom(UK),AustraliaandNewZealand(NZ),andsubsidizedbytheUS DepartmentofVeteransAffairsNationalFormulary(VANF),UKNationalHealth Service(NHS),AustralianPharmaceuticalBenetsScheme(PBS)andNZ'sPHARMAC. Wecomparelicensedandsubsidizedmedicinesintermsof:(i)totalnumbersof entities(uniqueATCcodes),(ii)timessincerstregistration("age")ofentities,and (iii)numbersofinnovativeentities. METHODS: AllproductslistedinamajorprescribingreferencetextineachcountrywereclassiedbyATCcodeandtheirregistrationdatesrecorded.Innovativeentitiesgiven"fasttrack"approvalbytheUS FoodandDrugAdministrationor"breakthroughorsubstantialimprovement"statusbytheCanadianPatentedMedicinesReviewBoardwereidentied. RESULTS: Ofthe918entitiesand64innovativeentitieslicensedintheUS,505and20respectivelywereVANFsubsidized.IntheUK,thiswas1020and58(1016and58NHS subsidized),Australia879and49(567and30PBSsubsidized)andNZ765and39(503 and19PHARMACsubsidized).WiththeexceptionoftheUK,USlicensedentities were"newer"thanelsewhere:USmedian"age"6,607days(VANF8,203days,p 0.001),UK7,319days(NHS7,319days,p0.903),Australia7,795days(PBS8,065 days,p0.406),NZ8,936days(PHARMAC10,724daysP 0.001).NHSsubsidized entitieswere"newer"thanelsewhere.IntheUSandNZ,subsidizedentitieswere signicantly"older"thanlicensedentities. CONCLUSIONS: Differentpharmaceuticalcostcontainmentpoliciesappeartoimpactthenumberand"age"oflicensed andsubsidizedentities,alongwithaccesstoinnovativeentities.TheNewZealand systemhadthestrongestcostcontainmentlevers,butsubsidisedthefewestand "oldest"entities,andfewestinnovativeentities. PHP15 ISSUESASSOCIATEDWITHBIOLOGICAGENTS:HEALTHCARESTAKEHOLDER SURVEYToscaniM1,VogenbergR1,NashD1,PeskinS2 1JeffersonSchoolofPopulationHealth,Philadelphia,PA,USA,2MediMedia,Yardley,PA,USAOBJECTIVES: Theobjectivesofthesurveyweretogatherinformationonperceptionsaboutbiologicagentsfrompayersandproviderspresentlyandoverthenext 18months.Areasoffocusincluded:(1)Perceptionsofhealthcarereformonbiologic productdevelopment,accessandreimbursement,andoutcomesresearchrequirements;(2)Biosimilarpricing;(3)Useofspecialtypharmacyproviders;(4)Improvementsinqualityofcare,outcomesandcostsoftargetdiseases. METHODS: An onlinesurveywasconductedwithaselectsampleofnationalstakeholdersrepresentingpayersandprovidersduringthe4thquarter2010. RESULTS: Over80re spondentsrepresentingmultipletypesofpayersandproviders(ex:healthplans, fedandstategovernmentagencies,healthsystems,)participatedinthesurvey. Fiftypercent(50%)ofpayersand30%ofprovidersdidnotknowthepercentageof theirorganization'shealthcarespendonbiologics,butnearlyallknewthattheir organization'shealthcarespendonbiologicsisincreasinglyimportant.Payersand providersagree(7.7/10and7.6/10)thatpatientswillincurmoreoutofpocketexpensesforbiologicsoverthenext6-18months.The3greatestareasindicatedby providersandpayersforimprovementinqualityofcareandoutcomesare:Oncology-earlystage,RA,andAD.Itisperceivedthatgenericallyinterchangeablebiosimilarswillbeavailableatanapproximate40%discount.Respondentsfeltestablishingavaluepropositionandpromotingappropriateusewillbekeydriversof importanceforbiologicagentsinthefuture.Respondentsagreethattheirneeds regardingbiologicagentsshouldfocuson:outcomesmeasurement,comparative effectivenessstudies,specialtypharmacyusage,benetdesigns,andregulatory developmentsthrougheducationalinitiatives. CONCLUSIONS: Biologicsremaina stronggrowingareaofinterestamonghealthcarestakeholdersandmonitoring changesinmarketplacedynamicswithadditionalresearcheffortswillbekeyareas offocusinthefuture. PHP16 ANALYZINGTHEEUROPEANEXTERNALREFERENCEPRICINGSYSTEMUSING SPECTRALGRAPHTHEORYLindgrenP1,kerborg…2 1i3Innovus,SanDiego,CA,USA,2i3Innovus,Stockholm,SwedenOBJECTIVES: Manycountriesemployanexternalreferencepricingsystemwhere thepriceofpharmaceuticalisinpartgovernedbythepriceinselectedothercountriesthuscreatingawebofrelationshipsbetweenthepricesindifferentcountries. Theobjectiveofthisstudywastoinvestigatewhichcountriescouldbeconsidered themostinuentialinthisnetwork. METHODS: Thepricereferencingpattern betweencountriesintheEuropeanEconomicAreawasrepresentedasadirected graphwitheachindividualcountryasanodeandtheexistenceofpricereferencing fromonecountrytoanotherasadirectededge.Theimportanceofeachcountry wasestimatedbycalculatingtheeigenvectorcentralityofeachnodebasedonthe adjacencymatrixrepresentingtheedgesofthegraph.Inanextension,wealso performedananalysiswhereedgeswereweightedbythesizeofthepharmaceuticalmarketineachcountry. RESULTS: Whenconsideringunweightedreferencing betweencountries,GermanyandFrancewerethetwomostinuentialcountriesA14VALUEINHEALTH14(2011)A1–A214

PAGE 23

followedbyItaly,SpainandHungaryaheadofagroupofsixcountriesofabout equalimpact.Whenweighingedgesbetweencountriesbymarketsizethepicture changedsomewhatwithGermanystillthemostimportantcountryaheadofItaly, UK,FranceandSpain. CONCLUSIONS: Takingtheentirewebofpricereferencing relationshipsintoconsiderationprovidessomewhatdifferentconclusionsthan justcountingthenumberofcountriesreferencingasinglecountrywhenassessing theimportanceofthatparticularcountry.Likewise,whenincorporatingthemarketsizeofeachcountry,thisgivesadifferentresultthanjustrankingcountriesby marketsizealone. PHP17 PRESCRIPTIONDRUGMONITORINGPROGRAMS:THEPATIENTPERSPECTIVEGoodinA,BlumenscheinK,FreemanPR,TalbertJ UniversityofKentucky,Lexington,KY,USAOBJECTIVES: Manystateshaveimplementedprescriptiondrugmonitoringprograms(PDMPs)toaddressabuseanddiversionofcontrolledsubstances.Whilethe perceptionsofprescribersanddispensershavebeenexamined,thepatientperspectiveonPDMPsisnotablyabsentfromexistingresearch.Thisprojectexamined patientperceptionsofKentucky'sPDMPbyaskingMedicaidbeneciariesabout theirexperienceswiththeKentuckyAllSchedulePrescriptionElectronicReporting program(KASPER). METHODS: Medicaidbeneciariesoverage18werecontacted bymailtocompleteamodiedversionoftheConsumerAssessmentofHealth PlansSurvey.Areminderpostcardandsecondsurveywassenttonon-responders twoweeksaftertherstmailing.1,305surveyswerereturned(n 5,297,response rate 24.64%).Responseswerecodedand25%ofsurveyswerere-codedtotest inter-coderreliability.StatisticalanalysiswasconductedinSTATAv11.0. RESULTS: Of461respondents,63(13.67%)reportedthattheirhealthcareprovider haddiscussedtheirKASPERreportwiththem.Fewerrespondentsreportedthat KASPERpreventedthemfromgetting(8.11%)orlling(8.53%)aprescription.Patientswithchronicpainconditionsweremorelikelytohaveahealthcareprovider discussaKASPERreportwiththem(10.18%)thannon-chronicpainpatients (3.95%).ChronicpainpatientswerealsomorelikelytoreportthataKASPERreport preventedthemfromgettingaprescriptionthannon-chronicpainpatients(6.01% vs.1.75%).Chi-squaretestingdemonstratesthatthesedifferencesarestatistically signicant(discussedKASPERreportp 0.006;KASPERreportpreventedgetting prescriptionp 0.013).Therewerelimiteddifferencesinpatient-reportedexperienceswithKASPERbaseduponpatientgender,race,Hispanicethnicity,oreducationlevel. CONCLUSIONS: Mostrespondentsreporttheyareunaffectedbythe KASPERprogram.Respondentswithchronicpainaresignicantlymoreaptto reportdifcultygettingandllingprescriptions.Statesseekingtoformoralter PDMPsshouldconsiderpatientinputaswellasprescriberanddispenseropinions. PHP18 OPTIMALHEALTHPOLICYANDTHELINKBETWEENOUTPUTMARKETSAND DEVELOPMENTPhilipsonT TheUniversityofChicago,Chicago,IL,USAOBJECTIVES: ToexaminewhetherR&Dformedicalproductsdiffersfromother formsofR&Dandshouldthereforebeguidedbydifferentpolicysolutionsinattemptingtostimulateit. METHODS: Conceptualanalysisandregressionanalysis. RESULTS: Wearguethat,asopposedtostandardproducts,thereisalinkbetween outputmarketsanddevlopmentofmedicalproductsinthesensethatthecost developingnewmedicalproductsdependsonthequalityandpriceofconventional careinoutputmarkets.Thisisbecauseclinicaltrialsarethemajordriverofdevelopmentcostsandthatrecruitmentofsubjectsintotrialsrequiresthemtoforego conventionalcare.Thisoutputmarketlinkinducesnon-standardeffectsonmedicalR&Dcostsandinnovativereturns.First,medicalR&Dhasaself-limitingeffect becausehighqualityinnovationsmakefuturedevelopmentmoredifcultthrough prolongedrecruitment.Second,outputmarketpolicies,suchasuniversalcoverage andpricecontrols,havenotonlydemand-sideeffectsonR&D,butalsosupplyside effectsbyalteringthecostofdevelopment.Third,incontrastwithotherproduct markets,thedevelopmentspendinginacountrymaynotbedrivenbyitsrelative R&Dtalentbutbyoutputmarketpolicies.Weconsiderevidenceofthediscussed linkbetweenoutputmarketsanddevelopmentforthebreak-throughHIVtherapiesintroducedonthemarketin1996.Wedocumentadramaticdropof35%in yearlytrialrecruitmentinducedbythesenewinnovations,slowingdowndevelopmentandloweringinnovativereturns. CONCLUSIONS: Wearguethatthethefutureefcinecyanalysisofvariousformsofhealthpolicyinoutputmarketsshould beevaluatedintermsofhowitaffectsdevelopmentincentives. PHP19 TRENDSINON-LABELANDOFF-LABELPRESCRIBINGOFMONTELUKASTINA COMMERCIALLYINSUREDPOPULATION1998-2008AliMM,MartinB UniversityofArkansasforMedicalSciencesCollegeofPharmacy,LittleRock,AR,USAOBJECTIVES: Montelukasthasbeenapprovedfortheprophylaxisandchronic treatmentofasthmainpatients12monthsandolder,acutepreventionofexercise inducedbronchospasminpatients15yearsandolderandreliefofsymptomsof seasonalandperennialallergicrhinitisinpatientsasyoungas6months.We examinedtrendsinmontelukastuse,newuse,andon-labelandoff-labelusefrom 1998to2008. METHODS: Inthisretrospectivecohortstudyofadministrative claims,eligiblesubjectsintheLifelinkdatabasebetweenJanuary1997andDecember2009wereanalyzed.Studysubjectshadtobecontinuouslyenrolledfortwo years.OnlabelusewasdeterminedbysearchingforICD-9-CMcodesforapproved indications;asthma,exerciseinducedbronchospasm,andallergicrhinitis.OLS regressionwasusedtoassesslineartrendsovertime. RESULTS: 0.16%ofsubjects wereprescribedmontelukastin1998and2.44%in2008andthelineartrendof increaseinusewassignicant(R2 0.978,p 0.0001).37,793subjectswereeligible fortheanalysisofoff-labelusewithameanageof31.22yearsand43.6%weremale. Off-labelusewaslowestin1999(12.56%)andhighestin2007(24.54%)withasignicantincreasingtrend(R2 0.5776,p 0.0066)butitincreasedmoremodestly from2002(20.6%)to2008(23.91%).Overall23.02%ofsubjectsusedmontelukastoff labelandtheelderly(31.45%)andpersonsfromthesouth(28.16%)hadthehighest ratesofmontelukastofflabeluse. CONCLUSIONS: Montelukastprescribingincreasedsubstantiallyoverthepreviousdecadewherein2008,1in40personswith commercialinsurancehadatleastoneprescriptionlled.Overthreequartersof montelukastprescribingissupportedbyadiagnosisforalabeledindication.Off labeluseisgreatestintheelderlywhichmaybepartiallyexplainedbymoreevidencebasedofflabelusesintheliteratureforadults. PHP20 MEDICATIONUSESURVEYOFINPATIENTSWITHBASICMEDICALINSURANCE (2010)XiongX ChinaHealthInsuranceResearchAssociation,Beijing,ChinaOBJECTIVES: Understandingthecharacteristicsanddistributionofinpatientswho havebasicmedicalinsurancewithinChina,Analyzingthetreatmentstatusand inpatients'expenses,Studyingandsummarizingtheapplicationofvariouskinds ofmedication,Providingdatabasetogovernmentforsupportingdrugregulatory policy. METHODS: WeconductedasamplingsurveyofinpatientswithbasicmedicalinsuranceinalltheprovincesinChinabycity.In2010,weextracteddataof 157,577inpatients,andaccountingfor0.5%ofthetotalsamplesize.Thesampled datawasstatisticallyanalyzedbySQLServer2003. RESULTS: Averagehospital expensespervisitincreasedrapidlywhiledrugexpensesremainedhigh.In2009, theaveragehospitalexpenseofinpatientswithbasicmedicalinsurancereached RMB7,670Yuanpervisit,a13.6%increasedfromthepreviousyear.Thedrugexpensewasaccountedfor49.1%ofthetotal.Overallthespectrumofdiseasesof inpatientsisstablewhiletheproportionoftumorpatientswasincreasing.In2009, amongthoseinpatientsafictedbyvariousdiseases,theCardiovascularsystem diseasesaccountedfor26.75%,whichstillkepttherstposition,andTumorsaccountedfor12.35%(10.95%in2008). CONCLUSIONS: 1) Effectsfromtheagingpopulation,thehealthinsurancefundspentmostlyontheelderly,aremoreobvious.In 2009,amongurbanbasicmedicalinsuredpeople,inpatientsthatareoversixty accountedfor49.2%,with55.26%ofthetotalexpense. 2) TheMedicalInsurance Fundhadahighconcentrationofresourceuse.In2009,9.3%ofthedrugsaccounted for80%thetotaldrugexpense. 3) Therationaluseofdrugsisnotoptimistic.The expenseofantibioticmedicinesaccountedfor29.3%ofthetotalexpenseofchemicalmedicines,thenumberoftheinpatientsthathavetakenconbinationalantibioticsaccountedfor63.7%ofthetotalinpatients.AbuseofTraditional-ChineseMedicine-injectionanddrugsforadjuvant-treatmentalsoneedstobemonitored. PHP22 ASSOCIATIONOFSELF-REPORTEDHEALTHSTATUSWITHCOST-RELATED MEDICATIONNONADHERENCEANDOUT-OF-POCKETPHARMACYSPENDING AMONGSENIORMEDICAREBENEFICIARIESTangY UniversityofPittsburgh,Pittsburgh,PA,USAOBJECTIVES: Toexaminetheassociationofself-reportedhealthstatuswithcostrelatedmedicationnonadherence(CRN),aswellaswithprescriptiondrugout-ofpocketspending(OOP)amongseniorMedicareenrollees. METHODS: Weanalyzed the2008HealthandRetirementStudy(HRS),anationalsurveyofadultsaged50or older.Self-reportedhealthstatuswasclassiedintothreecategories:excellent (excellentandverygood),good,andpoor(fairandpoor)health.Weestimatedthe associationbetweenself-reporthealthstatusand1)cost-relatedmedicationnonadherenceusingLogisticregression;2)monthlyout-of-pocketpharmacyspending usinggeneralizedlinearmodel(GLM)controllingforsocio-demographiccharacteristicsandphysicalhealthconditions.Allestimateswereweightedtobenationally representative. RESULTS: Among9,090Medicarebeneciariesaged65orolder (meanage74.7years),56.1%werefemale,5.8%reportedtakinglessmedication becauseofcost.Comparedtothosewithexcellenthealthstatus,beneciarieswith goodhealth(OR 1.643,p 0.0001,95%condenceinterval[CI]1.263-2.138)and poorhealth(OR 2.947,p 0.0001,95%CI2.264-3.834)weremorelikelytoreport cost-relatedmedicationnonadherence.TheGLMregressionresultsshowedthat comparedtoexcellenthealthbeneciaries,theaveragemonthlyout-of-pocket pharmacyspendingwas$13.844(p 0.0001,95%CI7.500-20.186)higherforgood healthand$27.245(p 0.0001,95%CI17.695-36.794)higherforpoorhealthsubjects. CONCLUSIONS: Drugspendingisheavilyconcentratedamongindividuals65years orolderandcost-relatedmedicationunderuseamongseniorshasbeenabigpublic healthconcern,especiallybeforetheimplementationofMedicarePartDin2006. Thisstudyndthatworsehealthstatusissignicantlyassociatedwithincreased likelihoodofcost-relatedmedicationnonadherenceandhigherout-of-pocketexpendituresforprescriptiondrugs.SeniorMedicarebeneciarieswithpoorhealth stillfacedhigherOOPburdencomparedwiththeirhealthycounterparts. PHP23 THEIMPACTOFMEDICAREPARTDONPRESCRIPTIONVOLUMEAND OUT-OF-POCKETCOSTSOFMEDICALLYNECESSARYDRUGSSepulvedaB,DoyleJ,WhiteC Quintiles,Hawthorne,NY,USAOBJECTIVES: WiththeJanuary2006introductionofMedicarePartDPlans(PDP) intendedtoincreasedrugaccess,wehypothesizedthatitwouldimpacttrendsin prescriptionvolume(TRx)andout-of-pocketcosts(OPC).Weevaluatedtheseun-A15VALUEINHEALTH14(2011)A1–A214

PAGE 24

dercommercialplansandPDPfor2high-usedrugsineachofthesixtherapeutic classesdeemedmedicallynecessarybyCMSforbeneciariesaswellastwostatins, forreference. METHODS: TRxandOPCwerecollectedannuallyfrom2002to2009 forthetwomostprescribeddrugsinthesixmedicallynecessaryclasses:anticonvulsants,antidepressants,oralantineoplastics,antipsychotics,immunosuppressants,andHIV/AIDSplusstatins(14drugstotal).DataarefromSDI'sVONAand VOPA,analyzedwith2-wayANOVA;signicanceatp 0.05. RESULTS: Alldrugs analyzedhadannualincreasesof 10%peryearinPDPprescriptionvolume.However,ninedrugshadsignicantlygreaterannualincreasesincommercialplan prescriptionvolumesbeforePDPthanafter,comparedtoonlyonewithsignicantlygreaterincreasesafterPDP.Sixdrugshadsignicantlygreaterincreasesin overallprescriptionvolumeafterPDPcomparedtobefore.Fivedrugshadannual OPCincreasesofat 10%peryearinPDPcomparedtotwothatpostedatleast10% annualdecreases.SixdrugshadsignicantlygreaterannualOPCincreasesbefore PDPcomparedtoafterundercommercialplans;onlyonehadgreaterincreases afterPartDcomparedtobefore.Fivedrugsshowedsignicantlygreaterincreases inoverallOPC(allpaymentformsaveraged)beforePDPcomparedtoafter. CONCLUSIONS: Incommercialdrugplans,theadventofPartDcorrelateswith slowerincreasesindrugutilization.However,PartDutilizationhasincreasedsignicantlyforalldrugs,indicatingagreatershifttowardPartDplans.Thatonlyve drugsunderPartDincreasedinOPCbymorethan10%(andtwodecreasedoverall) suggeststhatthisshiftmaybeduetogreaterpatientaccess. PHP24 NUMBERANDTYPEOFDRUGSUSEDBYSENIORSONPUBLICDRUGPROGRAMS INCANADA,2002TO2008HuntJ,GaucherM CanadianInstituteforHealthInformation,Ottawa,ON,CanadaOBJECTIVES: Theuseofmultiplemedicationscanincreasetheriskofadverse effects,druginteractionsandnon-compliancewithdrugtherapy,allofwhichmay resultinless-than-optimalhealthoutcomes.Seniorsareataparticularlyhighrisk ofadverseeffects.Althoughitmaybeappropriateforapatienttobetakingahigh numberofmedications,theadditionalrisksshouldbeconsidered.Thisanalysis examinesthenumberandtypesofdrugsbeingusedbyseniors,andhowutilization changesasseniorsage. METHODS: ClaimsleveldatafromtheNationalPrescriptionDrugUtilizationInformationSystem(NPDUIS)Databasewereanalyzedfor 1,039,642seniorsonpublicdrugprogramsinsixCanadianprovincesbetween2002 and2008,representingover80%oftheseniorpopulationinthoseprovincesin 2008.DrugclassesweredenedusingtheWorldHealthOrganization'sAnatomical TherapeuticChemicalclassications.Thenumberofdrugswascalculatedasthe numberofuniquedrugclassesapersonclaimedinagivenyear. RESULTS: In2008, 21.4%ofseniorsonpublicdrugprogramshadclaimsfor10ormoredrugclasses,a slightincreasefrom2002(18.6%).Thenumberofdrugclassesusedbyseniors increasedwithage.Themostcommonlyuseddrugclasseswereusedtotreat chronicconditionssuchashighcholesterol,highbloodpressure,heartfailure,and emphysema.3-hydroxy-3-methyl-glutaryl-CoenzymeAreductaseinhibitorswere themostcommonlyuseddrugclassamongseniorsaged65to84,whilesingleingredientangiotensin-convertingenzymeinhibitorswerethemostcommonly usedclassamongthoseaged85andover. CONCLUSIONS: Findingssuggestahigh proportionofseniorsmaybeatriskfordruginteractionsandotheradverseevents duetothenumberofmedicationstheyaretaking.Thisillustratestheimportance ofmedicationmanagementstrategiesforseniors,andtheneedforcommunication betweenhealthcareprovidersregardingseniors'drugregimens. PHP25 THEPOTENTIALVALUEOFTHE2014MEDICAIDEXPANSIONSepulvedaB,DoyleJ Quintiles,Hawthorne,NY,USAOBJECTIVES: Withtheimminent$40billionexpansionofMedicaidin2014,health insurersfacenewopportunitiestobenetbyprivatelymanagingthesegrowing plansforsomeofthelargerstates.Overall,theStateswillseea32%increase approximately16millioninMedicaidenrolleesin2014(UnitedHealth).Weanalyzedthepotentialwindfallofthisexpansioninthetenstateswiththelargest projectedgrowthinveofthelargestdrugmarkets. METHODS: Prescriptionvolume(TRx)andout-of-pocketcosts(OPC)inveofthelargestdrugmarkets(depression,diabetes,hypertension,cholesterol,pain)wereobtainedforthetenstates withthelargestMedicaidenrollmentincreases(California,Texas,Florida,Pennsylvania,Ohio,Michigan,Georgia,NorthCarolina,Illinois,Louisiana;Source:UnitedHealth).DatawereobtainedfromJune2006toJune2010;SDIVONAandVOPA databases. RESULTS: MedicaidprescriptionsrepresentasignicantportionofTRx inthe10statesanalyzed,from2.7%(MI)to8.9%(IL).Furthermore,MedicaidaccountsforaconsiderablefractionofTRxineachofthedrugmarketsanalyzed, from1.5%(cholesterol)to3.2%(pain).TheaverageannualchangesinMedicaid prescriptionswerehighlyvariablefrom 7.5%(OH)to8.2%(IL).Averagenational OPCamongthesedrugmarketshavedecreasedanaverageof20.1%peryearcomparedtoadecreaseof4.8%amongprivateinsurers. CONCLUSIONS: Medicaidalreadyrepresentsasignicantportionofprescriptionvolume,andthatpatientaccess isconstantlyincreasingwithprecipitouslydecreasingOPC.Withtheimpendingrisein Medicaidenrollmentin2014,prescriptionvolumeswillincreasefurther,representing agreatopportunityfortheexternalmanagementofMedicaidplans. PHP26 ANALYSISOFTHESIXPROTECTEDMEDICATIONCLASSESBASEDONPLAN TYPEANDLOWINCOMESUBSIDYSTATUSBlackwellS,WaldronC,EvansM CentersforMedicare&MedicaidServices(CMS),Baltimore,MD,USAOBJECTIVES: Theprimaryresearchquestionwastoidentifytrendsincostofthesix protectedmedicationclassesinthePartDprogrambasedonplantypeandlow incomestatusbetweencalendaryears2007and2008.Thesecondresearchquestionwastoidentifytrendsingapphaseandcatastrophicphaseentry. METHODS: TheprimarydatasourcewastheprescriptiondrugeventdataintheChronic ConditionWarehouse(CCW)for2007and2008.Datawereination-adjusted. Resultsarebasedonanalysisof100%dataforMedicarebeneciariesinthe CCW.Thesixprotectedmedicationclassesunderstudywereanticonvulsants,antidepressants,antineoplastics,antipsychotics,antiretrovirals,and immunosuppressants. RESULTS: BeneciariesenrolledinPDPshadhigherdrug costsperpersonacrossalldrugclassesforbothyearscomparedtothoseinMAPDs.LISenrolleeshadhigherdrugcostsperpersonforeachyearforthedrug classescomparedtonon-LISenrolleeswiththeexceptionofantineoplastics.These ndingswereinthecontextofnumberofprescriptionslledperenrolleeincreasinglessthanonepercentbetweencalendaryears2007and2008foralldrugclasses. OddsratiopointestimatesbetweenthetwoyearsforgapphaseentryandcatastrophicphaseentryindicatethatPDPandLISenrolleesweremorelikelytoreach thegapandcatastrophicphasesascomparedtotheircounterparts(i.e.,MA-PDand non-LISenrollees)forbothcalendaryears. CONCLUSIONS: Findingssuggestthat drugcostsperpersonincreasedfrom2007to2008.Ingeneral,theincreasewas higherforMA-PDenrollees(comparedtoPDPenrollees)aswellasforLISenrollees (comparedtonon-LISenrollees).Ahigherlikelihoodofgapphaseandcatastrophic phaseentryexistedforPDPenrollees(comparedtoMA-PDenrollees)andforLIS enrollees(comparedtonon-LISenrollees)forallclasses. PHP27 CANADIANPUBLICDRUGPROGRAMSPENDINGONSENIORS,2002TO2008HuntJ,GaucherM CanadianInstituteforHealthInformation,Ottawa,ON,CanadaOBJECTIVES: Seniorsareestimatedtoaccountforabout40%ofallCanadianretail spendingonprescriptiondrugs,andmakeupasignicantproportionofpublic drugprogramexpenditures.Thisanalysisexaminesthetypesofdrugsaccounting forthemajorityofpublicdrugprogramspendingforseniors,andthedistributionof programspendingacrossseniors. METHODS: Thisstudyexaminedclaimsfor 1,039,642seniorsonpublicdrugprogramsinsixCanadianprovinces(Alberta,Saskatchewan,Manitoba,NewBrunswick,NovaScotiaandPrinceEdwardIsland)from 2002to2008,representingover80%oftheseniorpopulationinthoseprovincesin 2008.Publicdrugprogramspendingonseniorsrefersonlytotheportionofprescriptioncostspaidbythepublicdrugprogram,includingprofessionalfeesand markup,whereapplicable. RESULTS: In2008,thetop10drugclasses,intermsof publicdrugprogramspending,accountedforalmosthalf(48.3%)ofallspendingon seniorsinthesixprovinces.3-hydroxy-3-methyl-glutaryl-CoenzymeAreductase inhibitorsaccountedforthehighestproportionofdrugprogramspendingforseniors,while4ofthetop10classeswereusedtotreathypertension.Almosthalf (45.5%)ofpublicdrugprogramspendingwasforasmallgroupofseniors(14.7%), wherethedrugprogramspaid$2,500ormoreoftheirannualdrugcosts. Tumour necrosisfactoralphainhibitorswerethefastestgrowingdrugclass,intermsofspending,growingatanaverageannualrateof58.4%between2002and2008. CONCLUSIONS: Findingsshowalmosthalfofdrugprogramspendingforseniorsison asmallnumberofdrugs.Thissuggeststhatanycostsavinginitiativesthatimpact spendingonthesedrugscouldhaveasignicantimpactonoverallpublicdrugprogramspending. PHP28 COMPLIANCEWITHTREATMENTGUIDELINERECOMMENDATIONS:USEOF OVER-THE-COUNTERMEDICATIONSINPATIENTSRECEIVING BISPHOSPHONATES,ANTIEPILEPTICS,ORCHRONICOPIOIDSBarnesKD,HeatonPC UniversityofCincinnati,Cincinnati,OH,USAOBJECTIVES: ThisstudydeterminedthefrequencyofOTCmedicationusefollowingtreatmentguidelines,specicallythepercentageofpatientvisitsrecordedin theNationalAmbulatoryMedicalCareSurvey(NAMCS)andNationalHospitalAmbulatoryMedicalCareSurveyoutpatientdepartment(NHAMCS)thatincluded(1) bisphosphonateswithcalcium,(2)divalproexorcarbamazepinetherapywithfolic acid,and(3)fentanylormethadonetherapywithalaxative. METHODS: Aretrospective,observationalstudywasconductedusingNAMCSandNHAMCSdata from2006-2008.Dataforvisitsincludingbisphosphonateswithcalcium,divalproexorcarbamazepineinwomenofchildbearingagewithfolicacid,andmethadoneorfentanylinadultsoverage40yearswithlaxativeswerecollected.Descriptivestatisticsandlogisticregressionwereusedtoanalyzeweighteddatato producenationalestimates. RESULTS: Calciumwasusedin25.3%ofvisitswith bisphosphonatetherapy;visitsassociatedwithantiepilepticsandchronicopioids hadlowerratesofsupplementation,with5.0%includingfolicacidand3.7%includinglaxatives.VisitsbyAsians(OR 2.74;CI1.42-5.27)andpatientsfromthenortheast(OR 1.57;CI1.10-2.24)ormidwest(OR 1.65;CI1.15-2.37)weremorelikelyto includecalcium,whilevisitsbypatientswithMedicaid(OR 0.48;CI0.30-.0.78)were leastlikelytoincludecalcium.Invisitswithantiepileptics,Asianwomen(46.0%) andwomenwithprivateinsurance(9.5%)weremorelikelytoreceivefolicacid.In visitsthatincludedchronicopioids,only1.2%ofpatientsage40-49yearsincluded alaxativewhile6.7%ofvisitsbypatientsover70yearsincludedalaxative.Laxative usewaslowinvisitsbyAfricanAmericanpatients(1.8%)andpatientsresidingina zipcodewith 12.84%ofthepopulationholdingabachelor'sdegree(1.9%). CONCLUSIONS: ThemajorityofpatientswerenotreceivingOTCsupplementation vitaltothetreatmentofosteoporosis,epilepsy,andchronicpain.Targetedinter-A16VALUEINHEALTH14(2011)A1–A214

PAGE 25

ventiontoimprovetreatmentguidelineadherencewouldincreasecompliance withrecommendations. HealthCareUse&PolicyStudies–EquityandAccess PHP29 EVALUATINGTHEPERFORMANCEOFANINNOVATIVEPUBLICHEALTH INSURANCE:THECASEOFADECENTRALIZEDPROVINCEINARGENTINAMaceiraD CEDES,BuenosAires,ArgentinaOBJECTIVES: InArgentina,thepublichealthsystemisdeeplydecentralizedand organizedmainlyattheprovinciallevel.Inthiscontext,differencesregarding incomedistributionandaccesstohealthservicesrequirethecreationofregulatory devicesandStateintervention.OneoftheanswersprovidedbytheBuenosAires HealthMinistrywasthecreationofProvincialPublicHealthInsurance.Thisprogramincludedaninnovativemechanismofhumanresourcespaymentforthe publicsectoracapitatedsystemwiththepossibilitytogenerateaplusoverthe xedsalary-andaimedtoreachthecoverageof2.4millionofcitizens.Theobject ofthisresearchistoanalyzetheperformanceofthestrategyintermsoffocalizationanddevelopmentofpreventivetasksassociatedtoahigherqualityatthe primaryhealthcarelevel. METHODS: Adatabaseof1.7millionofconsultationsin 59municipalitiesduringtheperiod2004to2006wasanalyzed,consideringthe evolutionofpreventiveandcurativeconsultations.Logisticregressionmodels wereimplementedtodeterminethevariablesconditioningthosetendencies. RESULTS: Theprogramwasappropriatelyfocalizedinpoormunicipalities.Preventiveconsultationsincreased107%duringthethree-yearperiod,andcurativeconsultationsdecreasedby56%.Preventiveconsultationsweresignicantlyassociatedwithyoungandfemalepatients,aswellasyoungerphysicians.Also,poorand populatedmunicipalitiesshowedhighassociationswithpreventiveconsultations. Differencesbetweenmunicipalitieswererelevant,showingsignicantassociationsonbothsigns. CONCLUSIONS: Theinsurance'shiringandpaymentmechanisms,resulttobeinnovativeandsuccessfulinthecontextofpublichealthsubsector,conditioningtheprofessionalstoincreasetheireffortstowardshigher qualitypreventivecare.Thestrategymightbeseenasoneoftheinstrumentswith thepotentialtoenhancecarequalityandperformance.Nevertheless,municipalitiesshowparticularcharacteristicsregardingtheirmanagementandadministrativestructuresthataffectthesuccessoftheprogram. PHP30 CANPRICINGSCHEMESIMPROVEMARKETACCESSFORINNOVATIVEHIGH PRICEDDRUGS?AnastasakiE,ColasanteW,Imbeah-AmpiahR PriceSpectiveLtd.,London,UKOBJECTIVES: Europeanpayersarefacinghighlevelsofuncertaintyaboutthevalue ofinnovativedrugsandtheirbudgetimpact.Theseuncertaintiesarerelatedto boththedrug'sactualusagesuchasdosingrequirementsaswellasexternalfactorssuchaspatients'clinicalresponseanddrug'srelativeeffectiveness.Overthe years,pricingschemeshavebeenproposedtoincreasebudgetpredictability.Due tothevaryingperceptionsofattractiveness,theseschemesarenotsimilarlyadoptedacrossEurope.Theresearchassessedtheattractivenessofvariousschemes asawaytoreducebudgetimpactuncertaintiesandimprovepatientaccessto innovativehighpriceddrugs. METHODS: Bothprimaryandsecondaryresearch wasconducted.Thesecondaryresearchofpublisheddata,suchaspayers'assessmentsofpricingschemesandcountries'pastpoliciesallowedustodevelopa frameworktounderstandtherelativeweightsofthefactorsaffectingpayers'uncertaintyregardinginnovative-highpriceddrugs.Theframeworkwasthenvalidatedthroughqualitativeprimaryresearch. RESULTS: Theresultsrevealedpayers' preferencetowardsthecostpercycleproposition,becauseofitseaseofimplementation.Nevertheless,ofthecountriesexamined,onlytheUKandItalywereexperiencedwithpricingschemes.Fewothersappearedtoberelativelyopento schemes,sometimesatthelocalleveland/orwhenmoretraditionalapproaches arepresent.Mostcountrieswereveryresistanttopricingschemes,primarilydueto theperceivedcomplexityofschemesmanagementandthetransparencyassociatedwithit.Respondents'opinionsregardingpricingschemesalsovariedinterms ofthelineoftherapytheyaretargetedto. CONCLUSIONS: Theheterogeneityof Europemakesitdifculttohaveoneschemetall.Schemescanreduceuncertaintiesassociatedwiththerapyoutcomesandbudgetaryexpenditures;howeverpayersarenotwillingtotakeadditionalrisksassociatedwithpricingschemesfor innovativehighpricedrugsandprefertraditionalstraightdiscounts. PHP31 REVIEWOFHIGHLYSPECIALIZEDMEDICALSERVICESINKAZKAHSTAN WITHINANINTRODUCTINOFUNIFIEDNATIONALHEALTHSYSTEMYergaliyevK,IssayevaR,UmralinT MinistryofHealthofRepublicofKazakhstan,Astana,KazakhstanOBJECTIVES: Asaconsequenceofthesweepingreformsin2010Kazakhstanhas introducednewlyUniedNationalHealthSystem(UNHS),whichaimsatimprovingthehealthofcitizensthroughfreechoiceofdoctorsandmedicalorganizations, andintroductionofnewnancialmodelsandeconomicincentivesforhealthcare providers,amongothers.WithinUNHSreimbursementofhighlyspecializedmedicalservices(HSMS)isbasedonincreasedtariffs.AccordingtoLawofKazakhstan, HSMSareservicesprovidedusingnew(high)healthtechnologies.Objectivesfor thepaperweretoreviewcurrentconditionofHSMSinthecountry,toidentify problemsandtoprovideappropriaterecommendationstomeetneedsofevolving UNHS. METHODS: RetrospectiveanalysisofhealthservicesprovidedbySecondary careregionalmedicalcenters(Regionalcenters)(n 848)andRepublicanleveltertiarycaremedicalcenters(Republicancenter)(n 22)over10monthsin2010,reviewofregulatorydocumentsrelatedtoHSMS. RESULTS: 6%ofRegionalcenters and59%ofRepublicancentersprovideHSMS.Only12%ofcasemixofRepublican centerisHSMS.AlistofHSMStechnologiesthatisapprovedbyMinistryofHealth mostlyconsistsofsurgeryprotocols(94%).Therearenoclearcriteriatodene HSMS.Republicancentershavetransferred37hightechnologiestoRegional centers. CONCLUSIONS: IntroductionofUNHScreatedabaseforcompetitiveenvironmentinKazakhHealthcaresystem.Firsttimeinahistory,Regionalcenters havestartedtoprovideHSMSandRepublicancentersreportaboutsignicantdeclineinanumberofpatientsthatneedHSMSandreductionofwaitinglist.Asa result,anaccessofKazakhpopulationtoHSMShasimproved.Thesendingshave shownanurgentneedforhealthtechnologyassessmentmethodologyandtools, asithelpstodenetrulyeffectivehealthtechnologyandtomanagescarcegovernmentbudgetforhealthservices. HealthCareUse&PolicyStudies–FormularyDevelopment PHP32 ANALYSISOFPHARMACISTS’INTERVENTIONSONELECTRONICVERSUS TRADITIONALPRESCRIPTIONSINTWOCOMMUNITYPHARMACIESGilliganA,MillerK,MohneyA,MontenegroC,SchwarzJ,WarholakTL UniversityofArizona,Tucson,AZ,USAOBJECTIVES: Thepurposeofthisstudywasto:1)measuretheincidenceofprescriptionproblemsthatrequirepharmacistintervention;2)determinethetypes andrelativefrequencyofprescriptionconveyancewhichcontainproblemsthat requirepharmacistintervention;and3)estimatethepharmacypersonneltimeand relatedpracticeexpensesforprescriptionsrequiringintervention. METHODS: This prospectivestudyutilizedanadaptationofthemedicationtherapyintervention (MTI)datacollectionform.Datawerecollectedattwochaingrocerystorepharmaciesbydirectlyobservingpharmacistsfor3weeks,duringthehoursof9amto6pm. Informationrecordedwitheachinterventionincludedthetypeofprescription, medicationinquestion,reasonsforintervention,interventionoutcome,andtime spentperformingintervention.Chisquareanalysiswasconductedtocompare percentageofinterventionratesbetweenprescriptionconveyances.Kruskal-Wallisranktestwasutilizedtocomparethetimeontaskvaluesfortheinterventions. Poissonregressionwasusedinordertodetermineifsignicantdifferencesexisted amongthetotalnumberofinterventionspergroup. RESULTS: Pharmacistsreviewed1678newprescriptionsandintervenedon157(9.4%)overa13dayperiod.A totalof12hoursand11minuteswasrequiredtoperformallinterventionsforan overallaverageof4.79(SD 0.34)minutesperintervention.Percentageofinterventionratesbetweenprescriptionconveyanceswasnotstatisticallysignicant (p 0.21).Thetimeontaskvaluesfortheinterventionswerealsonotstatistically signicant(p 0.39).However,comparedwithhandwritteninterventions,e-prescribinginterventionsoccurred1.57timesless(p 0.0001),faxedinterventionsoccurred1.52timesless(p 0.0001),andverbalinterventionsoccurred2.05timesless frequently(p 0.001),onaverage. CONCLUSIONS: E-prescribinginterventionsoccurredlessfrequentlycomparedwithhandwritteninterventions,whichmaypresentapotentialbenet.Futureeffortstodevelopandenhancee-prescribingare neededtoaddmeasurablevaluetopatientcare. PHP33 IMPACTOFPHARMACYANDMEDICALPLANINTEGRATIONONOVERALL MEDICALCOSTSBunzTJ,NguyenHV,RegineML CIGNAHealthCare,Bloomeld,CT,USAOBJECTIVES: StudiespublishedbyAetnaandHealthPartnershavedemonstrated thevalueofintegratedmedicalandpharmacybenetsintheirpopulations.Combinedcoverageprovidestheopportunityforcoordinatedpatientoutreachand alignedpolicydecisions.Thepurposeofthisanalysisistodetermineifintegrated benetsresultinlowertotalmedicalcostswhencomparedtoplansthatutilizea carve-outpharmacybenet. METHODS: Aretrospectivematchedcohortstudywas utilizedinordertocomparethetotalmedicalcostsassociatedwithcustomersin integratedhealthplanswithcustomerswithacarve-outPBM.Individualcustomerswereincludediftheywerecontinuouslyenrolledforatleast12monthswithin thesameplantype,hadafullsetofdemographic,risk,andclinicaldata,andwere betweentheagesof18and65.Thetwogroupswerematchedbasedondemographic,comorbidity,andemployercharacteristics. RESULTS: Thisanalysis matched39,896customersin2007(19,948ineachgroup)and151,144customersin 2008(75,572ineachgroup)forthenalanalysis.Medicalcostswere$13.77(p-value 0.067)and$11.39(p-value0.006)percustomerpermonthlowerforcustomerswithan integratedpharmacybenetascomparedtocarve-outcustomersin2007and2008 respectively.In2007and2008thebulkofthesavings($6and$7pcpmrespectively) wereassociatedwithlowerutilizationofoutpatientservices. CONCLUSIONS: When controllingfordemographic,clinical,andemploymentcharacteristics,thereisa signicantsavingsassociatedwithintegratingmedicalandpharmacybenets. PHP35 DECISIONMAKINGINBRAZILBASEDONHEALTHTECHNOLOGYASSESSMENT: THEGOOD,THEBAD,ANDTHEFUTUREMachadoM1,FonsecaE2,FonsecaM3 1GlaxoSmithKlineBrazil,RiodeJaneiro,Brazil,2FederalUniversityofS‹oPaulo,S‹oPaulo, Brazil,3Axia.Bio,S‹oPaulo,BrazilOBJECTIVES: DespitethepremiseofuniversalaccesssupportedbyitsConstitution, theBrazilianhealthcaresystemissubjecttosimilarnancingissuesasinother jurisdictionsworldwide.AmisunderstandingbetweenConstitutionalrightsandA17VALUEINHEALTH14(2011)A1–A214

PAGE 26

healthcarebudgetrestrictionsleadstothecontroversialmodelofproceduraljustice.Theobjectivewastodescribethecharacteristicsofthecurrentdecision-makingscenarioinBrazilusingHealthTechnologyAssessment(HTA),itsmainactors, consequencesandfutureopportunities. METHODS: AreviewofHTAregulationin Brazilandthepublishedliteratureonthetopicwasconducted.TheBrazilianHTA modelwasdiscussedandstrengthsandweaknesseshighlighted. RESULTS: Over theyears,drugexpendituresandhealthcaredemandhavegrownexponentiallyin Brazil.In2006,theBrazilianMinistryofHealth(MoH)createdaHTAprocess(Regulation#152).Themainobjectivewastoalignthedynamicsofhealthtechnology decisionswithpopulationsocialneeds,epidemiologicalprole,andhealthcare systemnancingcharacteristics.Theprincipalcomponentsareaclinicalreview andeconomicanalysisoftechnologiesunderevaluation.Multiplegovernment playersareinvolved,butnoreferencetoexternalinvolvement.Healthtechnology decisionstakeintoconsideration:a)therelevance/impactofthetechnologywithin thesystem,andb)thesafety,effectiveness,andcost-effectivenessoftheappraised technologies.Noreferencewasfoundtowhatisconsideredcost-effective/efcient forthesystem.ProposedHTAregulationchangesareunderwayandinclude:rapid response,societalengagement(i.e.,publicconsultationofMoHdecisions),and changesindecisiondeterminants(i.e.,economicimpacttobeleastimportant). CONCLUSIONS: Currently,Brazilholdsaninformedhealthtechnologydecisionmakingprocess.However,clear/objectiveparametersforaidingdecisionshavenot beendescribedoryetproposed.Uncertaintystillremainswithregardstoefcient healthtechnologydecisions.CurrentinitiativestoimproveHTAprocessarecontroversial.Inthefuture,HTAcouldserveasamediatorbetweenconstitutional obligationsandcurrentbudgetaryrestrictionsinthecountry. HealthCareUse&PolicyStudies–HealthCareCosts&Management PHP36 CATASTROPHICINJURY-RELATEDENROLLMENTANDCOSTSINASTATE MEDICAIDFEE-FOR-SERVICEPROGRAMSmithMJ1,ZaloshnjaE2,MillerTR2 1UniversityofOklahoma,Tulsa,OK,USA,2PacicInstituteforResearch&Evaluation,Columbia, MD,USAOBJECTIVES: 1)Calculatethenumberofenduringenrolleeswhoenteredastate Medicaidprogrambecauseofcatastrophicinjury,and2)estimatelifetimeMedicaidcostsoftheenrollees. METHODS: De-identiedadministrativedatafroma stateMedicaidprogramthatincludedpaidfee-for-serviceclaimsformedicalservicesandprescriptionmedicationsdatedbetween2000and2005,aswellasan eligibilitylelistingenrollmentperiodsofrecipients,werethedatasource.The studyisaprospective,longitudinal,historicalcohortdesign.Inpatienthospitalizationclaimswithadiagnosisofinjuryinanyoftherstthreediagnosiseldswere extracted.Thestudysamplecomprisedrecipientslessthan65yearsoldwhose eligibilityforMedicaidbeganontherstofthemonthinwhichahospitalizationfor injuryoccurred.Thecohortwasfollowedforthedurationofeachrecipient'sinitial eligibilityperiod.Costswerediscountedat3%andfromtheperspectiveof Medicaid. RESULTS: Eligibilityof2089recipientsstartedthemonththeywerehospitalizedforinjuryoverthestudyperiod,arateofapproximately1.1per1000 recipientsperyearand6.0per1000newrecipientsperyear.Amongthe2089 cohort,364(17%)recipientshadopen-endedeligibilitythatcouldlasttheduration oflife;amajorityofwhomwereover20yearsold.Adultswithinjurywithopenendedeligibilityaccountedforapproximately1.5per1000newadultrecipientsper year.TheestimatedlifetimeMedicaidcostformedicalservicesandprescription drugsfora30-yearoldmalerecipientwithinjurywithlifetimeeligibilityequaled $76,833. CONCLUSIONS: CatastrophicinjurycanleadanindividualtoenterMedicaid.OneinsixentrantsmayrequirelifetimeMedicaidcoverage.Inthisstudy,the rateoflifetimeeligibilityforMedicaidoccurredwithlowincidence.However,the costofinjurytotheMedicaidprogramoftengreatlyexceedsthecostofinjury treatmentalone. PHP37 AREGISTRYPERSPECTIVEOFOFHOSPITALDISCHARGECOSTS:PILOTRESULTS OFTHE10MOSTFREQUENTDIAGNOSISANDPROCEDURESOFTHE ARGENTINE-HEALTHCARECOSTANDUTILIZATIONPROJECT(A-HCUP)InsuaJT1,GiuntaD2,IoliP3,VillalonR1 1HospitalUniversitarioAustral,UniversidadAustral,Derqui,Argentina,2HospitalItaliandode BuenosAires,Caba,Argentina,3HospitalPrivadodelaComunidad,FundacionMedicadeMardel Plata,MardelPlata,BuenosAires,ArgentinaOBJECTIVES: Thepriorityofhospitalhealthcareresultsandcostsintransitional countries,needstoavoidnon-standardized,unreliableornon-existentdata.To achievethisgoalwedevelopedahospitaldischargeregistryinArgentina METHODS: AMinimumDischargeDataSet(MDS),designedandextractedfroma1 yearoutput,2007-2008,of3non-prothospitalswithinformationsystems.Cost andresultindicators,harmonizedaccordingtoWHOandHCUPS(USA)terminology,groupers(ClinicalClassicationSoftware-CCSsinglelevel-SL,2009)ofprimary diagnosis(D 1)andprocedures(P 1),(#CCS[descriptiveterm]),totalcosts(CT$) andmedianperdischargecost($,25P-75P-percentiles),ininternationaldollarsPPP, (UNData:1Arg$ 1.608PPP,2008)wereobtained,generatingaParetorankingof10 mostfrequentCCSdischargecodes. RESULTS: 58116dischargeswith17125Dx1 and9163Px1whereobtained.Therst10CCS-SLDx1were:#218[livingrecently born]($229,117-251);#45[chemotherapy,radiotherapy],($4705;1946-8212);#196 [pregnancyandnormaldelivery],($4.829;2.958-5.985);#149[biliarydisorders],($ 5.932;4.700-9.268);#122[Pneumonia],($2.193;1.280-4.856);#101[coronaryatherosclerosis],($22.090;5.586-38.692);#108[congestiveheartfailure,non-hypertensive],($1.918;989-6.156);#195[othercomplicationsofdelivery],($6.940;3.9718.341);#144[unspeciedneoplasia],($12.975;4.511-28.960),y#203[osteoarthritis], ($13.490;8.753-20.305);comprised29,4%ofdischarges,withaCT$$142.852.765 (23,7%).TheprocedurerelatedCCSofdeliveryarecesareansection(#134)andother birthrelated(#137,#135)represented6.95%ofprimaryprocedures(P 1).Themost costlyprocedurewasP 1#($15.834;4.860-32.168).Therst10P constitutesa 17,6%ofdischargesanda23.9%ofprocedureCT$. CONCLUSIONS: Transitional countriescandevelopthisinformation,butharmonizationisrequired.Weobtained30%oftotalcosts(CT$)andperdischargecosts($)withaParetorank.Future studiesneedtoimprovewithincountrygeneralizabilityofresults. PHP38 UTILIZATION,PRICEANDSPENDINGOFANTI-TUMORNECROSISFACTOR BIOLOGICSINTHEUNITEDSTATESMEDICAIDPROGRAMAtzingerC,GuoJJ UniversityofCincinnati,Cincinnati,OH,USAOBJECTIVES: Anti-TNF( tumornecrosisfactor) drugsareextremelyeffectiveforthe treatmentofmanyauto-immunediseasessuchasrheumatoidarthritis,multiple sclerosis,andpsoriasis.ThehighpriceofthesemedicationshasasignicantimpactontheMedicaidbudget.Theobjectivewastodeterminethetrendsinutilization,priceandspendingforanti-TNFbiologicsandtoanalyzethemarketshare competitionbetweenbiologics. METHODS: Aretrospective,descriptiveanalysis wasconductedusingnationalsummarylesfromtheMedicaidStateDrugUtilizationDatafrom1998to2010.Quarterlynumberofprescriptionsandreimbursementdatawerecalculatedforthetrendsanalysis.Studydrugsincludedallbrand andgenericnamesofanti-TNFbiologics(e.g.,abatacept,adalimumab,anakinra, certolizumab,etanercept,golimumab,iniximab,andrituximab).Marketshares forbiologicswerequantiedandcomparedbetween2006and2010.Thequarterly priceperprescriptionwascalculatedbydividingthetotalreimbursementbythe totalnumberofprescriptions. RESULTS: Thedatashowanincreaseinthenumber ofprescriptionsreimbursedovertime.Etanerceptisthemostcommonlyprescribedbiologicinthisclass,followedbyadalimumab.Medicaidhasreimbursed over$1billionforetanercept,and$500millionforadalimumab.Duringtheperiod from2006Q1to2010Q1,etanerceptlost35%ofit'smarketshare.Whileallother compoundsincreasedtheirmarketshare,etanerceptlost40%ofitssalesmarket share.All studybiologicshaveexperiencedanincreasereimbursedpricesince initialmarketentryafteroneyearofentryintomarket.Olderbiologicslike etanerceptandadalimumabincreasedover100%and60%,respectively.Certolizumabwhichhasonlybeenonthemarketfortwoyearshasincreased15%. CONCLUSIONS: Biologicmedicinesrepresentsomeofthemostcomplexbutcostly products.Itwillbecriticaltodevisecreativemethodsandincentivestocontrolthe costsofthesemedicineswithoutstiingtheinnovation. PHP40 HEALTHCAREREFORMSUNDERECONOMICCRISIS:THEGREEKCASEKritikouP,LatsouD,YfantopoulosJ NationalandKapodistrianUniversityofAthens,Athens,GreeceBACKGROUND: SincetheestablishmentoftheGreekNationalHealthCareSystem in1983severalreformshavebeenlegallyintroducedwithoutbeingimplemented. Therecenteconomiccrisisaggravatedtheperformanceofthehealthcaresystem increasingtheinefcienciesandtheinequalities. OBJECTIVES: Theaimofthe studyistoperformasystematicreviewandcriticalappraisalofefciency,quality, effectivenessandequalityoftheGreekhealthcaresystemsincethe1980s. METHODS: Asystematicliteraturereviewwasperformedusingthefollowingdatabases:CenterforReviewsandDissemination(CRD),PubMed,ScienceDirect, EconLitandEconPapers.Thekeywordsusedare"Greece","health"and"health care".Theretrievedarticlessearchedwere338.Ofthese73paperswereincludedin thestudyanalyzingefciency,equality,andeffectivenessoftheGreekhealthcare systemovertheperiod1990-2011.TheevolutionofpublicandprivatehealthexpendituresinconjunctionwithGDPandhealthoutcomeindicatorsisconcerned usingbothdescriptiveandeconometricmodels. RESULTS: TheGreekhealthcare systemisfragmentedwithamixofpublicandprivateactivities.Privatenancing inGreeceisthehighestintheEuropeanUnionwitholigopolistictrendsinprivate hospitalsanddiagnosticcenters.Satisfactionwithpublicservicesisoneofthe lowestintheEU27countries.Informalpaymentsconstituteasignicantshare creatingdistortionsintheefciencyofthepublicsector.Furthermore,informal paymentsareregressiveimposingahighereconomicburdeninthelowerincome classes. CONCLUSIONS: Thesystematicreviewhighlightsthepersistentproblems oftheGreekhealthcaresystemandsigniestheimportanceofhealthcarereforms towardsgreaterefciencyandconvergencewiththerestoftheEUcountries. PHP41 THECHANGINGLANDSCAPEOFORPHANDRUGPROVISIONINTHEEU5AND THEUNITEDSTATESMcConkeyD DoubleHelixConsulting,London,UKOBJECTIVES: Theglobalhealthcareenvironmentisinaconsiderablestateofux withmanynationalhealthsystemsundergoingsomedegreeofreform.Orphan drugshavehistoricallybeenfundedatanationallevelthroughdifferentsystemsto therestofthepharmacopeia.Withthisinmind,theauthorsetouttoassessthe impactofhealthcarereformonhoworphandrugsarelikelytobeprovidedforin thefuture. METHODS: Thisprojectinvolvedastudyoforphandrugprovisioninthe pastandmorerecently,includingbothprimaryandsecondaryresearchelements. Trendsinthechanginghealthcaremarketandspecicreformswereconsideredas partofaprospectivestudyontheorphandruglandscape. RESULTS: Orphandrug provisionusuallyutilisesaseparateinfrastructuretomanagetheallocationof funding,withthissysteminsulatedtoadegreefromgeneralhealthcarereform.In countrieswhererarediseasesaretreatedinspecicinstitutions(e.g.Italy)thereisA18VALUEINHEALTH14(2011)A1–A214

PAGE 27

likelytobelittleornochangeinhowthisismanaged.Conversely,incountries wheretheorphandrugprovisionisintegratedwiththehealthcaremacro-structure,suchasintheUKwiththeformerregionalSpecialCommissioningGroups, thiswilllikelychangealongwithstructuralreformstotheoverallsystem.Finally, inmarketswherehealthcarecanbeconsideredpredominantlyprivatelyfunded (e.g.USA),unlesstherearechangesmadetolegislation,orphandrugprovisionand fundingareunlikelytobesignicantlyaffectedinthefuture. CONCLUSIONS: Despitetheissuesoftighteningbudgets,restrictionsonprescribingandstructural reform,orphandrugprovisionisunlikelytodecreasefromcurrentlevels,butbudgetaryconstraintsmaymakethefuturelandscapemorehostiletodrugsfornew orphanindications. PHP42 EVALUATINGDRUGCOSTANDRESTRICTIONPROCESSESOFCOMMONLYUSED PRESCRIPTIONDRUGSUNDEREACH2011CALIFORNIASTAND-ALONE MEDICAREPARTDPLANPatelR,KimMJ,WalbergM,WoelfelJ UniversityofthePacic,Stockton,CA,USAOBJECTIVES: AlthoughtheMedicarePartDbenethashelpedimproveprescriptionmedicationaccessandlowerout-of-pocketspendingforsome,variabilityin drugcostandaccessrestrictionsofPartDplanscontinuestopresentachallenge forothers.Thepresentresearchsoughttoexaminethereporteddrugcostand restrictions(StepTherapy(ST),QuantityLimits(QL),PriorAuthorization(PA),and FormularyCoverage)associatedwitheachofthe100(65genericand35brand name)mostcommonlylleddrugsbyMedicarebeneciariesundereverystandaloneprescriptiondrugplan(PDP)availableinCaliforniain2011. METHODS: The listofthetop100DrugsbytotalllsunderPartDplansin2008wasretrievedfrom theCentersforMedicare&MedicaidServices.Eachofthesedrugswasenteredinto theMedicarePlanFinderTool(www.medicare.gov).DatafoundviauseofthePlan FinderToolrevealedthefullcost,formularycoverageandrestrictionprocessesof everyexamineddrugundereachofthe2011CaliforniaPDPsforwhichdatawere available. RESULTS: Thedifferencebetweenthelowestandhighestplanreported fullcostofeachgeneric(brand)drugfromallavailablePDPsrangedfromalowof $3.37($5.41)toahighof$134.34($106.33).Inaddition,differenceswerefound betweengenericandbrandmedicationsintermsofdrugrestrictions(3.08%generic vs.57.1%branddrugsrequiredST),(58.46%genericvs.94.3%branddrugsrequired QL)and(4.62%genericvs.31.4%branddrugsrequiredPA)acrossthePDPs.Finally, 13(20%)genericand25(71.4%)brand-namedrugswerenotcoveredononeormore PDPformularies. CONCLUSIONS: Signicantdifferencesinmedicationcostsand planimposedrestrictionprocessesexistbetweenthevarious2011CaliforniaPDPs. SuchdifferencesmaypresentadditionalbarrierstoPartDplanaccessandpotentiallyimpacthealth-relatedoutcomesandcostsofMedicarebeneciaries. PHP43 THEINFLUENCEOFPATIENTACCESSSCHEMESONAPPRAISALDECISIONSBY NICEINTHEUNITEDKINGDOMTimmB,KuselJ,HamerN,CostelloS CostelloMedicalConsultingLtd,Cambridge,UKOBJECTIVES: PatientAccessSchemes(PAS)havebecomeincreasinglyimportantin technologyassessmentbyNICE;apreviousISPORabstractshowedanincreasein PASinclusioninsubmissionsfrom0in2006to29%in2009.However,concerns havebeenraisedregardingtheincreasingcomplexityandthesuitabilityofthese schemescomparedtoothercost-containmentmethods.Weaimedtoevaluatethe inuenceofincludingPASonappraisaldecisionsbyNICE,andconsidertheimpact thatthiscouldhaveonfuturesubmissions. METHODS: AllNICEtechnologyappraisalspublishedin2010andappraisalsindevelopmentwerereviewed.SubmissionsinwhichaPASwasincludedwereidentied;PASweredenedasthemanufacturerprovidingapre-denedreductioninoverallcostoftreatmentthrough risk-sharingorrebateschemes. RESULTS: Ofthetechnologyappraisalspublished byNICEin2010,24%(6/25)includedaPAS.Oftheappraisalsindevelopmentfor whichanalappraisaldeterminationhasbeenpublished,44%(7/16)includeda PAS,eitherintheoriginalsubmissionorintherevisedmanufacturer'smodel. Oncologyproductsrepresented10ofthe13(6published 7indevelopment) submissionsidentied.Ofthese13submissions,3publishedin2010receiveda negativeappraisal,alongwith4oftheappraisalsindevelopment(subjecttoappeal).Inthemajorityofthesecases,evenwhenthePASwasincludedintheeconomicmodeltheproposedICERwasstillabovetheassumedwillingness-to-pay (WTP)thresholdof£30,000perQALY.ExcessiveadministrativeburdenofimplementingthePASwasalsonotedinonerejectedsubmission. CONCLUSIONS: PAS areanincreasinglyimportantelementofNICEsubmissionstoenhancethecosteffectivenessofinnovativedrugs.NegativeappraisaldecisionsmadeonsubmissionsincludingPAShighlighthowmanufacturersmustensurethattheseschemes areeasytoimplementandenhancecost-effectivenesstoaleveldeemedacceptablebyNICE,inordertopositivelyinuenceappraisaldecisions. PHP44 DURATIONOFPATIENTSVISITSTOTHEEMERGENCYDEPARTMENTKaracaZ,WongH,MutterR AgencyforHealthcareResearchandQuality(AHRQ),Rockville,MD,USAOBJECTIVES: Thisstudyexploresthedurationthatpatientsstayintheemergency department(ED)forvisitswheretheyaretreatedandreleased.Durationfortreatand-release(T&R)EDvisitsisassessedbyadmissiondayandhour,patientdemographicsandhospitalcharacteristics. METHODS: Aretrospectivedataanalyses wasconductedtocharacterizethedurationofT&REDvisits.Durationforeachvisit wascomputedbytakingthedifferencebetweenadmissionanddischargetimes. Sensitivityanalyseswereperformedtoassesstherobustnessofresults.The HealthcareCostandUtilizationProject(HCUP)StateEmergencyDepartmentDatabases(SEDD)for2008wereusedintheanalysis.TheSEDDemployedinthisstudy include4.7millionT&REDvisitsinArizona,Massachusetts,andUtah. RESULTS: Durationvariedsignicantlyacrossadmissionhouranddayoftheweek.At95thpercentile,theaveragedurationofT&REDvisitswerelonger(197.8202.6,minutes).Theaveragedurationforpatientsadmittedat8a.m.onMondays,other weekdays,andweekendwererespectively184,189,and172minutes.Similarly,the averagedurationforpatientsadmittedat4p.m.(12a.m.)onMondays,otherweekdays,andweekendwererespectively210,202and179(231,246and234)minutes. Medicarepatientshavethelongestaverageduration(238minutes).Blackpatients have22minuteslongerdurationcomparedtowhitepatients.Therewassignicant variationinaveragedurationacrossdiseasegroups(e.g.,284minutesformental disordersand160minutesforinjuryandpoisoningrelateddiseases).Theaverage durationatteaching(non-teaching)hospitalswas225(166)minutes.Hospitals withlargebedsizewereassociatedwiththelongestdurationofvisits(222minutes) whencomparedtohospitalswithsmallbedsize(172minutes). CONCLUSIONS: ThedurationofT&REDvisitsvariedsignicantlyacrosspatientcharacteristics, hospitalcharacteristics,admissionhour,anddayoftheweek. PHP45 TRENDSOFHEALTHCAREUTILIZATIONAMONGLOYALPATIENTSINA MEDICALCENTERUNDERTAIWAN’SUNIVERSALHEALTHINSURANCE PROGRAMLinHW1,YuIW2,ChangCK2,LinCC1,WuHC2,HungJS2 1ChinaMedicalUniversity,Taichung,Taiwan,2ChinaMedicalUniversityHospital,Taichung, TaiwanOBJECTIVES: Thosepatientsvisitedtowardasinglehospitalloyallyhaveaccountedfor3.5%ofallbeneciariesandconsumed19.3%ofcoveredmedicalexpenditureunderTaiwan'sNationalHealthInsurance(NHI)program.Thisstudy wastoexaminethetrendsofserviceutilization,medicalexpenditureanddisease patternsamongloyalpatientsvisitedtowardChinaMedicalUniversityHospital (CMUH),a2000-bedmedicalcenterinTaiwan. METHODS: Patients'visitsmade towardCMUHaccountedformorethan50%oftheirtotaloutpatientvisitsinTaiwanduringJanuary2008toJune2009(retrospectiveperiod)wereassignedas CMUHloyalpatientsbyBureauofNHI.Thecorrespondingdatadocumentedin CMUHfromJanuary2008toApril2010wereretrievedandanalyzed.ThoseassignedloyalpatientspersistentlyvisitedCMUHduringJulytoSeptember2009were groupedintofourlevelsofutilizationbaseupontheirnumbersofoutpatientdepartment(OPD)visitsmadetowardCMUHduringretrospectiveperiod.Thecomparisonsofdiseasestatuses,healthcareutilization,medicalexpenditureandvisits towarddifferentspecialtiesamongdifferentutilizationgroupsandacrossdifferent studyperiodsinCMUHweremade. RESULTS: 11,846CMUHloyalpatientswere groupsintofourlevelsofOPDutilization( 20,21 23,24 35, 36)during18 months(meanOPDvisit 1.83/person/month).Thoseloyalpatientsusedmore OPDvisitswereelder,moreco-morbidities,takencarebymorephysiciansand medicalspecialties,andprescribedmoremedications.ThemorevisitsmadetowardOPD,thehigherlikelihoodtoconsumemoremedicalexpendituresandutilize moremedicalresourcesinCMUH.Althoughmorethan75%ofOPDvisitswere madetowardoneortwomedicaldepartments,moremedicalresourcesinCMUH wereattributabletothosepatientsvisitedmorethanthreemedicalspecialties. CONCLUSIONS: Resultssuggestin-houseprogramstofacilitateintegratedcareand medicationreconciliationtowardhighutilizationloyalpatientsaremoredemandedthanotherloyalpatientsinCMUH. PHP46 THESTATEOFHEALTHECONOMICEVALUATIONINSOUTHAFRICA: ASYSTEMATICREVIEWGavazaP1,RascatiKL2,OladapoAO2,KhozaS2 1AppalachianCollegeofPharmacy,Oakwood,VA,USA,2UniversityofTexasatAustin,Austin, TX,USAOBJECTIVES: Thestudyassessedthestateofhealtheconomics(includingpharmacoeconomics)researchinSouthAfrica. METHODS: AliteraturesearchwasconductedtoidentifyhealtheconomicsarticlespertainingtoSouthAfrica.Tworeviewersindependentlyscoredeacharticleinthenalsampleusingthedata collectionformdesignedforthestudy. RESULTS: Intotal,108studiesinvestigating awidevarietyofdiseaseswereincludedinthestudy.Thesearticleswerepublished in39differentjournalsmostlybasedoutsideofSouthAfricabetween1977and 2010.Onaverage,eacharticlewaswrittenbyfourauthors.Mostrstauthorshad medical/clinicaltrainingandresidedinSouthAfricaatthetimeofpublicationof thestudy.Basedo na1to10 scale,with10indicatingthehighestquality,themean qualityscoreforallstudieswas7.59(SD 1.42)andhalfofthearticleswereofgood quality(score8-10)Thequalityofstudieswasrelatedtothecountryinwhichthe journalpublishingthearticlewasbased(outsideSouthAfrica higher),current residenceoftheprimaryauthor(outsideSouthAfrica higher),methodofeconomicanalysis(economicevaluationshigherthancoststudies),typeofdataused (secondaryhigherthanprimary),primarytrainingoftherstauthor(healtheconomics/pharmacoeconomics higher),typeofmedicalfunction(diagnosis higher),studyperspective(society higher),primaryhealthintervention (pharmaceuticals higher),studydesign(modeling higher),numberofauthors (more higher),andyearofpublication(morerecent higher)(p 0.05). CONCLUSIONS: Halfofthearticleswereofpoororfairquality.Measuresare neededtopromotethecommissioningofmoreandbetterqualityhealtheconomicsandpharmacoeconomicsstudiesinSouthAfrica.A19VALUEINHEALTH14(2011)A1–A214

PAGE 28

PHP47 ADVERSEDRUGEVENTSANDELECTRONICMEDICALRECORDS:RESULTSFROM THE2007-2008NATIONALAMBULATORYMEDICALCARESURVEYHoggJM,HeatonPC UniversityofCincinnati,Cincinnati,OH,USAOBJECTIVES: Thegoalofthisstudywastoevaluatetheimpactofelectronicmedical record(EMR)useonthefrequencyofvisitsduetoanadverseeventasreportedinthe NationalAmbulatoryMedicalCareSurvey(NAMCS)2007-2008database.Thespecic objectivewastodeterminethefrequencyofadverseeventsinpatientvisitsthatincludedtheuseofanEMRcomparedtovisitswithouttheuseofanEMR.Anadverse eventwasdenedasanadverseoutcomefrommedicalorsurgicalcareoranadverse drugevent. METHODS: Thestudydesignwasaretrospective,cross-sectional,observationalanalysisoftheNAMCSdatabasefrom2007-2008.Patientvisitsthatwerethe resultofanadverseeventwereselectedfromallestablishedpatientvisits.ThefrequencyofadverseeventswasanalyzedwithrespecttotheuseofanEMR.Patientvisits wereweightedwithastatisticalmultipliertogeneratenationalestimates. RESULTS: Anadverseeventoccurredin44,035,493(2.6%)patientvisits.Ofthesevisits,26,067,600 (1.5%)didnotincludetheuseofanEMR,while17,967,893(1.1%)includeduseofan EMR.ThemajorityofvisitsincludingtheuseofanEMRwerebyfemalepatients (59.4%),whitepatients(83.2%),orpatientsage45-64years(29.2%).Only40.2%ofestablishedpatientvisitsincludedtheuseofanEMR. CONCLUSIONS: Adverseevent frequencywaslowerinpatientvisitsthatutilizedanEMRascomparedtopatientvisits thatdidnotincludetheuseofanEMR.IncreasingEMRusewillallowhealthcare professionalstofurtherpreventadverseevents. PHP48 STATUSOFPHARMACISTATCOMMUNITYLEVEL:ANEXPLORATORYSTUDY INQUETTA,PAKISTANSaleemF1,Al-QazazHK1,AzharS1,AhmadN1,AtifM1,HaqN1,AsifM2,IqbalQ3 1UniversitiSainsMalaysia(USM),Pinang,Penang,Malaysia,2TheIslamiaUniversityof Bahawalpur,Bahawalpur,Punjab,Pakistan,3UniversityofBalochistan,Quetta,Balochistan, PakistanOBJECTIVES: TostudythestatusofpharmacistsworkingatcommunitypharmaciesinQuettacity,Pakistan. METHODS: Across-sectionalsurveywasdesignedto conductthisstudy.Aquestionnaireaboutstatusofpharmacistsworkingatcommunitypharmacieswasconstructed,contentvalidatedandused.Pharmacists werehiredfordatacollectionandanalysesweredonebySPSS15. RESULTS: All415 communitypharmaciesofthecitywereapproachedand392responded(94.45%). Majorrespondents(81.25%)didnotpossessthebasicrequirementofPharmacy degreealthoughthecommunitypharmacieswerelicensedunderaqualiedperson.Therewereonly18.75%pharmacieswereaqualiedpharmacistpossessing pharmacydegreewaspresent.Peoplewithhighersecondaryeducation(31.25%) werefoundmostlyinvolvedinrunningcommunitypharmacies.Theknowledgeof medicinesandgenericswasquitehigh(74.95%).Drugguides(45.25%)andinformationgivenbymedicalsalesrepresentatives(30.25%)playedakeyroleinthegenerationofthisinformation.Itwasalsoreportedby65.50%thatasbusinessisless protable,pharmaciescannotaffordafulltimepharmacistwhereas41.50%describedunemployment/lackofopportunitiesinothersectors,thereasonofrunningacommunitypharmacy.61.25%oftherespondentshadanexperienceofmore than5yearsworkingatcommunitypharmacies.Allpharmaciessoldpackedmedicationsanddispensingwasnotdoneonanyofthecommunitypharmacy. CONCLUSIONS: Pharmacistatthecommunitylevelisimportanttominimizedrug useproblems.Theremustbestrictmonitoringbytheconcerndepartmentsand licenseissuingauthoritiesinordertopromotepropermedicationpracticesatcommunitylevels.Inaddition,thepresenceofPharmacistatpharmaciesshallbedeclaredmandatory. PHP49 UTILIZATIONANDCOSTOFTHEHEALTHPLAN-EFFECTSOFSPOUSAL BEREAVEMENTReisNetoJ1,ArantesACL2 1CAPESESP/CAPESAUDE,RiodeJaneiro,RJ,Brazil,2CASADOCUIDAR,SaoPaulo,SP,BrazilOBJECTIVES: Toevaluatetheeffectsofwidowhoodontheoutpatientandhospital healthservicesusedbybeneciariesofaprivatehealthcareplaninBrazil. METHODS: Cohortstudyofbeneciarieslistedasmarriedofthehealthplanin thedatabaseofJanuary2007,consistingof29,932couples.Inafollow-upperiod of36monthswasveriedthedateofdeathofaspouse,thisbeingthedateof entryintothestudy.Afterexclusionofonecase,487hadrecordsofuseandcost ofhealthcareexaminedduring12monthsbeforeandafterthedeath.The covariablesinthestudywerethetypeofprocedure(consultations,testsand admissions),sexandage(upto59years,60-69,70-79,80ormoreyears).The statisticalanalysisusedtheStudentttestforpairedobservationsandacondenceintervalof95%.Thesignicancelevelwassetatp 0.05. RESULTS: Betweenthetwoperiods,theaveragenumberofconsultationsincreasedfrom3.77to 4.66(p 0.001),testsfrom16.88to23.16(p 0.001)andadmissionsfrom0.18to0.30 (p 0.007).Regardingmedicalandhospitalexpenditures(directcosts),thisincreasedfromU$1,246.69toU$1,941.47(p 0.001).Inthe segmentedanalysisby agegroups,variationsinquantitiesandcostsofvisitsandtestswerestatisticallysignicant(p 0.05),differentlyofadmissions(p 0.05).Asfortheevaluationbygender,onlythevariationsofthemeannumberofadmissionsandtotal costwerenotsignicant(p 0.05). CONCLUSIONS: Thisstudydemonstratedthe increaseduseofhealthservicesanddirectcostsofmedicalandhospitalappointments,testsandadmissionsafterthelossofaspouse.Themourningshouldnotbe overlookedfromtheperspectiveofmanagement,maybeapredictorofmorbidity anduseofhealthplan. PHP50 MIXEDEVIDENCEFORTHE”HEALTHYADHEREREFFECT”INASAMPLEOF 22,070ADULTSWITHCHRONICDISEASEINTHEUNITEDSTATESMcHorneyC,GadkariA Merck&Co.,Inc,NorthWales,PA,USAOBJECTIVES: The"healthyadherereffect"positsthatpartoftheassociationbetweennon-adherenceandsuboptimalhealthoutcomesisthepossibleconfoundingeffectofworselifestylebehaviorsofnon-adherers.Ourobjectivewastotestthe "healthyadherereffect"amongUSadultswithchronicdisease. METHODS: We conductedacross-sectionalsurveyof22,070adultswithasthma,hypertension, diabetes,hyperlipidemia,osteoporosis,GERD,depression,oranxietyfromtheHarrisChronicDiseasePanel.Respondentswereclassiedintooneofthreemutuallyexclusivecategories:(1)persisters(i.e.,currentlytakingmedication);(2)non-persisters(i.e.,discontinuedmedicationinthepastyear);and(3)non-fulllers(i.e.,did notobtaintherstllforanewprescriptioninthepastyear).Respondentsalso completeditemsonlifestylebehaviors:weightandheighttoderiveBMI,smoking status;alcoholuse;physicalexercise;seatbeltuse;andreceiptofaninuenza vaccinationinthepastyear.Linearandlogisticregressionwasusedtomodel individuallifestylebehaviorsonself-reportedadherencestatuscontrollingfordemographiccharacteristics. RESULTS: Multivariateregressionsrevealednostatistically-signicantassociationsbetweenadherencestatusandBMI(continuous), weekly-moderateexercise,weekly-vigorousexercise,bingedrinking,andobesity asdenedbytheBMI.Multivariateregressionsrevealedthatnon-persistersand non-fulllersweresignicantlylesslikelytoreporthavingaushotinthepast year(OR 0.76and0.73,respectively).Comparedtopersisters,non-persisterswere morelikelytobecurrentsmokers(OR 1.19,p .0043)andwerelesslikelytoalways useseatbelts(OR 0.80,p .0009). CONCLUSIONS: InthisInternetsampleof22,070 adultswithchronicdisease,wefoundmixedevidencefora"healthyadherereffect."Non-persistersandnon-fulllersdifferedfrompersistersonsome,butnot all,lifestylebehaviors.Futureresearchshouldlinkbothadherencebehaviorsand life-stylebehaviorstoriskofmortalityandhospitalization. PHP51 MORTALITYAFTERSPOUSALBEREAVEMENTINBENEFICIARIESOFTHE HEALTHPLANReisNetoJ,ArantesACL CAPESESP/CAPESAUDE,RiodeJaneiro,RJ,BrazilOBJECTIVES: Toevaluatethemortalityrateinbeneciariesofahealthplanin Brazilafterspousedeath,comparedwiththeexpectedinthepopulation. METHODS: Cohortstudyofbeneciariesofthehealthplanlistedasmarriedin 2007,consistingof29,932couples.Thereafterbeganafollow-upof36monthsfor vericationofdeathofeitherspouse.Duringthisperiod,enteredthestudyatthe dateofdeathofspouse,308widowsand180widows.Themortalityrateobserved inthegroupwascomparedwithexpectedmortalityforthepopulationofreference, accordingtosexandage.StatisticalanalysiswasconsideredtheStandardMortalityRatio(SMR),withevaluationofsignicancebythemethodofPoissonapproximationwithByarandtheFisherexacttest,withancondenceintervalof95%.The signicancelevelwassetatp 0.05. RESULTS: Themeanageformaleswas65.24 years(95%CI51.03to79.45)and63.91yearsforfemales(95%CI53.15to80.67). Observedinperiod35deathsinthegroup(7.17%).Intherst12monthsafter bereavement,mortalitywas12foranexpected6.15,SMRto1.95(p 0.0477).The SMRin24monthswas2.02(p 0.0026)andat36monthsof1.85(p 0.0018). CONCLUSIONS: Thisstudyshowedhigherthanexpectedmortalitytosignicant levelsinpeoplewhohaveexperiencedthedeathofaspouse.Measurestosupport andhealthcareforfamilieswholivesgriefcanbeaninterestingalternativeinthe socialaspectandsomehowcontributetoreducingtheriskofmortalityatthe expectedlevelforsexandage. PHP52 INCREASEINUNINTENTIONALFATALPOISONINGSBYNOXIOUSSUBSTANCES INKANSASCITY,MISSOURI,FROM1999TO2008LiuY1,CaiJ2,HongL3,HoffGL2,OkahFA1 1TheUniversityofMissouriKansasCity,KansasCity,MO,USA,2KansasCityHealth Department,KansasCity,MO,USA,3TheUniversityofTennessee,Memphis,TN,USAOBJECTIVES: 1)Todescribethetrendsofunintentionalfatalpoisoningsbynoxious substancesinKansasCity,Missouri,from1999to2008and2)toidentifydemographicriskfactorsassociatedwithunintentionalfatalpoisonings. METHODS: ThiswasaretrospectivestudyusingthedeathdataforKansasCity,Missouri.The dataincludedthecauseofdeath,sex,race,age,annualincome,andHealthZone foreachindividual.Unintentionalfatalpoisoningsbynoxioussubstanceswere identiedbyICD-10codesX40-X49.Age-adjusteddeathratesandage-specic deathratesoffatalpoisoningswerecalculatedforeveryyear.Chi-squaretests wereusedtocomparedeathratesforeachdemographicvariable.Amultiplelogisticregressionanalysiswasconductedtomeasuretheoddsratio(OR)offatalpoisoningsduetosex,age,annualincome,andHealthZone. RESULTS: Outof38,785 deathsfromallcauses,473(1.2%)wereidentiedtobeduetounintentionalfatal poisonings.Age-adjusteddeathratesfromunintentionalfatalpoisoningsincreasedfrom7per100,000populationin1999to14in2008,whiletheall-cause age-adjusteddeathratesdeclinedfrom1999to2008.From1999to2003,theage group35-44yearsoldhadthehighestdeathrateofunintentionalfatalpoisonings, whereasfromyear2004to2008,theagegroup45-54yearsoldhadthehighest deathrate.Inthelogisticregression,comparedwiththeagegroupof55yearsand older,theagegroupof35-44yearsoldhadanORof44.06,andtheagegroupof45-54 yearsoldhadanORof21.49.ThereweredifferencesinORsbysex,annualincome,A20VALUEINHEALTH14(2011)A1–A214

PAGE 29

andHealthZone. CONCLUSIONS: Age-adjusteddeathratesofunintentionalfatal poisoningsincreasedforKansasCity,Missouri,from1999to2008.Sex,age,annual income,andHealthZone,wereriskfactorsforunintentionalfatalpoisonings. PHP53 ARAPIDEVALUATIONUNITINCREASESEMERGENCYDEPARTMENTVISITSTwanmohJ1,FlemingAE2,BrooksD3,JiangY4,VillacortaR4,McCombsJ4 1St.AgnesHospital,Baltimore,MD,USA,2JohnsHopkinsUniversity,Baltimore,MD,USA,3Johns HopkinsMedicine,Baltimore,MD,USA,4USCSchoolofPharmacy,LosAngeles,CA,USAOBJECTIVES: EmergencyDepartment[ED]overcrowdingisanationalproblem. Thisstudyevaluatestheeffectofaprovider/nursetriagemodel(RapidEvaluation Unit,REU)designedtoimprovetheefciencyoftheEDonEDpatientvolumeand ambulancediversionhours. METHODS: Dataondailypatientcounts[inpatient,ED] andambulancediversionhoursforasinglehospitalinBaltimorewereusedto evaluatetheimpactofREUonhospitalperformance.Datacovered2008,theyear beforetheREUwasimplemented,and2009duringwhichtheREUwasimplementedincrementally.TheREUwasoperatingonMonday-WednesdayfromJanuary1stthroughMarch31st,andthenMondaythroughFridayfromApril1stthroughDecember31st.Regressionmodelswereestimatedforthedailycountsfor admissions,EDvisitsandambulancediversiondayscontrollingfortimetrend effects,dayoftheweekandmonthoftheyear.Asplinefunctionalformwasused totesttheeffectsoftheREUonpre-2009timetrendsandtoimprovemodelt. Modelswereadjustedforautocorrelationandheteroskedasticityasneeded. RESULTS: TheREUsignicantlyincreasedEDpatientvisitcountsby5.5visitsper day[p 0.01]withoutincreasingthenumberofhoursthehospitalwasdiverting ambulancesduetoovercrowding[ 0.83hoursperday,p 0.05].ForEDvisits,all monthsexhibitedsignicantlyhigherdailyEDvisitcountsthaninDecember,rangingfrom 9to 28visitsperday.Theearlyworkweek[Monday-Wednesday]also exhibitedsignicantlyhigherratesofEDvisitsthanSaturday[range:5-12visitsper day]. CONCLUSIONS: TheREUwaseffectiveatincreasingEDvisitswithoutincreasingambulancediversionhours.Wetheorizethatthismaybeduetoareductionin thetimespentbypatientswaitingfortreatment. PHP54 THEIMPACTOFPHARMACIST-CONDUCTEDMEDICATIONRECONCILIATION WITHPATIENTCOUNSELINGATHOSPITALADMISSIONChenCC1,HuangYF1,HoSW2,LuSH2,LinHY2,LiuHP1 1TaipeiMedicalUniversity-ShuangHoHospital,Taipei,Taiwan,2TaipeiMedicalUniversity, Taipei,TaiwanOBJECTIVES: Medicationreconciliationistheprocessofobtainingandmaintaining acompleteandaccuratelistofthecurrentmedicationtherapyofapatientacross healthcaresetting.Thecombinationofrecordedandreported(assessedthrough patientcounseling)medicationusemayincreasetheaccuracyofmedicationreconciliationprocess.Theobjectiveofthisstudyistopreventdrug-relatedproblems andevaluatetheeffectofpharmacist-conductedmedicationreconciliationwith patientcounselingathospitaladmission. METHODS: Patientswhowereadmitted toInternalMedicineinlocalhospital,withageover65years,morethantwocomorbiditiesandreceivingmorethanthreeregularmedicationswererecruited. Patientswithcancer,admittedtotheICU,dischargedwithin48hours,orunableto communicatewereexcluded.Then,thesepatientswereinterviewedbypharmacistswithin48hoursafteradmission.Theunintentionaldiscrepancieswerediscussedwiththephysiciansanddrugtherapywouldbeadjustedaccordingly.The primaryoutcomeofthisstudyincludedtherateofunintentionaldiscrepancies identiedbypharmacistsbeforeandafterpatientcounselingandtherateofunreconciledmedicationsresolvedbypharmacists. RESULTS: Forty-ninepatients wererecruitedinthisstudyand129unreconciledmedicationswerefoundbetween October17andNovember18,2010.Themostcommontypeofdiscrepancieswas omissionswithrateof58.7%.Ontheotherhand,therateofunintentionaldiscrepanciesidentiedbypharmacistwhichmaycauseharmincreasedfrom6.6%to 12.4%afterpatientscounselingand81.3%ofdrug-relatedproblemwasresolved afterpharmacistintervention.Additionally,allergyhistorywasestablishedin 16.3%ofpatients. CONCLUSIONS: Pharmacist-conductedmedicationreconciliationwithpatientcounselingimproveddrugtherapyandcompletenessandaccuracyofallergyhistory.However,thisstudywasonlyundertakeninpatientsadmittedtoInternalMedicineduringonemonth.Implementationofmedication reconciliationwithpatientcounselingtofurtherpatients,suchassurgicalpatients maybeconsidered. PHP55 THEINFLUENCEOFCOST-EFFECTIVENESSISSUESOFTHEAPPRAISEDORPHAN DRUGSONRECOMMENDATIONSOFAGENCYFORHEALTHTECHNOLOGY ASSESSMENTINPOLAND(AHTAPOL)Oerska-SujkowskaG,Jagodzinska-KalinowskaK,MatusewiczW AgencyforHealthTechnologyAssessmentinPoland,Warsaw,PolandOBJECTIVES: Theaimofthepresentanalysiswastoidentifywhatwastheinuenceofpricesandthecorrespondinglowcost-effectivenessestimatesoftheappraisedorphandrugsonAHTAPolrecommendations.ThemaintaskofAHTAPol, establishedin2005,istoassessandappraiseallmedicaltechnologiesandservices claimingpublicmoneyfounding.Pharmacoeconomicevaluationsofnewtherapies arerequiredforallreimbursementdecisionsandorphandrugmanufacturerscannotbeexemptedfromprovidingafullpharmacoeconomicorHTAreports. RecommendationsissuedbyAHTAPolhavebeenbasedonManufacturerssubmission, additionalofciallypublisheddata,expertsopinionsandPolishpublicpayer (NationalHealthFund)evaluation. METHODS: Allrecommendationsissuedbythe AHTAPoluntiltheendof2010werereviewedandanalyzedfromtheofcialWebsite ofAHTAPol.Theorphandrugsrecommendationswereidentiedandcategorizedinto typesofrecommendations(positiveornegative)andreasoned. RESULTS: Among286 AHTAPoldecisionsstudied,26(9%)appliedtoorphandrugs.AHTAPolgavepositive recommendationsforreimbursementto19of26oforphandrugsubmissions(73%).7 (27%)werenotapprovedandreceivednegativerecommendations.Onlyinonecase cost-effectivenessestimatesandthecorrespondinghighpriceswereemphasizedasa mainreasonofnegativerecommendations.In1casethereasonwasconnected withdrugsafetyissues,whilein5caseswithinsufcientevidenceofefcacy. CONCLUSIONS: Thecost-effectivenessissuesoftheappraisedorphandrugswerenot importantargumentinnegativerecommendationsoftheAHTAPol.Infacttheinsufcientevidenceofclinicalefcacywerethemostimportantandthereforeprevailing argumentforissuingnegativeopinionsofAHTAPolrecommendations. PHP56 DEMONSTRATINGCLINICAL-EFFECTIVENESSUSINGINDIRECTANDMIXED TREATMENTCOMPARISONANALYSIS:AREVIEWOFMANUFACTURERS’ SINGLETECHNOLOGYAPPRAISAL(STA)SUBMISSIONSTOTHENATIONAL INSTITUTEFORHEALTHANDCLINICALEXCELLENCE(NICE)BendingMW1,HuttonJ1,McGrathC2 1UniversityofYork,YorkHealthEconomicsConsortium,York,UK,2Pzer,Surrey,UKOBJECTIVES: IndirectandMixedTreatmentComparisons(ITC/MTC)canprovide valuableinformationfordecision-makers,especiallywhendirectcomparisonsbetweenmedicinesareunavailable.Theobjectiveofthisstudyistoinvestigatethe methodsandimpactofITCandMTCsubmittedbymanufacturersontheNICE committee'sappraisalofpharmaceuticals. METHODS: AsearchoftheNICEwebsitewasconductedformanufacturers'submissionsthatreportedtocontaineither anITCorMTCsince2006.Datawereextractedandanalysedforthetypeofnetwork meta-analysis,numberoftrials,head-to-headtrialevidenceavailable,disease area,treatmentcomparisons,studyselectionjustication,sensitivityanalysisof trialselectionandoutcome(s).Theimpactwasassessedbyanalysinginformation onEvidenceReviewGroup(ERG)reviewofevidencesynthesis,Appraisalcommittee'scommentsandnaldecision. RESULTS: Thesearchidentied24submissions thatincludedeitheraMTCorITC.TheissuesmostfrequentlyraisedbytheERG werelackofreportingoftrialcharacteristics,descriptionofmethodsandquality assessmentofalltrialsincluded.Themostfrequentvalidityconcernswereboth clinicalandstatisticalheterogeneitybetweentrials,inappropriatemethodsof analysis,exclusionofrelevanttrialsandsparsenumbersoftrialsinthenetwork. Thecommitteeconsideredallevidenceandreportedtheanalysistobeplausibleon oneoccasion,statedconcernswiththevalidityin18appraisalsanddidnotdirectly commenton5occasions.FiftyeightpercentofappraisalsincludinganMTCorITC resultedinarestricteddecision. CONCLUSIONS: ITCandMTCanalysishasprovidedadditionalusefulinformationforNICEappraisalsbuttherehasbeenwide variationinthereportingandvalidityofanalysisperformed.Reimbursement agenciesshouldestablishguidelinesfortheconductofITCandMTCtoreducethis variation.TheemergingISPORGoodResearchPracticeguideonindirecttreatment comparisonswillprovideabasisforthese. PHP57 PATTERNSOFINSURANCECOVERAGEINTHEUNITEDSTATES:ANALYSISOF THE2004-2007MEDICALEXPENDITUREPANELSURVEYS(MEPS)BerryE1,KeltonCM1,HeatonPC1,GuoJJ2,FairbrotherG3,HansemanD2 1UniversityofCincinnati,Cincinnati,OH,USA,2UniversityofCincinnatiMedicalCenter, Cincinnati,OH,USA,3CincinnatiChildren'sHospitalMedicalCenter,Cincinnati,OH,USAOBJECTIVES: Theobjectivewastounderstandtheextentoftheinsurance-loss problemintheUnitedStatesoverafour-yearperiod(2004-2007)by(1)determining thenumber(andproportion)ofindividualswhoexperiencedlossininsurance coverageand(2)forthosewhoexperiencedgapsincoverage,determiningthe averagelengthofthoseperiodsoftimewhentheywerewithoutcoverage. METHODS: Thisstudywasaretrospective,longitudinaldataanalysisusing monthlydatafromthe2004-05,2005-06,and2006-07panelsoftheMedicalExpenditurePanelSurvey(MEPS).Individualswerecategorizedonthebasisofcontinuity ofinsurancecoverage:acohortofcontinuouslyinsuredindividuals;acohortof continuouslyuninsuredpeople;andacohortofpeoplewhoexperiencedgaps (fromoneto23months)incoverage.Thenumberandpercentofindividualswho wereineachgroupforeachpanelwerecalculated.Achi-squaretestwasusedto determineifthereweredifferencesinthedistributionofhealthinsurancepatterns betweenthepanels,andANOVAwasusedtodetermineifthereweredifferences betweentheaveragegaplengthovertheyears. RESULTS: Poolingacrossthethree panels,therewere157,641,235(17.37%)individualswhoexperiencedgapsininsurancecoverage,82,826,100(9.12%)whowerecontinuouslyuninsuredfor24months, and616,864,391(67.95%)whohadcontinuouscoverageoveratwo-yearperiod.The proportionofcontinuouslyuninsuredrosefrom8.73%in2004-05to9.23%in200607.Theaveragegaplengthwas8.56,8.44,and8.54monthsin2004-05,2005-06,and 2006-07respectively.Therewerenosignicantdifferencesacrossthepanelsin eitherthedistributionofinsurancecoveragepatternsorbetweentheaveragegap lengths. CONCLUSIONS: Patternsinhealthinsurancecoveragehaveremainedstableoverthepastseveralyears.Examinationofchangingpatternsofinsurance coveragewillbeimportantashealthreformisintroducedbetweennowand2014. PHP59 THEAVAILABILITYANDSIGNIFICANCEOFCOST-EFFECTIVENESSDATAFOR INFORMINGDECISIONSTOADDNEWDRUGSTOTHENATIONALLISTOF HEALTHSERVICESINISRAELGreenbergD1,WachtO1,SiebzehnerMI2 1Ben-GurionUniversityoftheNegev,Beer-Sheva,Israel,2TheIsraeliCenterforTechnology AssessmentinHealthCare,TelHashomer,IsraelA21VALUEINHEALTH14(2011)A1–A214

PAGE 30

OBJECTIVES: TheNationalListofHealthServices(NLHS)inIsraelhasbeenupdated annuallysince1999butresultsfromeconomicevaluations(EE)werenotusedto supportcoveragedecisions.WeexploredthepotentialavailabilityofEEresultsto thecommitteeresponsibleforupdatingtheNLHSatthetimesofcoveragedecisionsandwhetheravailabilityanduseofthesedatacouldhavealteredthese decisions. METHODS: WeusedtheTuftsMedicalCenterCost-EffectivenessAnalysisRegistry(http://www.cearegistry.org)tosearchforrelevantcost/QALYEEfor alldrugsandtheirrelevantindicationsaddedtotheNLHSfrom1999through2008. Foreachpairofdrugandcost/QALYpublicationwerecordedthepublicationdate, theintervention(s)andcomparator(s)consideredandtheincrementalcost-effectivenessratio(ICER)todeterminevalueformoney.BasedonavailableICERswe qualitativelyclassiedeachcoveragedecisionintooneofthreecategories:1)The coveragedecisioncanbejustiedonEEgrounds(EEsuggestthedrugiseither dominant/cost-savingorprovidesgoodvalueformoney);2)Thecoveragedecision cannotbejustiedonEEgrounds;3)TheevidencefromEEismixedandwecould notdeterminewhetherthecoveragedecisioncanbejustiedornot. RESULTS: Relevantcost/QALYanalyseswerefoundfor181(40%)of451drugsincludedinthe updatesoftheNLHSofwhichonly71(16%)ofdrugshadrelevantEEpriortothe coveragedecision.BasedontheevidencegatheredfromEEpriortoandfollowing thecoveragedecision,wesuggestthatdecisionswerecorrectin56%ofthecases, incorrectin17%andambiguousin27%. CONCLUSIONS: TheuseofEEtosupport coveragedecisionscouldhavealtereddecisionsinasizableproportionofdrugs addedtotheNLHSinIsrael.AvoidingtheuseofresultsfromEEtosupportpublic fundingofdrugsmayleadtoanon-optimaluseofscarcehealthcareresources. PHP60 ORPHANDRUGSFACETOUGHERSCRUTINYINSECURINGFAVORABLEPRICING ANDMARKETACCESSGrosvenorA,SarafS,JonesK PriceSpectiveLtd.,London,UKOBJECTIVES: Orphandrug(OD)legislationhasbeenhighlysuccessfulinincentivisingpharmaceuticalcompaniestoinvestindevelopingmedicinesforpreviouslyignoredrarediseases.Since2005,athirdofallnewdrugapprovalsintheUShave beenODs,andworldwidethemarketisforecasttogrowataCAGRof6%,reaching $112bnin2014.Althoughtheseproductsstillonlyaccountfor2-3%oftotaldrug budgetsintheUSandEU,theirburgeoningnumberisplacingincreasedpressure onfunding.ThisposteraimstoexplorehowpayersintheUSandEUarerespondingtothesenancialdemands. METHODS: WereviewedallODsapprovedinthe USandEUsince1stJanuary2000and,foreach,calculatedtheaveragecostper patientperyear,basedonpublicpricesintheUSand5EU.Wealsoreviewed publishedpayerassessmentsoftheseproducts,andallgovernmentproposals regardingODpoliciessince2005. RESULTS: Since2007,therehasbeenanobvious dropinpricessecuredfornovelODs.IntheEUalone,theaveragecostperpatient peryearforproductsapprovedin2008to2010( 34,890)is73%lowerthanforthose approvedin2000to2007( 129,228).Furthermore,ouranalysissuggestspayersare adoptingamorediscerningapproachtothewaytheyevaluateorphandrugs,especiallythoseperceivedtobeexploitingtheoriginalintentofthelegislation.For example,Germany'srecenthealthcarereformshighlightedplanstotargetODs thatfailtodemonstratemeaningfulpatientbenets,suchassurvival,andallODs exceedingannualsalesof 50mwillhavetosubmitacost-benetanalysis. CONCLUSIONS: Payersareclampingdownon"opportunistic"ODsandonlythose witharobustbodyofevidencesupportingboththeclinicalandeconomicargumentsfortheirusewillsecurefavourablepricingandmarketaccess. PHP61 WIDERCONSULTATIONINHEALTHTECHNOLOGYASSESSMENT(HTA) DECISIONS:BETTERUNDERSTANDINGORALOBBYINGOPPORTUNITY?BendingMW1,HuttonJ1,McGrathC2 1UniversityofYork,YorkHealthEconomicsConsortium,York,UK,2Pzer,Surrey,UKOBJECTIVES: HTAagenciesworldwidehavevaryingprocessesthatallowconsultationwithstakeholdersduringdecision-making.Theobjectiveofthisstudyisto determinetheimpactoftheNationalInstituteforHealthandClinicalExcellence (NICE)SingleTechnologyAppraisal(STA)consultationstageonreimbursement decisionsofpharmaceuticals. METHODS: Documentationwasaccessedfromthe NICEwebsiteforallSTA'sconductedbetween2006andAugust2010.Detailsofthe rstAppraisalConsultationDocument(ACD)draftdecision,subsequentACDs, FinalAppraisalDetermination(FAD)andnalguidancedecisionwereextracted. Thedecisionswerecategorisedwithrespecttothelicensedindication(recommended,restricted,notrecommended,onlyinresearch).Detailsofthefurther analysisandevidencesubmittedbythemanufacturerasaresultofconsultation wereextracted.Thesedatawereanalysedforthedifferentstagesof decision-making. RESULTS: Thewebsitesearchidentied55NICEappraisalsof whichoverftypercentwereforcancermedicines.Finaldecisions(draftrst provisionaldecision)included36%(13%)recommended,36%(20%)restricteddecision,16%(56%)notrecommendeddecisionand11%(11%)terminateddecision.One appraisalcontainedonlyinresearchrecommendationsinadditionforuseinroutinepractice.AnACDwasproducedin42appraisals,followedbythemanufacturer providingfurthereconomicanalysisin26appraisals,apatientaccessschemein5 appraisalsandnewclinicalevidencein2appraisals.Typesoffurthereconomic analysisprovidedbythemanufacturerwereforothertreatments/strategies;differentmodellingassumptions;alternativesurvivaldistributions;furthersensitivityanalysis;andother. CONCLUSIONS: NICE'siterativeconsultationprocessallowsconsiderationofevidenceandwideconsultationwithstakeholders.This resultsinevidencethatismoreappropriatefortheevaluationofpharmaceutical's andpartlyexplainsthehigherrecommendationratewhencomparedwithsimilar internationalreimbursementagencies.Thereisaneedforfurtherresearchtounderstandtheimpactofthedifferentprocessesemployedacrosscountries'decision-making. PHP62 WHATISTHEIMPACTOFCOMPARATIVEEFFECTIVENESSANDVALUEBASED PRICINGONAPRODUCT’SVALUEANDMARKETACCESS?WalkerR,Ng-HaingJ,KoruthR,SparrowhawkK PriceSpective,London,UKOBJECTIVES: Comparativeeffectiveness(US)andvalue-basedpricing(VBP)(UK) areanticipatedtobringchangesfromafree-pricing'systemfordrugstoonewhere pricesareinuencedbygovernmentalauthorities.Aproduct'svaluewilltakeinto accountadditionalfactors,suchaswidersocietalbenetsandtherapeuticinnovation.TheaimofthisresearchistodeterminetheimpactofcomparativeeffectivenessandVBPonthepriceandmarketaccessofnewdrugsindepression. METHODS: Aliteraturereviewwasconductedusingelectronicdatabases(Medline, Embase,GoogleScholar).Thesearchwasperformedfortheyears2009-2011and keytermsincludedcomparativeeffectivenessUSA,valuebasedpricingUK,drug valueandsocietalbenets.Inaddition,ananalysisofaUKDepartmentofHealth consultationpaperandtheUSAgencyforHealthcareResearchandQualitypolicy documentswasperformedinordertodeterminehowcomparativeeffectiveness andVBPmayaffectthemarketaccessandcommercialviabilityofproductsin diseaseareaofdepression(SSRIs,SNRIsandatypicalantidepressants).Aspreadsheetwasusedtocapturedataandacomparisonwasundertakentocompareand contrastthetwomarketsandthedifferentimplications. RESULTS: Theadditionof comparativeeffectivenessandVBParetotakeaccountofindirectcostsassociated withalldiseaseareas.Depressionhashighindirectassociatedcostsassociated and,assuch,thevalueofnovelantidepressantswillincrease.Thisislikelyto enablebetteraccesstonewproducts. CONCLUSIONS: Themovetocomparative effectivenessandVBPislikelytochangethemarketaccessofproducts,particularlyincertaindiseaseareas.Therewillbemorepositivedriversforinvestmentin thediseaseareaofdepression.Moreover,decisionstakenatthemarginduringthe drugdevelopmentprocesswillbeimpactedasanychangeinmarketaccessislikely toaffecttheGo/Nogo'decisioncriteria. PHP63 WHATINFLUENCESPHARMACEUTICALREIMBURSEMENTDECISIONS?A SYSTEMATICREVIEWOFFACTORSREPORTEDTOINFLUENCEDECISIONSIN OECDCOUNTRIESBendingMW1,HuttonJ1,McGrathC2,GlanvilleJ1 1UniversityofYork,YorkHealthEconomicsConsortium,York,UK,2Pzer,Surrey,UKOBJECTIVES: Manyfactorsinuencepharmaceuticalreimbursementdecisions. Thisstudyaimstodeterminetheinuenceoffactorsconsideredintheevaluation ofpharmaceuticalsonthereimbursementdecisionsofgovernmentfundedbodies inOECDcountries. METHODS: Asearchof MEDLINE,EMBASE,EconLit,HealthManagementInformationConsortium,NHSEEDandREPECEconomicworkingpapers until July2010wasconducted.AhandsearchoftheInternationalJournalofTechnology AssessmentinHealthCarewasundertaken(1990-2010).Thefollowingstudydesignswereeligible:experimental,quasi-experimental,retrospective,prospective, caseseriesandsurveysorquestionnairesdesign.TheinuentialfactorswerereviewedacrossandwithinOECDcountries. RESULTS: Thesearchidentied12 quantitativestudiesand23qualitativestudies.Thequantitativestudiesconsideredthecorrelationbetweenfactorsanddecisionseitherthroughregressionanalysisofretrospectivedecisionsordiscretechoiceexperiments.Cost-effectiveness wasfoundtobeconsistentlyinuentialforreimbursementdecision-makingin Australia,England,CanadaandTheNetherlands.Therewasvariationinthedenitionofclinicalconsiderationsandotherfactorsinstudiesconductedincountries.Thislimitedcomparabilitywithinandacrosscountries.Studiesreported mixedevidenceoftheinuenceofthequality,quantityandtypeofclinicalevidence,robustnessofeconomicmodels,sensitivityanalysis,budgetimpact,lackof alternativetherapyandseverityofdiseaseonreimbursementdecisions.Qualitativestudiesreportednarrativedescriptions,casestudiesandinterviewswithdecision-makers.Thesestudiessupportedtheinuenceofcost-effectivenessfound inthequantitativeevidence.Theyadditionallydescribedtheinuenceofthecompositionofthedecisionpanel,committeedeliberations,stakeholderinvolvement andlobbyingondecisions. CONCLUSIONS: ThereislimitedevidenceontheinuenceofevaluationfactorsonreimbursementdecisionsinafewOECDcountries withestablishedreimbursementprocesses.WiderinvestigationofthefactorsinuentialinothercountrieswouldallowcomparisonofthesimilaritiesanddifferencesacrossOECDcountries. PHP64 THERELATIONSHIPBETWEENFINANCIALIMPACTANDTHELIKELIHOODOF DRUGREIMBURSEMENTINTHEAUSTRALIANHEALTHCARESYSTEMMauskopfJ1,ChirilaC1,MasaquelC1,BoyeKS2,BowmanL2,BirtJ2,GraingerD3 1RTIHealthSolutions,ResearchTrianglePark,NC,USA,2EliLillyandCompany,Indianapolis, IN,USA,3EliLillyandCompany,Sydney,AustraliaOBJECTIVES: InthispaperweestimatetherelationshipbetweenthenancialimpactofanewdrugonthehealthcaresysteminAustraliaandtheprobabilityofthe drugbeingrecommendedforreimbursementbythePharmaceuticalBenetsAdvisoryCommittee(PBAC). METHODS: DatainthePBACsummarydatabaseregardingdrug-reimbursementdecisionsmadebetweenJuly2005andNovember2009 wereabstracted.FinancialimpactwascategorizedasA$0orless,greaterthanA$0 upthroughA$10million,andgreaterthanA$10millionperyear.Descriptiveanal-A22VALUEINHEALTH14(2011)A1–A214

PAGE 31

ysis,logisticanalysis,survivalanalysis,andrecursivepartitioningdecisionanalysiswereusedtoestimatetherelationshipbetweenthenancialimpactofanew drugindicationandtheprobabilityofitsreimbursement.Themultivariableanalysescontrolledforotherclinicalandeconomicvariablesthathavebeenshownto becorrelatedwiththeprobabilityofreimbursement,includingthecostperqualityadjustedlife-yeargained. RESULTS: Inallanalyses,nancialimpactwasasignificantpredictoroftheprobabilityofreimbursement.Forexample,inthelogistic analysis,theoddsratioofreimbursementforadrugsubmissionwithanancial impactgreaterthanA$10millioncomparedwithA$0orlesswas0.12(95%condenceinterval[CI]:0.03-0.55);theoddsratioofreimbursementforadrugsubmissionwithanancialimpactgreaterthanA$0upthroughA$10millioncompared withA$0orlesswas0.16(95%CI:0.04-0.60).Similarresultswereobtainedinthe survivalanalysis.Intherecursivepartitiondecisionanalysis,therstsplitofthe datawasforsubmissionswithapositivenancialimpactcomparedwiththose withanegativenancialimpact. CONCLUSIONS: InAustralia,nancialimpacton thehealthcaresystemisanimportantdeterminantofwhetheranewdrugis recommendedforreimbursement,evenwhencost-effectivenessestimatesand otherclinicalandeconomicvariablesarecontrolled. PHP65 HEALTHOUTCOMESANDECONOMICSRESEARCHFORCELLULARTHERAPIES ANDREGENERATIVEMEDICINES:LESSONSFROMAHEALTHTECHNOLOGY ASSESSMENTANDREIMBURSEMENTANALYSISINTHEUNITEDSTATESFaulknerEC,SpinnerDS RTIHealthSolutions,ResearchTrianglePark,NC,USAOBJECTIVES: Cellulartherapiesandregenerativemedicines,arepoisedtohavethe sameparadigm-shiftinginuenceonhealthcareasmonoclonalantibodies(mABs) andpersonalizedmedicine.Whilethesetherapiesholdsimilaritiestoconventional biopharmaceuticals,theyalsodifferinmaterialwaysincludingattributesofboth medicaldevicesandpharmaceuticals;useofmultipleprocedurestoprepareand delivercells;andthepotentialtocuresomediseases.Becauseoftheircomplexity, thesetechnologiesarealsoanticipatedtobecostlyandfaceheavyscrutinyof value.Theobjectiveofthisresearchistoevaluaterecentreimbursementpolicies onregenerativemedicines,comparethemtocurrentbiopharmaceuticals,and evaluatelessonsforhealtheconomicsandoutcomesresearch(HEOR)andreimbursementplanning. METHODS: AsearchofUSHTAsfromtheCentersforMedicareandMedicaidServices,theAgencyforHealthcareResearchandQuality (AHRQ),theBlueCrossBlueShieldTechnologyEvaluationCenterandpubliclyavailablecommercialpayercoveragepolicieswasconductedtoidentifyreimbursement recommendationsandsupportingrationale.Areviewoftheliterature,including theCochraneLibraryandPubMedwasalsoconductedusingrelevantMeSHterms andtextwordstoidentifyadditionalreimbursementissuesassociatedwithregenerativemedicines. RESULTS: Althoughanascenttreatmentarea,over15technologyassessmentsandcoveragepoliciesonregenerativemedicineswereavailable fromUSHTAagenciesandpayers.Differentfrommostothertechnologies,some noncoveragepositionshavebeenestablishedpriortotheemergenceofspecic technologiesorsupportingevidence.TheliteraturerelevanttoHEORissueson regenerativemedicinesisparticularlylimitedtodate. CONCLUSIONS: Whilepayer policiesonregenerativemedicinesreecteddecisionfactorscommonlyassociated withbiopharmaceuticals,otherfactorsbeyondconventionalbiologicswereidentied.Thesefactorsincludespecialevidentiaryconsiderationsforcellprocessing steps,inuenceoftheevidence-basesupportingmulti-proceduralstepsonreimbursement,andevaluationoftheentireprocedurevs.thebiopharmaceutical alone.KeyconsiderationsforHEORarediscussed. PHP66 DEMONSTRATING“DISEASEMODIFYINGTHERAPIES”?ANHTAPERSPECTIVEMcDonaldP,ColasanteW,OshinowoB,SarafS PriceSpectiveLtd.,London,UKOBJECTIVES: Theobjectiveofthisstudywastounderstandwhatstakeholdersin US,EU5,Canada&Australiainterpretasdiseasemodication.Inchronicprogressiveconditions,diseasemodicationversussymptomcontrolistheultimategoal ofhealthcarespecialists.However,thereisnoconsensusonwhat"diseasemodication"reallyis.FromaHTAperspective,notonlyisthereisdifcultyinvaluing diseasemodifyinginterventions,butalsoimpliedrisktopayersapprovingtoreimbursethesedrugsatlaunch. METHODS: Primaryresearchwasundertakenwith payersandmedicalspecialiststounderstandrequirementstosupportdisease modicationclaimsinHTAassessments.Variousattributeswereassessedduring in-depthdiscussionsandthroughdiscretechoiceconjoint.Over100respondents wereinterviewed. RESULTS: TheresultsshowthatthatefcacyisthemostimportantattributeconsideredindiseasemodicationforHTAassessments.Thecorrelationofbiomarkerstoclinicalendpointsalsohasutility.Themarketaccessimplicationsforsuchproductsacrossgeographiesvaryconsiderably.However,costeffectivenessremainsakeydriverinspecicmarkets. CONCLUSIONS: Disease modicationmeansdelayingorhaltingtheprogressionofadisease.Efcacyisthe mostimportantsinglefactor,withtheevidenceofmagnitudeanddurationof effectbothbeingessential.However,efcacyaloneisinsufcienttosupporta diseasemodicationclaim.Robustlongtermdataarealsorequiredanddatashowingameaningfulimprovementovercurrentlyavailabletherapies.Forregulators andpayers,acceptanceofadiseasemodicationclaimmeansthatpayerstakethe riskofreimbursingbasedonsomeextrapolationofdataatlaunch.Acommitment tointegrateaprocessofdatareviewofoutcomesovertimelinkedwithperformancewillensureanyriskismitigated. PHP67 UNDERSTANDINGTHEVIETNAMESEPRICINGANDREIMBURSEMENT ENVIRONMENTTHROUGHACOMPARISONWITHTHATOFCHINALewisS,DummettH DoubleHelixConsulting,London,UKOBJECTIVES: Vietnamisanemergingpharmaceuticalmarketthatisbothpoorly understoodandundergoingsignicantchange,withatargetforintroducinguniversalhealthcareby2014.TheobjectiveoftheresearchwastoplaceVietnam's pricingandreimbursementenvironmentinacontextthatbringsitsdynamicsinto clearerfocusandtogaugethelikelyfuturedirectionofitsevolution. METHODS: A comparisonwithitsneighbour,China,wasconductedbasedoninterviewsinboth marketswithgovernmentadvisers,healtheconomistsandhealthpolicyprofessors,aswellasKOLs.Acomparativeanalysiswasthenconductedofthemarket accessdynamicsanddrivers,aswellasofpolicyreformplans. RESULTS: Many similaritiesexistbetweentheP&Renvironmentsofthetwomarkets,althoughthey sitatdifferentplacesalongtheP&Rdevelopmentcontinuum.Inboth,branded drugsenjoyaconsiderablepricepremiumovergenerics,whichcouldbeashighas 40timesinVietnam.However,Chinaislookingtochangethesituationbyremovingthepremiumforoff-patentbrandeddrugs.Despiteeffortsintodeveloptheir reimbursementsystem,direct-saletohospitalsistheprimaryrevenuechannelfor pharmaceuticals,withKOL-endorsementamajormarketaccessdriver.Intermsof drugpriceregulation,themostcommonlyusedmeasureisthroughtheenforcementofpricecaps. CONCLUSIONS: TheVietnameseP&Rsystemissimilartothat ofChina5-10yearsago.Severalfundamentsintermsthestructureofthehealth system,formalisationofthereimbursementsystemandmarketaccessdriversare thesame,butChinaissignicantlyfurtherdownthelinethanitsneighbour.However,thehealthreformagendasofthetwomarketsarebothheadinginsimilar directions. PHP68 THE15YEARSEXPERIENCEOFNEWDRUGSADOPTIONANDREIMBURSEMENT INTAIWAN’SNATIONALHEALTHINSURANCEHuangWF1,HsiehCF1,ChenGT2 1NationalYang-MingUniversity,Taipei,Taiwan,2KooFoundationSunYat-SenCancerCenter, Taipei,TaiwanOBJECTIVES: TopresenttheempiricalexperienceofnewdruglistingandreimbursementunderTaiwan'sNationalHealthInsurance(NHI),andtodiscussthe performanceofsuchmechanism.Wealsoattempttoassessitsimpactonthe publicaccesstopharmaceuticalinnovations. METHODS: Thematerialsarebased onthedocumentationofTaiwan'sNHIDrugReviewCommittee(DRC)over15years period(1996 2010).Wedenedthecriteriaofpricingmethodsinto9categories: InternationalPriceComparison,ComparisonwithSimilarProductswithEquivalentTherapeuticEffects,PriceProportionMethod,PriceAddition,OrphanDrugs, TheLowestAvailableInternationalPrice,CostAnalysis,GroupingandOthersincludedrisksharingmanagements. RESULTS: Thetotalnumberofnewdrugsthat appliedforNHIlistingandreimbursementduring1996 2010was1103,andthe numberofpetitioncaseswas587(53%).Thetotalnumberofnewdrugswithnal pricingdecisionsinthisstudywas802.Amongthem,343itemswereissuedwith reimbursementpricewithoutpetition,andtheremainingnewdrugsreceivedtheir reimbursementpricesafterpetitiononinitialpricingdecisions.Theapprovedprice wasaveraged69%oftheinternationalmedianprices,andwasonly60%ofinternationalmedianpricesamongpetitioncases.Thetopthreemethodsofpricingare PriceProportionMethod,EquivalentTherapeuticEffectwithSimilarProduct (23.5%),andInternationalPriceComparison.BecauseNHIfacednancialcrisisin thepast10years,thepriceofreimbursementcameasapproximately70%oftheir respectiveapplicationpricesintherecentyears.Moreandmorerisk-sharing agreementswereintroducedinthepricenegotiation. CONCLUSIONS: ThepotentialpressureontheNHImaybeevengreaterforgrowingnancialgapinthefuture. Therearemoreeffortsneededtoseekthetransparencyinthelistingandpricing processforpublicgood. HealthCareUse&PolicyStudies–HealthCareResearch&Education PHP69 HOWMUCHFORAQALYINKOREAKimY1,ShinS1,ParkS1,SongH1,ParkJ1,BaeE2,AhnJ1 1NationalEvidence-basedHealthcareCollaboratingAgency(NECA),Seoul,SouthKorea,2Sangji University,Wonju,Kangwon-do,SouthKoreaOBJECTIVES: Tomeasurewillingnesstopay(WTP)foraQALYinKorea. METHODS: AsurveyquestionnairebasedonEQ-5DscenarioswasdevelopedtomeasureQALY improvementsinKorea.Doubleboundeddichotomouschoice(DBDC)questions alongwithanopenquestionwereusedtoelicitWTPs.Eachpersonwasaskedfor fourscenarioschosenfrom3-itemEQ-5Dscenarios( 1QALY)andanadditional scenariowithliveinperfecthealthfor1yearordienow( 1QALY).ThesameWTP questionswerealsorepeatedforQALYimprovementsofafamilymemberinstead ofself.Thequestionnairealsoincludedquestionsondemographics,diseasestatus, andavisualanalogscale(VAS)measureofeachscenariopresented.Consistencyof eachrespondentwascheckedbymatchingranksofvescenariosbetweenWTPs andQALYimprovementseitherbyVASorKoreanEQ-5Dtariff.InitialbidsforDBDC questionsweredeterminedbythequintilesofpilotsurveyWTPs.Surveyquestions foreachstudywerenetunedthoughtwofocusgroupinterviewsessionsper study.Ageneralpopulationsamplewasinterviewedface-to-facein2010. RESULTS: Ofthetotal1,017personssurveyed,933personspassedconsistencytest. ForthosewhopassedconsistencyandnotinMedicalAidsprogram,WTPfora QALYcalculatedfromthenalopenquestions(afterDBDCquestions)was19millionKRW(approximately16,000USD).WTPforfamilymemberwereconsistentlyA23VALUEINHEALTH14(2011)A1–A214

PAGE 32

higherthanself.UsingDBDCquestionsseemstohaveadditionalbenetofrening openendedquestions.TheseresultswerehigherespeciallyinparametricestimationmodelsusingonlyDBDCquestions.Thisobservationcautionsusofselectinga specicanalyticmethodmayinuencetheresults. CONCLUSIONS: Thisstudyis therststeptowardresolutionofcontroversiesaroundeconomicevaluationsof healthcareinKoreaandhopestoencouragemorelocalresearchonthisissue. PHP70 PREVENTABLEHOSPITALIZATIONSANDHEALTHINSURANCECOVERAGEIN HARRISCOUNTY,TEXASMurtyS,BegleyCE,SwintJM UniversityofTexasHealthScienceCenteratHouston,Houston,TX,USAOBJECTIVES: Preventablehospitalizations(PHs)arehospitalizationsthatcanbe avoidedwithappropriateandtimelyaccesstoprimarycare.LackofhealthinsurancehasbeenassociatedwithfrequentPHsinseveralcommunitiesintheU.S.In HarrisCounty,publicandprivatesafetynetclinicsprovidesubsidizedprimarycare touninsuredindividuals.Therefore,beinguninsuredmaynotresultinunnecessaryhospitalizations.Theobjectiveofthisstudywastodeterminetheprevalence ofPHsandtoexaminetheeffectofhealthinsuranceontheriskofPHsinHarris County(Houston),Texas. METHODS: Thisstudywasrestrictedtonon-elderlyadult (age18to64)residentsofHarrisCounty,Texas.TheAgencyofHealthcareResearch andQuality(AHRQ)guidelineswereusedtoidentifyahospitalizationaspreventableornotfromtheTexasHealthCareInformationCollectioninpatientdischarge data2007.PrevalencewasmeasuredastherateofPHs;i.e.thenumberofPHsper 100hospitalizations.Healthinsurancestatuswasclassiedas"uninsured","Medicaid-insured"or"privately-insured".A2-analysisfollowedbyamultivariable logisticregressioncontrollingfordemographicandsocioeconomiccharacteristics wasperformedtodeterminetheeffectofhealthinsuranceontheriskofPHs. RESULTS: Therewere17,810PHsamongnon-elderlyadultsinHarrisCountyin 2007(PHrate 7.83PHsper100hospitalizations).Healthinsurancestatuswas signicantlyassociatedwithPHs(2 781.83;p 0.001).Inthemultivariablemodel, theuninsuredhad60%higheroddsofhavingaPH(OR 1.628;p 0.001)ascomparedtotheprivatelyinsured.Medicaid-insurancewasalsoassociatedwitha marginallyhigherriskofPHs(OR 1.065;p 0.05). CONCLUSIONS: Lackofhealth insuranceisanimportantfactorindeterminingriskofPHsinHarrisCounty,despitethepresenceofsafetynetclinics.Ifrecenthealthcarereformsaresuccessful inincreasinghealthinsurancecoverage,alargeproportionoftheseunnecessary hospitalizationsmaybeavoided. PHP71 IMPACTOFTHEMEDICAREPARTDCOVERAGEGAPONPRESCRIPTIONDRUG UTILIZATIONANDMEDICATIONADHERENCENaikR1,BorregoM2,DoddM2,RaischD2,KhanN3,BakhirevaL2,GupchupG4,CunicoL2 1Allergan,Inc.,Irvine,CA,USA,2UniversityofNewMexico,Albuquerque,NM,USA,3Oxford OutcomesLtd.,Morristown,NJ,USA,4SouthernIllinoisUniversityEdwardsville,Edwardsville,IL, USAOBJECTIVES: ToassessimpactoftheMedicarePartDcoveragegaponprescription drugutilizationandmedicationadherenceusingdatafromMedicarebeneciaries enrolledinalargehealthplaninNewMexicoin2007. METHODS: Quasi-experimental,retrospective,pre-postwithcontrolgroupstudydesignwasutilized.Preandpost-coveragegapprescriptiondrugutilizationandmedicationadherenceof beneciariesenrolledinahealthplanwithnoprescriptiondrugcoverageduring thecoveragegap(nocoverageplan)wascomparedwithgenericdrugcoverage (genericcoverageplan)andfullprescriptiondrugcoverage(fullcoverageplan)plan beneciaries.Prescriptiondrugutilizationwasassessedusingtotalnumberofprescriptionspermember.MedicationadherencewasmeasuredusingMedication PossessionRatioandtheProportionofDaysCovered(PDC). Difference-indifferenceanalysis(DiD)wasusedtocomparepre-andpost-coveragegapprescriptiondrugutilizationandmedicationadherencebetweenthethreeplans. RESULTS: Ofthe14,846beneciarieswhometinclusionandexclusioncriteria, 2,661(17.92%)enteredthecoveragegapin2007.DiDanalysesindicatedthat beneciariesinthenocoverageplanlledsignicantlyfewerprescriptionsin thepost-coveragegapperiod,thanbeneciariesinthefull(14.67fewerprescriptions;p 0.001)andgeneric(12.52fewerprescriptions;p 0.001)coverageplans. Signicantdecreaseinpost-coveragegapmedicationadherencewasobserved betweennocoverageandfullcoverageplanbeneciarieswithrespecttostatins (5.8%),ARBs(16%)andPPIs(18.1%).Signicantdecreaseinpost-coveragegap medicationadherencewasobservedbetweennocoverageandgenericcoverage planbeneciariesutilizingstatins(1.1%)andARB's(12%).Nosignicantdifferenceswereobservedbetweenfullandgenericcoverageplanbeneciaries.SignicantdifferencesinadherencewerefoundonlywhenadherencewasmeasuredusingthePDC. CONCLUSIONS: Lackofprescriptiondrugcoverageduring theMedicarePartDcoveragegapmayleadtodecreasedutilizationandadherencetocertainclassesofprescriptiondrugs. PHP72 ASSESSINGTHEQUALITYOFPHARMACOECONOMICSTUDIESININDIA:A SYSTEMATICREVIEWDesaiP1,ChandwaniH1,RascatiK2 1UniversityofTexasatAustin,Austin,TX,USA,2UniversityofTexas,CollegeofPharmacy, Austin,TX,USAOBJECTIVES: TheaimofthestudywastoevaluatethequalityofpharmacoeconomicstudiesbasedinIndia. METHODS: Aliteraturesearchwasconductedusing PubMed,Medline,EconLit,PsycInfoandGoogleScholartoidentifypublishedwork onpharmacoeconomicsstudiesbasedinIndia.Originalresearchstudiesthatevaluatedpharmaceuticals,andwereconductedbetween1990and2010inIndiawere included.Tworeviewersindependentlyreviewedthearticlesusingasubjective 10-pointscale(10beingthehighest)andthe100-pointQualityofHealthEconomic Studies'(QHES)questionnaire(100beingthehighest). RESULTS: Theincludedarticles(n 29)werepublishedin23differentjournals.Only9articleswerepublished injournalsbasedinIndia.Therstarticlewaspublishedin1998.Eacharticlewas writtenbyanaverageofveauthors.ThemajorityofauthorsresidedinIndia(62%) atthetimeofpublicationandhadamedicalbackground(90%).Cost-effectiveness analysiswasthemostfrequentlyusedmethodofanalysis(79%).Thesourceof fundingandstudyperspectivewasnotlistedin45%and41%ofthestudiesrespectively.Thestudydesignwasarandomizedcontrolledtrialfor41%ofthestudies. Themeansubjectivequalityscoreofallthearticles(n 29)was7.8(SD 1.3)and themeanQHESscoresforthestudiesevaluatingcostsandoutcomes(n 24)was 86(SD 6).Thequalityscorewassignicantly(p 0.05)relatedtocountryofresidenceofprimaryauthor(non-India higher)andthestudydesign(randomized controlledtrials higher). CONCLUSIONS: Theneedforeconomicevaluationof pharmaceuticalsisimperative,especiallyindevelopingcountrieslikeIndia;becauseitcanhelpdecisionmakersallocatescarceresourcesinajustiablemanner. Standardizationofguidelines,andimprovedpharmacoeconomiceducation-startingfromtheundergraduateleveltospecialization-aretwosuggestionstohelpto improvethequalityofthepharmacoeconomicresearchinIndia. PHP73 SYSTEMATICLITERATUREREVIEWONHETEROGENEITYOFRELATIVE EFFECTIVENESSANDRELATIVEEFFICACYOFMEDICINESINEUROPEPuig-Peir—R,TowseA,Mestre-FerrandizJ OfceofHealthEconomics,London,UKOBJECTIVES: Tounderstandtheextentoflikelyvariationifanyinunderlying relativeefcacyandrelativeeffectivenessofdrugsusedinoneormoreofthe27 MemberStates(MS). METHODS: Asystematicliteraturereviewofstudiesinfour databaseswasconductedfortheperiodJanuary2000-May2010:Medline,Embase, EconLitandHealthManagementInformationConsortium(HMIC).Logicalcombinationsofkeywordsrelatedtoeffectiveness,generalisability,externalvalidity, transferability,Europeandreviewweresearched. RESULTS: Atotalof326articles wereinitiallyidentiedandtenformedthebasesofthereview.Ninewere reviewsororiginalcost-effectivenessstudiesandrelevantdataontheeffectivenessresultswasextracted.First,wefoundthatthereisacommonassumption intheliteraturethatrelativeefcacyisconstantacrosscountriesandtherefore generalisable.Lessisknownaboutwhetherrelativeeffectivenessisindeed similarordifferent.Second,wegroupedthefactorsthatcouldpotentiallyintroducevariationinrelativeeffectiveness:(i)patientsanddiseasecharacteristics;(ii)comparatorsused;(iii)differentmeasuresofhealthoutcomes;(iv)variationsinclinicalpractice.Third,noobservationalstudieswereidentied. Evidencefromstudiesusingefcacydata,mainlyfromRCTs,wasmixed;three studiesfoundnodifferencesinclinicaloutcomesacrosscountries,onestudy founddifferencesduetopatients'characteristicsandseverityandanother studyusingmultilevelanalysisfoundcountryandpatientscharacteristicsexplainingpartlytheheterogeneityoftreatmenteffects.Onereviewreportdifferencesinclinicalpracticeasonemajorcausesofvariationsinclinicaloutcomes betweencountries. CONCLUSIONS: Theliteraturewasscarceandtherefore moreevidenceisneededbeforeanystatementcanbeclaimedontheexistence ofvariationsinrelativeeffectivenessorefcacyindifferentcountries.Future collaborationamongtheMSharmonisingmethodology,generatingdataand sharingpatients'registrydatawillbecrucialtoproducethisevidence. PHP74 CLINICALTRIALLEARNINGCURVESAFFECTOUTCOMESMEASUREMENT, PATIENTSAFETYANDTRIALSUCCESS:CONTRIBUTINGFACTORSANDA POTENTIALSOLUTIONINCLINICALTRIALSIMULATIONSpinnerDS1,WalterJW1,Vazquez-GraggM1,BloiseT1,Stafford-SmithM2,FaulknerEC1, TaekmanJM2 1RTIHealthSolutions,ResearchTrianglePark,NC,USA,2DukeUniversityMedicalCenter, Durham,NC,USAOBJECTIVES: Evidencehasrecentlyemergedregardingthepresenceofinvestigator learningcurvesinthecontextofclinicaltrials(i.e.,theassociationbetweenclinical protocoldeparturesandpatientenrollmentsequence).Wereviewedtheliterature toidentifyfactorscontributingtoclinicaltriallearningcurvesandpotentialimpactsoflearningcurvesonoutcomesaccuracy,patientsafetyandoveralltrial success. METHODS: ThePubMeddatabasewassearchedforstudiesdescribing clinicaltriallearningcurveeffectsandpotentialsolutionstotheseeffects,using searchtermcombinationsincluding"clinicaltrial","clinicalprotocol","learning curve",and"enrollmentsequence". RESULTS: Learningcurveeffectswereidentiedintrialsincludingthosefortreatmentofsepsis(PROWESSandADDRESS)and high-riskcardiacdisease(VALIANT).Outcomesofthesetrialspotentiallyresulting fromlearningcurvesincludedambiguoustreatmentbenet,increasedadverse eventsandmortality,andoveralltrialfailure.Potentialcontributingfactorsinclude thoserelatedtostudysite(e.g.,largemulticentertrials;siteswithpoorenrollment orprevioustrialexperience;inexperiencedstaffandhighstaffturnover),study protocols(e.g.,impreciseorcomplexprotocols),anddiseaseseverity.Preventing suchlearningcurveeffectswouldlikelyhaverequiredextensivetrialsiteandstaff screeningandtraining,aswellasimprovedtrialprotocoldesignpriortorstpatientenrollment.Clinicaltrialsimulation,amethodologyidentiedinseveralreports,couldpotentiallyaddressclinicaltriallearningcurvesbyimprovingprotocol designandstaffcompetence.Thisapproach,whichfocusesonhumanfactors involvedinconductingclinicalprotocols,offersthepotentialtoimprovetheaccuracyofmeasuredtrialoutcomesanddecreasethelikelihoodofprematuretrialA24VALUEINHEALTH14(2011)A1–A214

PAGE 33

termination. CONCLUSIONS: Clinicaltriallearningcurveshavesignicantimplicationsforoutcomesresearchaccuracy,patientsafety,andoveralltrialsuccess. Clinicaltrialsimulationmayprovideabroadlyapplicablemethodologyforaddressingseveralfactorsassociatedwithclinicaltriallearningcurveeffectsandfor improvingtheaccuracyofclinicaltrialoutcomes. PHP75 USEOFELECTRONICMEDICALRECORDSFROM2001THROUGH2010: IMPLICATIONSFORCOMPARATIVEEFFECTIVENESSRESEARCHAnastassopoulosKP1,MannR2,KnightTG1,SudharshanL1,AckermanSJ2 1CovanceMarketAccessServices,Inc.,Gaithersburg,MD,USA,2CovanceMarketAccess Services,Inc.,SanDiego,CA,USAOBJECTIVES: RecentdevelopmentsintheUnitedStates(US)healthcarereform andfundingforcomparativeeffectivenessresearchsuggestthatuseofelectronic medicalrecords(EMR)inoutcomesresearchmayincreaseovertime.EMRcanbe particularlyusefulwhenoutcomesarenotwell-denedwithdiagnosisorprocedurescodesorwhenclinicaldataareneeded.Theobjectiveofthisstudywasto reviewtrendsintheuseofEMRduringthepastdecade. METHODS: Areviewof publishedliteraturewasconductedinPubMedforyears2001through2010toidentifyoutcomesstudiesintheUSthatusedEMR.Internalqualityassurancestudies andvalidationstudiesthatusedEMRwereexcluded.Thenumberofstudies,settingofcare,patientpopulation,whetherthestudywascomparative,andanynoted limitationswereexamined. RESULTS: Atotalof58EMR-based,outcomesstudiesin theUSwereidentiedoverthepastdecade;increasingfrom3in2001to12in2010. ThemajorityofstudiesincludedoutpatientEMR.Studiesincludedavarietyof patientpopulationswithoverone-thirdincardiovasculardisease,psychiatricdisease,anddiabetescombined.Thepercentofstudiesthatwerecomparativeranged from0%in2001to45%in2010.Measuresofeffectivenessvariedwidelyandincludedlabvalues,clinicalmeasures,andhealth-relatedquality-of-lifeoutcomes. SomenotedlimitationsontheuseofEMRdatainoutcomesresearchincludedlack ofrepresentativenessofallcaredeliveredacrosspracticesettings,lackofgeneralizabilityandstandardization,andrelianceonhealthcareproviderreporting. CONCLUSIONS: AlthoughtheuseofEMRinoutcomesresearchhasincreased slowlyinthepastdecade,theproportionofcomparativestudiesusingEMRhas increasedovertime.AstheindustryworkstostandardizeEMRandmoreadvanced outcomesarecollectedinEMRsystems,EMRdatamayplayalargerroleincomparativeeffectivenessresearch. PHP76 THEOUTLOOKOFLARGESIMPLETRIALSFORCOMPARATIVEEFFECTIVENESS RESEARCH:ANAPPLICATIONOFTHEPRECISFRAMEWORKVincentL QuintilesGlobalConsulting,Hawthorne,NY,USAOBJECTIVES: Inrecentyears,comparativeeffectivenessresearch(CER)hasgained increasinginterestandinvestmentacrossmajorlifescienceandhealthpolicy stakeholdergroups:regulators,biopharmaceuticals,physicians,patients,andpayers.Onemethod,largesimpletrial(LST),ahybridofobservationalcohortstudyand randomizedcontroltrial(RCT),isdesignedforlargenumbersofpatientsinpostapprovalresearchandisachoicestudydesignforcomparingtherelativestrengths andweaknessesofmedicalinterventions.TheobjectiveofthisstudywastoevaluatethecurrentlandscapeofLSTstocapturekeythemesinthefutureoutlookof CER. METHODS: Astructuredabstractreviewandfunnelanalysiswasconducted usingdataavailablethroughClinicalTrials.gov.Usingthesearchterms,"phaseIV", "post-marketing","randomized",and"multi-centered",2230clinicaltrialswere lteredwithrestrictedparametersfornumberofpatients(n 1000),evidenceof blinding,statedobjectiveofcomparativeeffectiveness,anddateoftrialinitiation. RESULTS: Togainaperspectiveoncurrentandfutureoutlook,onlytrialsinitiated between2009and2010(n 10)wereassessedinthisphaseofstudy.Usingthe PragmaticExplanatoryContinuumIndicatorSummary(PRECIS)frameworkoriginallydevelopedforpragmatictrials,theanalysisrevealedanimportanttheme. Whenapplyingkeydomaincriteriapractitionerexpertise,treatmentexibility, eligibilitycriteria,healthoutcomes,studyduration,intent-to-treat(ITT),andprimaryoutcomeanalysis,asdenedbythePRECISframeworklessthan50%ofthe LSTCERtrialsweredesignedtomeasureeffectiveness,asopposedtoefcacy. CONCLUSIONS: Accordingtondings,therehasbeenanuptakeontheuseofLSTs ininvestigatingtherelativeeffectivenessofinterventions.However,considering theincreaseingovernmentalpressuretoreducehealthcarespendingbyimproving qualityoftreatment;itisimportantthattheseLSTsareconductedtotrulycapture effectiveness.Forthatreason,thereisneedforfurtherresearchtoextractand developnewframeworksforevaluation. PHP77 THESTATEOFHEALTHECONOMICSANDPHARMACOECONOMICSRESEARCH INRUSSIA:ASYSTEMATICREVIEWGavazaP1,ShepherdMD2,ShcherbakovaN3,KhozaS3 1AppalachianCollegeofPharmacy,Oakwood,VA,USA,2CenterforPharmacoeconomicStudies, CollegeofPharmacy,UniversityofTexasatAustin,Austin,TX,USA,3DivisionofPharmacy Administration,CollegeofPharmacy,UniversityofTexasatAustin,Austin,TX,USAOBJECTIVES: ToinvestigatethestateofhealtheconomicresearchinRussiaavailableintheEnglishlanguagebydescribingthenumberandcharacteristicsofthe articles,andassessingthequalityofthesearticles. METHODS: Thestudyassessed thestateofhealtheconomicsandpharmacoeconomicsresearchinRussia.We conductedaliteraturesearchtoidentifyhealtheconomicsarticlespertainingto Russia.Eacharticleinthenalsamplewasscoredbytworeviewersindependently usingthedata-collectionformdesignedforthestudy. RESULTS: Intotal,16studies investigatingawidevarietyofdiseaseswereincludedinthestudy.Thesearticles werepublishedin15differentjournalsallbasedoutsideofRussiabetween1994 and2009.Onaverage,eacharticlewaswrittenbysevenauthors.Mostrstauthors hadmedical/clinicaltrainingandresidedintheUSA(n 8)atthetimeofpublicationofthestudy.Basedonascaleof110,with10indicatingthehighestquality,the meanqualityscoreforallstudieswas8.09(SD 1.29)and25%ofthearticleswere offairquality(score57).Thequalityofarticleswasstatisticallysignicantlyrelated(P 0.05)totheprimaryhealthintervention(pharmaceuticals non-pharmaceuticals)andprimarytrainingoftherstauthor(medical non-medical). CONCLUSIONS: Theconductofhealtheconomicsandpharmacoeconomicsresearch inRussiaintheEnglishlanguageislimitedand,onaverage,thepublishedarticles wereofgoodquality.However,aboutone-quarterofpublishedarticleswereoffair quality.MorehealtheconomicsresearchinEnglishiswarrantedinRussia. PHP78 CHARACTERISTICSOFHOMELESSINDIVIDUALSUSINGINPATIENTAND EMERGENCYDEPARTMENTSERVICESKaracaZ,WongH,MutterR AgencyforHealthcareResearchandQuality(AHRQ),Rockville,MD,USAOBJECTIVES: Thisstudycomparesthepatientcharacteristics,insurancecoverage, diseaseprevalence,andutilizationpatternsofinpatientandemergencydepartment(ED)servicesbetweenhomelessandnon-homelesspeopleusinganewdata sourcethathasnotbeenemployedinexistingstudies. METHODS: Aretrospective dataanalysiswasconductedtocomparedifferencesinpatientcharacteristics, insurancecoverage,diseaseprevalence,andutilizationpatternsbetweenhomelessandnon-homelessindividualswhovisitedahospitalorhospital-basedemergencydepartment.Foreachservicetype,separaterateswerecreatedforhomeless andnon-homelesspopulationsandthencompared.TheHealthcareCostandUtilizationProject(HCUP)StateInpatientDatabases(SID)andStateEmergencyDepartmentDatabases(SEDD)for2008wereusedintheanalysis.TheSIDemployedin thisstudyinclude15.9millioninpatienthospitaldischargesfromcommunityhospitalsintenstatesArizona,California,Colorado,Florida,Georgia,Massachusetts, Missouri,NewYork,PennsylvaniaandWisconsin.177,056ofthesedischargeswere homelesspatients.TheSEDDemployedinthisstudyinclude23.7millionvisitsto ED,wherepatientsweretreatedandreleased,fromsevenstates(i.e.,Arizona, Florida,Georgia,Massachusetts,Missouri,NewYorkandWisconsin).49,595of thesevisitswerehomelesspatients. RESULTS: Theuninsuredhomeless(nonhomeless)patientswereaccountablefor28.1%(4.6%)ofinpatientadmissionsand 42.85(21%)ofEDvisits.Medicaidcovered48.2%ofallinpatientdischargesand 34.7%ofallEDvisitsbyhomelesspatients.73.7%(50.6%)ofhomeless(non-homeless)inpatientadmissionscamethroughEDs.Homelesspatientswithmentaldisordersaccountedfor22.4%ofallhomelessinpatientdischargesand49%ofall homelessEDvisits.Amajorityofthementaldisorderdiagnosisinbothsettingsof carewerealcohol-relateddisorders,mooddisorders,andschizophrenia. CONCLUSIONS: Theproleofhomelessandnon-homelesspatientsdifferedsignicantlybyinsurancestatusandbyraceinboththeinpatientandemergency departmentsettings. PHP79 ROLEOFHTASYSTEMSINREIMBURSEMENTANDMARKETACCESS: COMPARISONOFTURKEYANDPOLANDKirpekarS,ShanklandB,DummettH DoubleHelixConsulting,London,UKOBJECTIVES: HealtheconomiesoutsideWesternEuropeareincreasinglyadopting evidence-baseddecision-making,butimplementationmethodsdiffer.CommonalitiesintheapproachtakeninTurkeyandPolandneverthelessillustratetherole ofapreviouslydevelopedreimbursementprocessintheadoptionofHTA.Insights fromthiscomparisonmayhaveimplicationsforindustry,governmentandprivate reimbursementauthoritiesworldwide. METHODS: TurkeyandPolandwereselectedforcomparisonduetotheirrelativelyrecentadoptionofHTAwithinan establishedreimbursementsystem.Atotalof18stakeholderswereinterviewedvia telephoneinterviewstounderstandcurrentandfutureutilisationofHTAinreimbursement.Thestudyevaluatedtheimpactofthefollowingdecision-makingdomains:clinicaldifferentiation,prioritisationofunmetneeds,price,andpresentationrequirements(e.g.budgetimpactanalysis).Acomparisonwasthenmadeona ratingscaledevisedtoaccountfortheseinuencingfactors. RESULTS: InTurkey thepharmacoeconomicsunitoftheSocialSecurityInstituteisresponsiblefor economicevaluation.HoweverthePolishHTAbodyAHTAPolismoredirectlyinvolvedintheHTAprocess.All9Polishrespondentsscoredhighontheinuenceof HTAinaccessdecisionswhilejust2respondentsinTurkeyconsideredHTAtobe extremelyinuentialinreimbursement.However,pricingdecisionsinbothcountriesarebasedoninternationalreferencingtovaryingdegrees. CONCLUSIONS: Despitesupercialsimilaritiesinthestructureoftheirreimbursementsystems, HTAinPolandismorecloselyintegratedwithdecisionmakingthaninTurkey.At presentHTAplaysaformalroleintheTurkishsystembutanegotiation-based approachremainstocustomarychannelforvaluearguments.However,therapid paceofchangeintheTurkishreimbursementsystemsuggeststheneedforfurther ongoingresearch. PHP80 RELATIONSHIPOFHERBALKNOWLEDGETOWARDSATTITUDEFORMATION AMONGFUTUREPHARMACISTSSuraSD,ChabriaA,SansgiryS UniversityofHouston,Houston,TX,USAOBJECTIVES: Trendsintheuseofherbalmedicationscontinuetorise.Pharmacists haveanopportunitytoprovideconsumerswithevidencebasedinformation.This studyexaminedpharmacystudent'sknowledgeregardingherbalmedicationfromA25VALUEINHEALTH14(2011)A1–A214

PAGE 34

coursesandtheassociatedoutcomestowardsattitudeformation. METHODS: Ina cross-sectionalprospectivelydesignedstudyasurveywasself-administeredto second,thirdandnalyeardoctorateofpharmacystudentsataUniversitywith diversestudentbody.Thequestionnaireincludedfoursections:attitude,education,knowledgeanddemographic.Pre-validatedinstrumentswereusedtomeasureattitudetowardsherbalmedicationsusingave-pointlikertscalewhile knowledgewasmeasuredusingatrue-falsescale.Descriptivestatisticswereused tosummarizedataandmultiplelogisticregressionwasconductedtoassessthe effectofpharmacystudent'sherbalmedicationknowledgetowardsattitudeformationwhilecontrollingforage,race,sex,yearofstudy,workexperienceand herbalmedicationrelatedcourses. RESULTS: Of206students,159completedthe survey(responserateof77.18%).Meanageofparticipantswas26( 3.2)yearswith 65.41%femaleand81.13%hadpreviouspharmacyworkexperience.Theaverage scoreontheherbalknowledgewas9.28 1.80(maximumscoreof15).Students withhigherknowledgehadpositiveattitudetowardsherbalmedications(OR 1.26; 95%CI 1.01-1.57).Studentswithworkexperiencehadpositiveattitudetowards herbalmedications(OR 2.93;95%CI 1.05-8.14).However,81%studentsbelieved thattheydidnothavesufcientknowledgeofherbalmedicationsand91%studentsreportedthattheyneededmoreinformation. CONCLUSIONS: Higherknowledgeleadstoapositiveattitudetowardsherbalmedicationsamongpharmacy studentsindicatingthattheymayrecommendsuchproductstopatients.Evaluatingtheroleofsuchrecommendationstoimprovepatientshealthoutcomesisthe directionforthefuture. PHP81 CAREGIVERWELL-BEINGANDHEALTHCAREACCESSANDQUALITYAMONG CHILDRENWITHCHRONICPHYSICALANDMENTALCONDITIONSRaneP,SambamoorthiU WestVirginiaUniversity,Morgantown,WV,USAOBJECTIVES: Thewell-beingofcaregiversofchildrenwithchronicconditionscan beinuencedbyseveralaspectsoftheirchild'scareandtheircomplexhealthcare needs.Theobjectiveofthisstudyistoexaminewell-beingandexperiencesof caregiversofchildrenwithdifferentchroniccondition. METHODS: Thestudywas cross-sectionalandusedresponsesofcaregiversof19,423children(aged6-17 years)fromtheNationalSurveyofChildren'sHealth,2007.Wecreatedahierarchy ofchronicconditionstoclassifychildrenfromouranalyticsamplein6mutually exclusivegroups:AutismSpectrumDisorder(ASD),Developmentaldelay/Learning disability,Attentiondecithyperactivitydisorder/behavioralproblems,Depression/anxiety,Speech/hearing/visionproblems,andDiabetes/asthma.Weassessed caregiverwell-beinginthreedimensions:physicalhealth,mentalhealth,andemotionalstress Healthcareexperiencesweremeasuredintwodomains:Access (healthinsuranceadequacyandconsistency;andunmethealthcareneeds),Quality(family-centeredcare-FCC,andeffectivecare-coordination-ECC).Allanalyses adjustedfortheforthecomplexsurveydesign,toderivenationalestimates.Chisquaretests,logisticandmultinomiallogisticregressionswereperformedinSAS 9.2. RESULTS: Physicalhealthofcaregiverswasnegativelyaffectedbyaccessto care(unmetneeds)andhealthcarequality(lackofFCC).Andpoormentalhealth wasnegativelyassociatedwithlackofECC.Ascomparedtocaregiversofchildren withasthma/diabetes,ahigherpercentageofcaregiverswithASDchildrenreportedpoorphysicalhealth(26.6%vs12.7%),poormentalhealth,andhighemotionalstress(26.5%vs2.1%). CONCLUSIONS: Child'sASDplacessignicantburden onthecaregiverwell-being,comparedtootherconditions.Inadditiontothetypeof child'sdisabilityandchroniccondition,caregiverswereathighriskforpoorphysicalhealthandhighemotionaldistressduetounmetneedsforchildren'shealth andlowerlevelsofthechild'shealthcarequality. PHP82 THEIMPACTOFINTEGRATEDMEDICALCARESERVICESFORLOYALPATIENTS INA2000-BEDMEDICALCENTERUNDERTAIWAN’SUNIVERSALHEALTH INSURANCEPROGRAMLinHW1,YuIW2,LinCH2,ChangCK2,ChangLY2,HsiehYW2,WuHC2,LinCC1, HungJS2 ,ChoDY2 1ChinaMedicalUniversity,Taichung,Taiwan,2ChinaMedicalUniversityHospital,Taichung, TaiwanOBJECTIVES: WithveryfewrestrictionsonchoosingphysiciansunderNational HealthInsurance(NHI)programinTaiwan,patientscouldvisitoutpatientdepartments(OPD)inatertiarymedicalfacilitytoseektheprimarycareotherthan secondarycare.ThisstudywastoexaminetheimpactofIntegratedMedicalCare (IMC)servicesinChinaMedialUniversityHospital(CMUH),a2000-bedmedical center,inTaiwan. METHODS: Thosepatientswhomademorethan50%oftheir totalOPDvisitsinTaiwantowardCMUHduringJanuary2008toJune2009(baseline period)wererecognizedasCMUHloyalpatients.Theywereinvitedtoutilizethe offeredpluralisticIMCservices,includingintegrated,geriatricandpharmaceutical careclinics,inadditiontousualprimaryandspecialtyclinics,duringDecember 2009toDecember2010(implementationperiod).Thosewhousedtovisitoneortwo specialistsinCMUHwereofthemostinterest.Thecomputerizedremindingand medicationutilizationevaluation(MUE)systemswereestablishedtofacilitatethe cooperativephysician-clinicalpharmacistmedicationtherapymanagement model.Themedicalandmedication-relatedissueswerereviewedbyprescribers, clinicalpharmacistsandtheclinicalMUEteamasneeded.ThedifferencesofmedicalexpenditurereimbursedbyNHI,OPDvisits,numberofprescribedmedication (Rx)andemergencydepartment(ED)visitsinCMUHduringbaselineandimplementationperiodswereexamined. RESULTS: Of11,902loyalpatients,75%made visitstowardoneortwospecialtiesandotherwise.Themedicalexpenditures,OPD visits,EDvisitsreduced2.4%,4.8%and6.3%perpersonpermonth,respectively,but thenumberofRxincreased0.2%.Thedifferencesofthehealthresourcesutilizationwerereducedfromthersttofourthquarterofimplementationperiod. CONCLUSIONS: TherewerepositiveoutcomesofofferingpluralisticIMCservices. However,theoutcomeswouldbemoresignicantacrosstimeiftheIMCservices wereemphasizedmoretowardthosewhovisitedmorethanthreespecialists. HealthCareUse&PolicyStudies–HealthTechnologyAssessmentPrograms PHP83 HTADECISIONDRIVERSFORACCEPTANCEOFHIGHICERSUBMISSIONSAND REJECTIONOFLOWICERSUBMISSIONSTernouthAM1,JakelA2,PlestedM1,ModhaR1 1HERONEvidenceDevelopmentLtd,Luton,UK,2HERONEvidenceDevelopmentLtd,Stopsley,UKOBJECTIVES: Frequently,technologyappraisaldecisionsarebasedontheICERs estimatedfromeconomicmodels.Thestudyobjectivewastoinvestigatereasons providedbythreeHealthTechnologyAssessmentauthoritiesforacceptanceof highICERsubmissions(commonlyassumedthresholdabove£30,000)andrejection oflowICERsubmissions(commonlyassumedthresholdbelow£30,000). METHODS: AllpublishedtechnologyappraisalssinceApril2005weredownloaded fromPBAC,SMC,andCADTH,websites.Themanufacturer'sbase-caseICERswere extracted.DecisiondriversprovidedbytheagencieswereextractedfromthetechnologyappraisalswhichwererejecteddespitealowICERoraccepteddespitea highICER. RESULTS: TheCADTHaccepted14highICERsubmissions;71%owingto statedclinicaleffectiveness,including42%onthebasisofrestrictionand15% restrictedduration.ThePBACaccepted26highICERsubmissions;53%owingto statedcost-effectiveness,30%withrestriction,and19%witharisksharingagreement.TheSMCaccepted22highICERsubmissions;68%owingtorestriction,and 31%owingtoorphandrugstatus.TheCADTHrejected2lowICERsubmissions owingtoalackofdemonstratedclinicalbenet.PBACrejected48lowICERsubmissions;73%owingtolackofdemonstratedclinicalbenet,40%owingtouncertaintyintheeconomicmodel,andand27%owingtoanunacceptablyhighICER. TheSMCrejected62lowICERsubmissions;34%owingtoalackofdemonstrated clinicalefcacy,71%owingtolackofarobusteconomicanalysis,and31%owingto cost-effectivenessnotbeingdemonstrated. CONCLUSIONS: Asidefromademonstrationofclinicaland/orcost-effectiveness,highICERswereacceptedonthe basisofarestrictiontocertainpatientgroups,pricereduction,orowingtoorphan drugstatus.LowICERsubmissionswererejectedonthebasisofuncertainorlackingclinicalbenet,uncertainmodelestimates,oralackofappropriateanalysis. PHP84 AHRQVERSUSNICE:DOTHECONCLUSIONSINCERREPORTSCORRESPOND CLOSELYTOTHECOMPARATIVEEFFECTIVENESSASSESSMENTSMADEINHTA REPORTS?AlnwickK HERONEvidenceDevelopmentLtd,London,UKOBJECTIVES: Whilenon-USagenciessuchasNICE(NationalInstituteforHealth andClinicalExcellence)intheUKhavebeenconductingHealthTechnologyAssessments(HTAs)forovertenyears,ComparativeEffectivenessResearch(CER)as conductedintheUSisrelativelyrecent.Thelackofeconomicconsiderationsisone well-deneddistinctionbetweenCERandHTA.Theaimofthisresearchisto compareCERpublicationsfromAHRQ(AgencyforHealthcareResearchandQuality)withHTA-relatedpublicationsfromNICEtodeterminewhetherthereareother consistent,cleardistinctionsofnote. METHODS: All22AHRQCERpublications publishedontheNationalInstitutesofHealthwebsitewereassessed.Conclusions relatingtothecomparativeeffectivenessofpharmaceuticalsonmajorclinicaloutcomeswerecompiled.TheNICEwebsitewassearchedforcorrespondingHTA guidance,andconclusionsandotherfeaturesofthepublicationswerecompared. RESULTS: Ofthe14AHRQCERpublicationsthatassessedpharmaceuticals,foronly twooftheseweretherecorrespondingNICEHTAs.TheCERpublicationonRheumatoidArthritisandPsoriaticArthritiscorrespondedtotwoNICEHTAs,which werebothingeneralagreementwiththeCERreport.Incontrast,theconclusion fromtheCERpublicationonLipid-ModifyingAgentsstatedthattherewasinsufcientclinicalevidencetoguidedecisions,whereasNICEwascondentenoughin theevidencetomakeasubsequentaccessdecision.InitsClinicalGuideline(CG) documentsNICEalsoprovidedstatementsrelevanttothreeotherAHRQCERreport topics;allwereingeneralagreementwiththeAHRQCERconclusions.Themost notabledistinctionsbetweenCERandHTAreports,however,wereintheirscope, breadth,andpurpose. CONCLUSIONS: Wherelike-for-likecomparisonswerepossible,thisresearchfoundmoreagreementthandisagreementbetweenAHRQand NICE.Themajordistinctionsofnoterelatedtothetopicschosenforassessmentby thetwoagencies,whichreectdifferencesinprioritiesbetweenhealthsystems. PHP85 APPLYINGFUZZYMULTIPLECRITERIADECISIONMAKINGTOESTABLISHA NEWHEALTHTECHNOLOGYASSESSMENTSYSTEMWITHCOVERAGE IMPLICATIONSINTAIWANNATIONALHEALTHINSURANCELinCW I-ShouUniversity,Kaohsiung,TaiwanOBJECTIVES: Itbecameobviousthatallcountriesconsiderhealthtechnologyassessment(HTA)differentlyduetothestructureofindividualhealthsystems.The objectiveofthisstudywastoselectandestablishanew,properHTAsystemin TaiwanNationalHealthInsurance. METHODS: Asystematicliteraturereviewwas undertakentoidentifyrelevantpapersofeachofshorttandempolymorphism analysis,positronemissiontomography,photodynamiclasertherapyandvideoassistedthoracicsurgeryandthenusedfocusgroupdiscussiontoselectcategory'sA26VALUEINHEALTH14(2011)A1–A214

PAGE 35

aspect,evaluationcriteriaandschemesofHTAsystem.Basedonanalytichierarch process(AHP)methods,Wemakeinterviewswithhealthofcials,healthcareprovidersandcustomersbyusingtheAHPquestionnaire.Fuzzymultiplecriteriadecisionmaking(FuzzyMCDM)wasadoptedtoproceedtheempiricalevidenceanalysisandevaluationontheschemesofHTAsystems. RESULTS: Atotalof65 participantswereincludedinthissurvey.Threecategory'saspect,twelveevaluationcriteriaandthreeschemesofHTAsystemweregenerated.Forthecategory's aspect,thesequenceofimportanceisTomeettheexpectationsofhealthcare providers',thenTosatisfytherequirementsofcustomer'andthelastisToaccomplishthecoveragepolicyofgovernment'(0.375,0.337and0.238,respectively). Thedeviationofbalancingweightamongthesethreeaspectsislittleandthe importanceisverysimilar.TheeffectivescoreofHTAsystems,whichisevaluated andselectedbywholeinterviewers,thesequencefromlargetosmallissystemof safetyandefcacy',systemofethicalstandards',andsystemofcosteffectiveness'(67.18,61.41and58.01,respectively). CONCLUSIONS: Thesystemofcost effectiveness'isnotsuitabletoadoptatcurrentstage.ItdoesnotmeettheexpectationsfromhealthcarestakeholdersifthereisonlyoneHTAsystem.TheBureau ofNationalHealthInsuranceshouldselectsystemofsafetyandefcacy'asthe basistoestablishanew,properHTAsysteminTaiwan. HealthCareUse&PolicyStudies–PopulationHealth PHP86 IRRATIONALUSEOFANTIBIOTICSANDROLEOFPHARMACISTS:ANINSIGHT FROMAQUALITATIVESTUDYINNEWDELHI,INDIAKotwaniA1,WattalC2,JoshiPC3,HollowayK4 1V.P.ChestInstitute,UniversityofDelhi,Delhi,India,2SirGangaRamHospital,NewDelhi, India,3DelhiUniversity,Delhi,India,4WorldHealthOrganization,NewDelhi,IndiaOBJECTIVES: Antimicrobialresistance(AMR)isaglobalpublichealthproblem. AntibioticmisuseisoneoftheimportantfactorsforincreasingtrendofAMR.In developingcountrieslikeIndia,pharmacistsoftendispense"prescriptiononly" drugs,likeantibioticstopatientswhodonothaveaprescription.Itisimportantto understandthedispensingpracticesandbehaviourofcommunitypharmacistsin ordertodeveloppolicyinterventionsthatwilleffectivelyimprovetheuseofantibioticsatthecommunitylevel. METHODS: Focusgroupdiscussions(FGDs)were heldforvemunicipalwardsofDelhiwithretailpharmacists,publicsectorpharmacistsandtheofce-bearersofpharmacists'associations.Dataonantibioticuse andresistancewascollectedearlierinthesevewards.FGDs(n 3with40pharmacists)wereanalysedthroughgroundedtheory. RESULTS: FromtheFGDsfour broadthemesidentiedwere:prescribinganddispensingbehaviour;commercial interests;advisoryrole;andinterventionstrategiesforrationaluseofantibiotics whichwerefurtherdividedintosubthemes.FGDswithpharmacistsworkinginthe publicsectorfoundthat,besidesthefactorslistedabove,overstockandnearexpiry,andundersupplyofantibiotics,promotedantibioticover-useorunder-use. Findingselucidatespharmacists'roleinincreasingAMRthroughinappropriate dispensingpracticesandprot-drivenirrationalprescribing.Suggestionsforinterventionsfrompharmacistswere(1)educationtoincreaseawarenessaboutrationaluseandresistancetoantibioticsforpharmacists;(2)involvingthemaspartners tocreateawarenessamongcommunitiesforrationaluseandresistancetoantibiotics;and(3)developinganeasyreturnpolicyfornear-expiryantibioticsinpublic sectorfacilities. CONCLUSIONS: Inascenariowherepharmacistsarenotonly intermediatesbetweenprescribersandusers,butarealsoprescribinganddispensing,theybecomeimportantstakeholdersforinterventionprograms.Amultidisciplinaryteamofpolicymakersshoulddesignsustainableinterventionprogramsto improvepharmacists'antibioticprescribingbehavior,aswellasawarenessprogramsforcommunitymembers. HealthCareUse&PolicyStudies–PrescribingBehavior&TreatmentGuidelines PHP87 IMPACTOFPRESCRIPTIONMONITORINGPROGRAMSONCONTROLLED SUBSTANCEPRESCRIBINGBEHAVIORFreemanPR1,TalbertJ2,BlumenscheinK1 1UniversityofKentucky,Lexington,KY,USA,2UniversityofKentuckyCollegeofPharmacy, Lexington,KY,USAOBJECTIVES: Prescriptiondrugmonitoringprograms(PDMPs)havebeenimplementedbystatesasameanstoreduceabuseanddiversionofcontrolledsubstances(CS).Relativelylittleisknown,however,abouttheireffectivenessin achievingthesegoals.Thepurposeofthisprojectwastoassesstheperceived impactoftheKentuckyAllSchedulePrescriptionElectronicReportingprogram (KASPER)onCSprescribingbehaviorandtoevaluatetheperceivedeffectivenessof KASPERasatooltoreducedrugabuse,diversionanddoctorshopping. METHODS: Asampleoftwothousandprescribers,includingphysicians,advancedpractice nursesanddentists,wasrandomlyselectedfromalistofover4,000registeredwith KASPER.Theprescriberswerecontactedviamailtocompleteasurvey.Areminder postcardandsecondsurveywassenttonon-responderstwoweeksaftertherst mailing.Fivehundredfty-seven(557)responseswerereceivedfromprescribers foraresponserateofapproximately28%.Responseswerecodedanddescriptive analysiswasconductedinSTATAv11. RESULTS: Themajority(89%)ofprescribers respondingtothesurveywerephysicianswhoreportedusingameanof19.7 KASPERreportsmonthlyfortreatmentdecisions.89%ofrespondentsindicated thatinformationinthereportsimpactstheirtreatmentdecisions.One-halfof respondentsindicatednochangeinCSprescribingbehavioroverthepastyearasa resultofKASPERwhile35%indicatedadecreaseinCSprescribingand13%indicatedanincreaseinCSprescribing.96%ofrespondentsbelieveKASPERisaneffectivetoolforreducingabuseanddiversionwhile95%believeitisaneffectivetool toreducedoctorshopping. CONCLUSIONS: RespondentsthatuseKASPERregularly ndtheinformationinreportsvaluableformakingtreatmentdecisions.Overall, thevastmajorityofrespondentsperceiveKASPERasaneffectivetooltoreduce drugabuse,diversionanddoctorshopping.Furtherresearchisneededtodirectly assesstheeffectivenessofPDMPs. PHP88 INFLUENCEOFHOSPITALCHOICESOFDRUGSONTHEPHARMACEUTICAL CONSUMPTIONINTHECOMMUNITYGalliniA1,LegalR2,TabouletF1 1UniversityofToulouse-INSERM1027,Toulouse,France,2FrenchDepartmentofHealth, Paris,FranceHospitaldoctorsinuencegeneralpractitioners'prescribingpatterns.But,no studyhasfocusedontheinuenceofhospitalchoicesofdrugsonprescribing patternsinthecommunityforcompetitivepharmacologicalclassesatthemacro level. OBJECTIVES: Toinvestigatetheinuenceofhospitalchoicesofdrugsonthe pharmaceuticalconsumptioninitssurroundingcommunity. METHODS: Quasiexperimentalstudy.Datafromthenationalsurveyaboutdrugsinhospitalsofthe FrenchDepartmentofHealthwereusedtoextractpurchasedquantitiesin2008by 25ofthe29Frenchuniversityhospitalsforprotonpumpinhibitors(PPI),serotonin antagonists(setrons),lowmolecularweightheparins(LMWH),erythropoietins (EPO),angiotensinconversionenzymeinhibitors(ACEI),angiotensinIIreceptors antagonists(ARA),statins,alpha-adrenoreceptorantagonists(AAA)andselective serotoninreuptakeinhibitors(SSRI).Reimbursedquantitiesfromgeneralpractitioners'prescriptionsforpatientslivinginthehospital'scatchmentareawereextractedfromthenationalhealthinsurancedatabase(concerning90%ofthepopulation).Quantitiesforeachbrandandeachdrugentitywereexpressedindened dailydosesper1000hospitalizations-day(DHD)or1000inhabitants-day(DID)and log-transformed.Correlationsbetweenconsumptioninuniversityhospitalsand theircatchmentareaswereassessedusingSpearmanrankcorrelationcoefcients. RESULTS: Universityhospitalsselectedbetween30and52%oftheavailabledrug entitiesforthese9classes.Acorrelationcoefcientsuperiorto0.20wascalculated for45%ofbrandnamesand75%ofthedrugentities.Signicantcorrelationswere foundfortheclassesofSSRIandACEI.Forstatins,ARA,LMWHandsetrons,correlationwasquitegoodbutdidnotachievestatisticalsignicanceboundary.For theremainingclasses(PPI,EPOandAAA),correlationvariedwiththedrugs,but wasgloballynonexistent. CONCLUSIONS: Correlationsbetweenhospitalandcommunityusesvariedwithclassesanddrugs.Thedirectionoftherelationship,as wellasitsdeterminants,willbefurtherassessedinamultivariateanalysis. PHP89 THEIMPLEMENTATIONOFPUBLICHEALTHGUIDANCE:PREFERENCEFOR CLINICALOVERBEHAVIOURALINTERVENTIONSHamerslagL,KuselJ,HaynesS,Brooks-RooneyC,CostelloS CostelloMedicalConsultingLtd,Cambridge,UKOBJECTIVES: Guidanceonpublichealthtopicssuchascontraception,smokingand obesityisregularlyprovidedindevelopedcountries.Theseguidelinesonsocial behaviourhaveatendencytobelessdistinctthanthoseforexplicitclinicalconditions,astheyoftenrecommendinterventionssuchassupportandaccessto informationratherthanspecicdrugsorsurgicalprocedures.Theaimofthisstudy wastocomparetheimplementationofbehaviouralguidancetothatofguidance recommendingclinicalinterventionsinvariouspublichealthtopics. METHODS: AllreportsontheimplementationofpublichealthguidanceintheUKthatwere listedinJuly2010bytheNationalInstituteforHealthandClinicalExcellence(NICE) werereviewedandcompared. RESULTS: Sixreportswerepublishedontheimplementationofpublichealthguidance.Thesefocusedonattentiondecithyperactivitydisorder,anxiety,long-actingreversiblecontraception,secondarypreventionofmyocardialinfarctions,obesityandsmokingcessation.Implementation wasmeasuredbyeitherthePrescriptionCostAnalysissystem,theelectronicPrescribingAnalysisCostToolorthroughanalysisofanonymisedpatientrecords. Generally,guidanceforclinicalrecommendations,suchasdrugsorsurgicalprocedures,wasaccuratelyadheredtointhetopicsassessed.Conversely,supportive recommendations,suchasaccesstoinformationandcounselling,displayedvery limitedimplementation.However,thesevariableswereverydifculttoassess throughthecurrentinformationsystemsavailable. CONCLUSIONS: Forpublic healthguidance,itisoftendifculttofollowuponhowwelltheadviceisimplementedthroughoutsocietyduetomethodologicaldifcultiesinmeasuringsupportiveinterventions.However,thereappearstobeacleartrendtowardsthepoor implementationofbehaviouralguidancecomparedtospecicrecommendations fordrugsandsurgicalprocedures.Inpublichealthtopics,whereguidanceisoften indistinct,itisimportantthatsupportiverecommendationsareemphasised enoughforthemtobeimplementedalongsideclinicalinterventions. PHP90 TRENDSINUTILIZATIONOFPREGNANCYRISKCATEGORYDANDXDRUGS AMONGPREGNANTWOMENINTHEUNITEDSTATESShahAJ,PatilDS,KadakiaA,AparasuRR UniversityofHouston,Houston,TX,USAOBJECTIVES: TheUSFoodandDrugAdministration(FDA)hasclassieddrugsused inpregnancyintovecategoriesnamelyA,B,C,DandXwithcategoryAbeingthe safestandcategoryXbeingabsolutelycontraindicatedinpregnancybasedonfetal riskassociatedwiththesemedications.TheobjectiveofthestudywastodeterminethetimetrendandpredictorsofFDA'scategoryDandXdrugsamongoutpatientvisitsinpregnantwomen. METHODS: Aretrospectivestudyusingthe1997-98A27VALUEINHEALTH14(2011)A1–A214

PAGE 36

and2005-06NationalAmbulatoryMedicalCareSurveyandtheoutpatientdepartmentcomponentoftheNationalHospitalAmbulatoryMedicalCareSurveywas conductedtoexaminematernalcharacteristicsandphysiciancharacteristicsassociatedwithCategoryDandXprescriptions.Descriptivestatisticsandlogistic regressionanalysiswereperformedtodeterminethetimetrendandpredictors associatedwithprescribingcategoryDandXdrugs. RESULTS: Approximately35 millionvisitsweremadebypregnantwomenannuallyduringtheseyears.Ofthese visits,1.4%(95%CondenceInterval(CI):1.13-1.77%)ofthevisitsinvolvedprescriptionofCategoryDdrugsand1.1%(CI:0.6-1.4%)ofthevisitsinvolvedprescriptionof CategoryXdrugs.Asignicantpositivetimetrendwasobservedinprescribingof categoryDandXdrugsin2005-06(OR-2.14,95%CI:1.23-3.73)comparedto1997-98. Pregnantwomenabovetheageof30hadgreaterodds(OR-1.62,95%CI:1.08-2.44)of receivingthesedrugsthanpregnantwomenbelowtheageof30.Pregnantwomen whoreceived 2medicationshadgreaterodds(OR-15.31,95%CI:9.02-26.01)ofreceivingthesedrugsthanwhoreceived 1medication.Signicantdifferences acrossphysicianspecialtyandracewerealsoobserved. CONCLUSIONS: BothmaternalandphysiciancharacteristicsplayaroleinprescribingofCategoryDandX drugs.WithincreaseinuseofcategoryDandXdrugs,moreresearchisneededto evaluatetheclinicalconsequencesoftheseprescribingpractices. HealthCareUse&PolicyStudies–QualityofCare PHP91 ASSOCIATIONBETWEENQUALITYOFCAREANDSHORTTERMOUTCOMESFOR VERYOLDPATIENTSHOSPITALIZEDFORACUTEILLNESSESChongWF1,DingYY2,SunY1,HengBH1 1NationalHealthcareGroup,Singapore,2TanTockSengHospital,SingaporeOBJECTIVES: Todeterminetheassociationbetweenqualityofhospitalcare,using selectedAssessingCareofVulnerableElders(ACOVE)qualityindicators,andshort termoutcomesforveryoldpatientsadmittedforacuteillnessesinatertiary hospital. METHODS: Thisisaretrospectivereviewofmedicalrecordsofarandom sampleofallpatientsaged80yearsandaboveadmittedtotheDepartmentof GeneralMedicineandDepartmentofGeriatricMedicinefrom2005to2008,withan equalnumberofcases(750)selectedforeachyear.Thequalityindicatorsselected werecognitiveandfunctionalassessment,dischargeplanning,deliriumevaluationandtreatment,mobilization,andaspirationprecautions.Ifpatientsmetthe inclusioncriteriaforeachspecicqualityindicator,theywereassessedtodetermineiftheprocess-of-carewascarriedoutanddocumented.Theoutcomesof interestwereinhospitalmortality,30-daymortalityandhospitalreadmission within15daysofdischarge. RESULTS: Fromthereviewof2,923cases,theadherenceforthequalityindicatorsrangedfrom33.8%forcognitiveandfunctional assessmentto88.1%fordischargeplanning.Inunivariateanalyses,adherenceto deliriumevaluationandtreatmentwasassociatedwithareductionininhospital (17.5%vs26.1%,p 0.004)and30-daymortalityrates(22.6%vs32.5%,p 0.002). Documentationofcognitiveandfunctionalassessmentwasassociatedwithareductionininhospital(6.2%vs16.2%,p 0.005)and30-daymortalityrates(10.3%vs 22.3%,p 0.005).Theassociationbetweenadherencetoqualityindicatorsandhospitalreadmissionwasnotstatisticallysignicant.Afteradjustmentsusingthe biovariateprobitregression,adherencetoaspirationprecautionswasassociated withlowerinhospitalmortality(rho -0.128(95%CondenceInterval: 0.233,-0.021). CONCLUSIONS: Adherencetoselectedprocess-of-carequalityindicatorswasgenerallylow.Adherencetoaspirationprecautionswastheonlyqualityindicatorthathadanimpactontheshorttermoutcomeforelderlypatients hospitalizedforacutemedicalconditions. PHP92 RESULTSOFAPHARMACYMANAGEMENTPROGRAMFORIMPROVING MEDICATIONADHERENCERegineML,BunzTJ CIGNAHealthCare,Bloomeld,CT,USAOBJECTIVES: CIGNA'sCoachRxpharmacymanagementprogramprovidesservices topharmacycustomerswithmedicationadherenceand/ormedicationsafetyissues/concerns.Inthisstudy,telephonicengagementwithaCoachRxpharmacist initiatedtheprocess.Theaimofthisanalysiswastoevaluatetheeffectofthe CoachRxprogramonmedicationadherence;useofmail-orderpharmacyservices; utilizationofgeneric,preferred,andnon-preferred-brandmedications;andingredientcosts. METHODS: Apre/postcomparisonspanningoneyear(6mopre/6mo post)wasconducted.Adherencewasmeasuredatthecustomerlevel.Continuouslyenrolledindividualswereidentied,andthefollowingdemographicinformationwascollected:age,sex,andemployer.Medicationpossessionratiowas usedtomeasureadherenceforveclassesofmedications:asthma,depression, diabetes,dyslipidemia,andhypertension.Useofmailorderpharmacyproducts, gaps-in-caremeasures,andpatient-requested,targetedmedicationswitcheswere evaluatedatthemedicationlevel. RESULTS: Acrosstheclassesofmedications, fouroftheveshowedanincreaseinadherencefollowingenrollmentinCoachRx. Amongthecustomersubgroupsub-optimallyadherent(MPR 80%)priortoenrollinginCoachRx,therewasastatisticallysignicantincreaseinadherenceforall vemedicationclasses.Compliancetoadherencerelatedgaps-in-careincreased by4%andthenumberoftreatmentslledsolelythroughmail-orderpharmacy increasedby11%followingenrollment.Therewasa14.4%decreaseintheshareof non-preferredbrandmedications,withnochangeinpreferredbrandmedications anda2.2%increaseingenericmedications.Amongthestudyparticipants,there were40conrmedpatient-requestedswitchesfromtargetedmedications,withan averageingredientcostsavingsof$661perswitchperyear. CONCLUSIONS: The maingoalofCoachRx,todriveappropriateuseofpharmacotherapy,wasachieved inthesix-month,post-enrollmentperiod.Improvementincareoversuchashort periodisapositivending.Theseimprovementsmayhaveanevengreaterimpact onfuturemedicalcostsandmayleadtocontinuedreductionsinmedicationcosts. PHP93 ANEXAMINATIONOFHOSPITAL-ACQUIREDCONDITIONSBYAGEANDPAYER INFIFTEENSTATES,2008MaedaJL1,ParlatoJ2,LevitK1,AndrewsRM3,JiangHJ3 1ThomsonReuters,Washington,DC,USA,2ThomsonReuters,Cambridge,MA,USA,3Agencyfor HealthcareResearchandQuality(AHRQ),Rockville,MD,USAOBJECTIVES: TheCentersforMedicareandMedicaidServices(CMS)hasidentied severalpreventableadverseevents,knownashospital-acquiredconditions (HACs),thatinmostcasesCMSwillnotallowadditionalreimbursementtohospitals.HACsarecommonorcostlyconditionsthatareacquiredduringthehospital stayandcouldhavebeenpreventedthroughtheapplicationofevidence-based guidelines.CMShasidentiedtencategoriesofHACsthataredeemedtobepreventablecausesofmorbidityandmortality.Theobjectiveofthisstudywasto documenttheprevalenceofHACsforall-payersandtodifferentiatetheimpactof HACsbypayerandagegroup. METHODS: Thestudydesignwasaretrospective cohortfromthe2008HealthcareCostandUtilizationProject(HCUP)StateInpatient Databases(SID)les.WeusedtheCMSdenitiontoidentifyHACsfromthepresent-on-admission(POA)indicatorassociatedwithsecondarydiagnosesintheSID. TheSIDaretheonlymulti-stateall-payerdataavailablethatincludethePOA indicatorneededtoidentifyHACs.Westudiedadult,non-maternal,medical-surgicalorselectedsurgicaldischarges,dependingontheHAC,fromcommunity, non-rehabilitationhospitalsinfteenstatesthatreportedPOAindicatorstoHCUP. RESULTS: FallsortraumawasthemostprevalentHACinoursample.Forselected surgicalprocedures,deepveinthrombosis/pulmonaryembolismhadthehighest rateofHACs(832per100,000hiporkneereplacementdischarges).Medicaiddischargeshadabouttwicetherateofmanifestationsofpoorglycemiccontroland morethantwicetherateofselectedsurgicalsiteinfectionsrelativetoMedicare discharges.FortheconditionsofstageIIIandIVpressureulcers,fallsandtrauma, catheter-associatedurinarytractinfection,anddeepveinthrombosis/pulmonary embolismtherelativeriskofacquiringaHACincreasedwithage. CONCLUSIONS: TherateofHACsdifferedbypayerandagegroup.Thus,itisimportanttoconsider examiningHACsbyallthird-partypayers. HealthCareUse&PolicyStudies–RegulationofHealthCareSector PHP94 THEROLEOFHEALTH-RELATEDQUALITYOFLIFEDATAINTHEDRUG APPROVALPROCESSESINTHEUNITEDSTATESANDEUROPE:AREVIEWOF GUIDANCEDOCUMENTSANDAUTHORIZATIONSOFMEDICINALPRODUCTS FROM2006TO2010MarquisP1,CaronM2,EmeryMP2,ScottJ3,AcquadroC2 1MAPIValues,Boston,MA,USA,2MAPIResearchTrust,Lyon,France,3MAPIValues, Maccleseld,UKOBJECTIVES: TheobjectiveofthisresearchwastoreviewtheextenttowhichHRQL andotherPROshaveplayedaroleindrugapprovalandlabelingsincetheFDA issueditsdraftguidanceontheuseofPROmeasuresandtheEMA,itsreection paperonHRQLin2006. METHODS: Thisresearchwasconductedthroughasystematicmanualreviewoftherapyarea-specicregulatoryguidelines(USandEU) andproductlabelingsissuedduringtheperiodofJanuary1,2006toNovember16, 2010. RESULTS: Fifteenand34guidancedocumentswerereleasedbytheFDAand theEMArespectively,containingrecommendationsfortheinclusionofPROendpointsinclinicaltrials.TheFDAreferredtoHRQL(asasecondaryendpoint)inthree ofthe15(20%)guidance.TheEMArecommendeduseofHRQLendpointsin22of the34(65%)guidance.TheFDAapproved93productswithlabelclaimsthatincludedPROendpoints(outof432totalapprovals).Ofthose,8products(8.6%ofall productswithaPROclaim)documentedtreatmentbenetscharacterizedasHRQL. TheEMAapproved54productsthatincludedPROendpoints(outof248totalapprovals),ofwhich16products(29.62%ofallproductswithaPROclaim)reected HRQLdata. CONCLUSIONS: OurreviewshowedthatpatientsperspectiveinclinicalresearchisimportantfortheEMAandFDA,withHRQLendpointsstillplaying aminorrole.OuranalysissuggeststhatEMA'sreceptivitytoHRQLendpointsis greaterthanFDA's,moresensitivetopatientsymptoms-baseddata.Forthefuture, webelievethatpatients'involvementshouldbeextendedtogetamoreprecise ideaofthePROsrelevanttothem.Otherkeyplayersinvolvedinreimbursement andpricingdecisionsmighthavetheirsaysincenancialhealthresourceswill becomescarceandthatassessingvalueformoneywillbemorecrucial. PHP95 EXPECTEDCHANGESINHEALTHCAREUTILIZATIONDUETOPATIENT PROTECTIONANDAFFORDABLECAREACTGhushchyanVH1,SullivanPW2,NaimA3,NairKV1 1UniversityofColorado,Denver,Aurora,CO,USA,2RegisUniversitySchoolofPharmacy, Denver,CO,USA,3CentocorOrthoBiotechServices,LLC,Horsham,PA,USAOBJECTIVES: OnMarch23,2010thePatientProtectionandAffordableCareAct (PPACA)wassignedintoalawprovidinghealthinsurancecoverageforapproximately32millionuninsuredAmericans.Theobjectiveistodeterminetheexpected changesinhealthcareutilizationfortheuninsuredafter2014whenkeyprovisions ofPPACAlawbecomeeffective. METHODS: MedicalExpenditurePanelSurveydata (2008)wasusedfortheanalysis.Thedifferencesinhealthcareutilizationbetween theuninsuredandthosewithprivateorMedicaidcoveragewasestimatedusingaA28VALUEINHEALTH14(2011)A1–A214

PAGE 37

negativebinomialregressionmodel,controllingforage,gender,race/ethnicity, incomeeducation,generalphysicalandmentalhealthandco-morbidityburden.In 2014dependingonfamilyincomeandemploymentstatus,theuninsuredwilleitherobtainprivateorMedicaidinsurancecoverage,thustheanalysiswasrestrictedtoindividualslessthan65yearswith12monthsofcontinuousprivateor Medicaidconvergeoruninsuredforthewholeyear.Ourstudysamplewasnationallyrepresentativeonbehalfof71.3%ofUSpopulation. RESULTS: Fivemeasuresof healthcareutilizationwereused(emergencyroom(ER)visits,outpatientvisits, ofcevisit,inpatientvisits,prescriptionuse).Uninsuredindividualshadlower utilizationforallhealthcareservicesexceptERvisits.Holdingeverythingelse constant,theuninsuredwillhave1.98(1.75-2.25)and1.61(1.24-2.1)timeshigher expectedrateofofcebasedvisits,2.39(1.81-3.15)and2.62(1.41-4.86)times higherexpectedrateofoutpatientvisits,2.17(1.58-2.97)and1.70(1.11-2.62) timeshigherexpectedrateofinpatientvisits,1.70(1.53-1.89)and1.92(1.57-2.34) timeshigherexpectedrateofprescribedmedicationuseafterobtainingprivate orMedicaidcoverage,respectively. CONCLUSIONS: Healthcarereformwillincreasethedemandforhealthservicesandprescribedmedications,exceptERuse fortheuninsured.Theseresultsmaybeusedbyvariousstakeholderstoestimate expectedchangesinhealthcareexpenditures. PHP96 REASONSFORREJECTIONOFPROLABELCLAIMS:ANANALYSISBASEDONA REVIEWOFPROUSEAMONGNEWMOLECULARENTITIESANDBIOLOGIC LICENSEAPPLICATIONS2006-2010DeMuroC1,ClarkM2,MordinM2,EvansE1,Copley-MerrimanK2,FehnelSE1, GnanasakthyA3 1RTIHealthSolutions,ResearchTrianglePark,NC,USA,2RTIHealthSolutions,AnnArbor,MI, USA,3Novartis,EastHanover,NJ,USAOBJECTIVES: PreviousanalysesofPROlabelclaimsconcentratedonlyonsuccessfullabelclaims.ThegoalofthisresearchwastoexplorethereasonswhyPROlabel claimswereeitherdeniedornotsought. METHODS: UsingtheFDADrugApproval ReportWebpage,allapprovednewmolecularentities(NMEs)andbiologiclicense applications(BLAs)betweenFebruary2006andDecember2010wereidentied.For identieddrugproducts,medicalreviewsectionsfrompubliclyavailablesummary basisofapprovals(SBAs)werereviewedtoidentifyPROendpointstatusandany FDAStudyEndpointsandLabelDevelopmentcomments. RESULTS: Outofthe116 NMEs/BLAsidentiedandaccompanyingSBAsreviewed,44.8%ofproductsincludedPROsaspartofthepivotalstudies;however,only24.1%receivedPRO claims.Primaryreasonsfordenial(wheredataavailable)includedalackofdemonstrationofcontentvalidity(inclusiveofgeneralmeasuressuchastheEQ5Dand SF-36)aswellasuseofPROstoassesssymptomsinanopen-labelsetting,lackof consensusonclinicallymeaningfulchange,interpretationoformissingPROdata, lackofmeasurementoffullconstellationofsymptoms,issuesofmultiplicityand concernsof"bias"incertainPROmeasures. CONCLUSIONS: Nearlyhalf(45%)of submissionsincludedPROsyetthisrateisnotreectedbyclaimsgranted.UnderstandingthenatureofPROclaimsgrantedunderthecurrentregulatoryguidanceis important.Additionally,aclearunderstandingofclaimsdeniedyieldsvaluable insightintowheresponsorsmayimproveimplementationofPROsinclinicaltrials andthePROevidencesubmittedinordertoincreasethelikelihoodofobtaining PROlabelclaims. PHP97 PROLABELCLAIMS:ANANALYSISBASEDONAREVIEWOFPROSAMONGNEW MOLECULARENTITIESANDBIOLOGICLICENSEAPPLICATIONS2006-2010MordinM1,ClarkM1,DeMuroC2,EvansE2,Copley-MerrimanK1,FehnelS2, GnanasakthyA3 1RTIHealthSolutions,AnnArbor,MI,USA,2RTIHealthSolutions,ResearchTrianglePark,NC, USA,3NovartisPharmaceuticals,EastHanover,NJ,USAOBJECTIVES: Wilkeandcolleagues(2004)previouslyconductedareviewofeffectivenessendpointsreportedinthelabelsofnewdrugproductsapprovedinthe UnitedStates(US)between1997-2002todeterminetheextentandtypeofPRO endpointsutilized.Theyreportedthat30%ofproductlabelsreviewedincluded PROs.Ourstudyaimedtobuilduponthisworkbydescribingthecurrentstateof PROlabelclaimsgrantedfornewmolecularentities(NMEs)andbiologiclicense applications(BLAs)followingreleaseofthedraftandnalFDAPROGuidancedocuments(i.e.,sinceFebruary2006). METHODS: UsingtheFDADrugApprovalReportswebpage,allFDAapprovedNMEsandBLAsbetweenFebruary2006andDecember2010wereidentied.Genericproductswithtentativeapprovalsgrantedin thisperiodwereexcluded Forallidentieddrugproducts,medicalreviewsections frompubliclyavailablesummarybasisofapprovals(SBAs)werereviewedtoidentifyPROendpointstatus.Productlabels(indication,clinicaltrialssections)were reviewedtodeterminethenumberandtypeofPROclaims. RESULTS: Ofthe116 NMEs/BLAsidentied,28(24.1%)weregrantedPROclaims.Themajority(n 24) wereforsignsandsymptoms.Nineofthesignsandsymptomclaimswerepainrelated.Ofthe28productswithPROclaims,aPROwasaprimaryendpointfor20 (71%).All20oftheseprimaryendpointsweresymptom-relatedandthemajority (12of20)werecollectedviadiary. CONCLUSIONS: PROclaimscontinuetobeapprovedbyFDA,with24%ofNMEsandBLAsgrantedPROclaims.SuccessfulPRO labelclaimsoverthepastveyearshavebeenlargelyinsupportoftreatment benetforsymptomsspeciedasprimaryendpoints.TheproportionofNMEswith PROlabelclaimsduringthepost-guidanceperiod(24.1%)waslowerthanthatofthe pre-guidanceperiod(30%). PHP98 ATRENDANALYSISOFNEWMOLECULARENTITIESWITHDRAWNFORSAFETY REASONSFROM1980TO2009INTHEUNITEDSTATESSeoane-VazquezE1,Rodriguez-MonguioR2,QureshiZP3,SzeinbachSL4 1MassachusettsCollegeofPharmacy&HealthSciences,Boston,MA,USA,2Universityof Massachusetts,Amherst,,MA,USA,3UniversityofSouthCarolina,Columbia,SC,USA,4Ohio StateUniversity,Columbus,OH,USAOBJECTIVES: Besidestheinuenceofeconomicfactors,prescribingpatterns,and marketdynamics,decisionstowithdrawproductsfromthemarketaredrivenby concernsoversafety.Thisstudyevaluatednewmolecularentities(NMEs)approvedbytheFDAintheperiod1980-2009thatwerewithdrawnfromthemarket forsafetyreasons. METHODS: DatawereobtainedfromtheFDAandtheUSFederal Register.Descriptiveanalyseswereusedtoclassifyproductdiscontinuationsby therapeuticcategory,year,andreasonfordiscontinuation. RESULTS: Therewere 740NMEsapprovedbytheFDAduringthestudyperiod.AsofDecember1,2010,the numberofdrugsdiscontinuedwas118(15.9%).Safetywasthereasonforwithdrawing27(3.6%)drugsfromthemarket.Therapeuticcategorieswiththemostsafety withdrawalsasapercentageofapprovalsinthe1980swererespiratory(28.6%), musculo-skeletal(23.1%),andnervoussystem(7.4%).Duringthe1990s,classes withthemostsafetywithdrawalsasapercentageofapprovalsweremusculoskeletal(18.8%),alimentarytractandmetabolism(12.0%),andbloodandblood formingorgans(7.7%).Therapeuticcategoriesaffectedbysafetywithdrawalsasa percentageofapprovalsinthe2000sweremusculo-skeletal(20.0%),alimentary tractandmetabolism(4.2%),andantineoplasticandimmunomodulatingagents (3.2%).Majorproblemsthatspurredsafetywithdrawalwerehepatictoxicity,severecardiovasculareffects,andgastrointestinalissues.Averagetimefromapprovaltosafetywithdrawalwas5.9(SD 5.0)years,witharangeof0.3-18.2years, anda95%CIof4.0-7.8years. CONCLUSIONS: ApproximatelyoneinsevenNMEs approvedintheperiod1980-2009wasdiscontinuedfromthemarket.Lessthan one-quarterofthediscontinuationswereattributedtosafetyreasons.Products remainedinthemarketforanaverageofsixyearsbeforesafetywithdrawal.An ongoingevaluationofnewdrugsthroughtheirproductlifecycleisimportantto determinetheirlong-termsafetyandvaluetosociety. PHP99 WILLBIOPHARMACEUTICALINNOVATIONSTILLBEAPICTUREOFHEALTH AFTERIMPLEMENTATIONOFHEALTHCAREREFORM?RossKD UniversityofWashington,Seattle,WA,USAOBJECTIVES: Todeterminewhatimpactshorteningorlengtheningthedataexclusivityperiod(DEP)forbiologicdrugshasoninnovation.Asapartofthis,thegoalis todeterminewhateffectsoninnovationthe12-yearDEPincludedinhealthcare reformwillhave. METHODS: AsimulationmodelisdevelopedtoassesstheprofitabilityofcandidatedrugsundervaryingDEPs.Allcostsandrevenuesarediscounted.Thedrugsarethengroupedinto10-drugportfoliosandtheprotabilityof eachportfolioisdetermined.Thepercentageofportfoliosthatareprotableunder eachDEPlengthisdividedbythepercentageofportfoliosthatareprotableunder aindeniteDEPtogivearelativelevelofinnovation. RESULTS: ADEPof0years yieldsa60%decreaseinthelevelofinnovationandtherearenoincreasesin innovationforDEPsabove34years.ForaDEPof12years,thereisanexpected8.1% decreaseinthelevelofinnovation. CONCLUSIONS: The12-yearDEPimplemented asapartofhealthcarereformislikelytodecreaseinnovationinbiologicdrugs.The expected8.1%decreaseininnovationmayormaynotbeworththeexpecteddecreaseinpricesoncebiosimilarcompetitorsenter.Themodelalsoindicatesthat therewouldbenoreturnstoinnovationbyincreasingDEPabove34years,andas such,itislikelythatthiswouldrepresentamaximumwhenselectingaDEP. PHP100 PRICINGANDREIMBURSEMENTOFORPHANDRUGSINCANADAKumarJ,BachmanEM HERONEvidenceDevelopmentLLC,Somerville,NJ,USAOBJECTIVES: TheCanadianOrganizationofRareDiseases(CORD)denesarare diseaseasonethatafictslessthat1personin200000.Signicantmarketaccess andpricingchallengesexistforODsinCanadabothatafederalandtheprovincial level.ThescopeofthisstudyistodescribetheODsregulationsinCanada,evidence requirementsbythenationalregulatoryagency,nationalandregionalfunding criteria,marketaccesschallengesassociatedwithODs,andapproachestoobtain accesstoODsinCanada. METHODS: Non-systematicPubMedsearch,HealthCanada,theCanadianAgencyforDrugandTechnologyinHealth(CADTH),Common DrugReview(CDR),CanadianOrganizationofRareDiseases(CODR)anddifferent provincesMinistriesofHealthwebsites. RESULTS: HealthCanadareviewsODsto ensurethatthedrugmeetsthecriteriaofefcacy,safety,andmanufacturingquality.TheCDRconductstheclinicalandcosteffectivenessreviewcomparedtoexistingtherapiesandmakespositiveornegativerecommendationstoprovincesto listODsintheirrespectiveformularies.Atthefederallevel,pricingofODsisregulatedbyPatentedMedicinePricingReviewBoard(PMPRB).Attheprovinciallevel, differentprovincescanmaketheirownindependentreimbursementdecisionirrespectiveofCDR'srecommendation.DuetothelargebudgetimpactofODs,most provincesdonotprovideaccess.Thespecializedaccessmechanism,criteriafor eligibility,extentofcoverage,anddifferentdatarequirementtoobtainaccessin threeimportantprovincesofCanada(Ontario,Alberta,andQuebec)willbediscussedintheposter. CONCLUSIONS: Intheabsenceofanationalorphandrug policy,patientssufferingfromrarediseasesfacechallengesinobtainingaccessto ODsinCanada.However,signicantopportunitiesexistformanufacturerstoprovideaccesstoODsinCanada.A29VALUEINHEALTH14(2011)A1–A214

PAGE 38

HealthCareUse&PolicyStudies–RiskSharing/Performance-BasedAgreements PHP101 VALUEBASEDPRICING:WHATISTHEFUTUREFORPATIENTACCESS SCHEMES?LovemanCM,MallinsonM,AreteouT,WhiteR DoubleHelixConsulting,London,UKOBJECTIVES: Inthefaceofuniversalbudgetrestrictions,pricingofpharmaceuticalsremainsadivisivesubject.TheUSgovernmentisincreasingitsinvolvementin healthcareprovisionandinsurerswillhavemorebudgetpressuresduetoincreasedcoverage.IntheUK,thegovernmenthasproposedavaluebasedpricing approachwhichwillpotentiallyincreasetheavailabilityofnewdrugs.Theobjectiveofthisresearchistobetterunderstandthefutureoutlookforpatientaccess schemes(PAS)inachievingaffordablebudgetimpactwhilstnotrestrictingpatient access. METHODS: 40telephoneinterviewswereundertakenwithpayersinthe UK,USandTaiwanrepresentinggovernmentagencies,HTAgroupsandinsurers. WeaskedintervieweesabouttheirperceptionsofPASpoliciestoproviderequired valueevidence,improveaffordability,strategiesthatcouldbeusedbymanufacturersandlikelyfuturechanges. RESULTS: Resultsdemonstratedthatmanufacturerswillaimforaparticularpriceorpricethreshold;ifthisisnotacceptedby payers,outcome-basedschemesorotherexiblepricingarrangementswouldbe proposedformanufacturerstoestablishthattheirproductiscost-effectiveandto achieverevenuetargets.UKpayerswillnditdifculttodeterminethelevelof weightingappliedtodifferentunmetneedsortoassignvariousrewardsonvalue. RespondentsstatedthatPAScouldbeamethodusedtodeterminethelevelof valuethroughmonitoringoutcomes,butadministrativeburdensneedtobeconsidered.IntheUS,marketevolutionwillacceleratetheuptakeofpatientaccess schemes.TaiwanislikelytointroducePASapproachesasameanstoreduceexpenditureonhighcostdrugs. CONCLUSIONS: Thecurrentbudgetconstraintsin thedifferentcountriesmayleadtoanincreasedinterestinimplementingPAS, whichinturnmaychangethewayhealthsystemslookatvalueforpatients. PHP102 WHICHCRITERIAAREUSEDINHEALTHCAREDECISIONMAKINGAND PRIORITYSETTING?ALITERATUREREVIEWFORANINTERNATIONALSURVEY OFDECISIONMAKERSGuindoLA1,WagnerM2,BaltussenR3,RindressD2,VanTilJ4,KindP5,GoetghebeurM2 1MScCandidate,Montreal,QC,Canada,2BioMedComConsultantsInc.,Dorval,QC,Canada,3RadboudUniversity,Nijmegen,TheNetherlands,4UniversityofTwente,Enschede, TheNetherlands,5UniversityofYork,York,UKOBJECTIVES: Resourceallocationisoneofthemostchallengingissuesfacedby healthpolicydecision-makersrequiringefcientconsiderationofmanyfactors. Objectivesofthisstudyaretoidentifycriteriausedindecisionmakingaroundthe world. METHODS: AnextensiveliteraturesearchwasperformedinMedlineand EMBASEtoidentifyarticlesreportingdecisioncriteria.Bibliographiesofrelevant articleswerealsosearched.Studiesconductedwithhealthcaredecisionmakers (e.g.,empiricalstudies,eld-testingofdecisionmakingtools,focusgroups,questionnaires,interviews),conceptualandreviewsarticlesaswellasarticlesdescribingdecisionmakingtoolswereincluded.CriteriareportedwereextractedandorganizedusingaclassicationsystemderivedfromtheEVIDEMframework. RESULTS: Atotalof2903recordswereidentiedthroughdatabasesearchingand 243additionalrecordswereidentiedthroughbibliographichandsearching;of these2790wereexcluded.356articleswereassessedforeligibilityand40articles wereincludedinthestudy.Largevariationsinterminologyusedtodenedcriteria wereobservedand338differenttermswereidentied.Thesewereassignedto58 uniquecriteriawhichwereclassiedin9differentcategoriesincluding:1)health outcomesandbenetsofintervention,2)typesofhealthbenet,3)impactof diseasetargetedbyintervention,4)therapeuticcontextofintervention,5)economicimpactofintervention,6)quality/uncertaintyofevidence,7)implementationcomplexityofintervention,8)priority,fairnessandethics,9)overallcontext. Themostfrequentlymentionedcriteriawere:equity/fairness(33times),efcacy/ effectiveness(28),healthcarestakeholderinterestsandpressures(28),cost-effectiveness(23),strengthofevidence(20),safety(19),missionandmandateofhealth system(17),need(16),organizationalrequirementsandcapacity(17)andpatient reportedoutcomes(16). CONCLUSIONS: Thedatasynthesizedinthisstudywill serveasthebasisfordevelopmentofaninternationalsurveyofhealthcaredecisionmakers.Theultimateobjectiveistodevelopmulticriteriaapproachestoefcienthealthcaredecisionmakingandprioritysetting. PHP103 COMPARATIVEEFFECTIVENESSREVIEWSANDTHEIRPOTENTIALIMPACTON FORMULARYACCESSINTHEUNITEDSTATESXiaAD HERONEvidenceDevelopmentLtd,London,UKOBJECTIVES: TheEffectiveHealthCareProgramoftheAHRQ(AgencyforHealthcare ResearchandQuality)hassponsoredvariouscomparativeeffectivenessreviews.The outcomesofthesereviewswereexaminedtodeterminetheirimpactonaccessof drugsintheUSandpotentialfutureimpactastheroleofHTAbecomesincreasingly important. METHODS: Initially,weexaminedtheAHRQonlinedatabaseandcompiledalistofconclusionsfromcompletedcomparativeeffectivenessreviewsinvarioustherapyareas.Then,wecomparedtheseconclusionstothecurrentaccessof thesetherapiesinaselectionofthelargestUSplansbylinescovered(Aetna,UnitedHealth,Cigna,Kaiser,Wellpoint)usingtheironlineformularydatabases. RESULTS: ACEIsvs.ARBs,NSAIDs,anti-depressants,epoetins,PPIsvs.H2As,andstatinswere reviewedbytheAHRQ.Someoftheseresultswereeitherinconsistentorinconclusive duetolackofevidence;however,severalfoundefcacyandsafetytobesimilaracross agentsinaclass.Asexpected,accesstotheseproductswithinthelargestUSplansare alsocomparable.Incertaininstances,aspecicproductorclassdemonstratedsuperiorefcacyortolerability.PPIsshowedgreaterefcacythanH2Asinresolutionof reuxdisease,andmorespecically,esomeprazoledemonstratedsuperiorefcacy overomeprazoleforreliefofsymptomsat4weeks.TopicalNSAIDs,suchasdiclofenac,demonstratedcomparableefcacytooralswithfewertolerabilityissues.However,thesesuperioritiesarenotalwaysreectedintheformularyaccessofthese productsintermsoftierplacementorrestrictions. CONCLUSIONS: Itappearsthat thesecomparativeeffectivenessreviewsbytheAHRQhavesomeindirectimpacton formularyaccessintheUS.However,priceandcontracting,inadditiontoefcacyand safetyareamongthekeydeterminantsforplans.Itwillbenecessarytocontinue monitoringthesereviewsmovingforwardsandattempttolterouttheirdirectimpactonaccessofdrugsovertime. PHP104 CURRENTUSESOFANDPERCEPTIONSABOUTFDAMASECTION114LinPJ1,HughesTE2,NeumannPJ1 1TuftsMedicalCenter,Boston,MA,USA,2i3Innovus,Carmel,IN,USAOBJECTIVES: Section114ofthe1997U.S.FoodandDrugAdministrationModernizationAct(FDAMA)stipulatestheconditionsunderwhichdrugcompaniescouldpromotehealtheconomicinformationtohealthcareplans.Thisstudyexaminedcurrent usesofandperceptionsaboutSection114. METHODS: Weconductedaweb-based surveyofaconveniencesampleof59expertsrepresentingpharmaceuticalcompanies andacademia.Weaskedabouttheirinterpretationof,andexperienceswith,Section 114,aswellastheirviewsregardingtheFDAissuingfurtherguidancetoadvisepharmaceuticalcompaniesonmakingpromotionaleconomicclaimstopayers. RESULTS: Thirty-threeofthe59experts(56%)completedthesurvey.Amongpharmaceutical companyhealthoutcomesdirectors,81%statedtheyalwaysorfrequentlyconsider usingSection114whenmakingpromotionalclaimsfordrugs,but63%statedthey wereunclearabouthowtoeffectivelycreate,approveoruseSection114information. ThereasonsfornotusingSection114included:notfeelingcomfortableusingthe Section(25%);thefactthateconomicvalueinformationnotincludedintheproduct label(13%);anduncertaintyaboutwhethercreatingaSection114piecewasworththe benet(13%).75%expectedtouseSection114toagreaterextentinthefuture.75%of outcomesdirectorand88%ofacademicexpertsstatedthattheFDAshouldissue guidanceonSection114,especiallyregardingwhatqualiesas"healthcareeconomic information"and"competentandreliablescienticevidence."69%ofoutcomesdirectorsand65%ofacademicexpertsagreedthattheincreasedfocusoncomparative effectivenessresearchwouldincreaseSection114use. CONCLUSIONS: PharmaceuticalcompaniesareconsideringandusingFDAMASection114tomakepromotional economicclaimsfordrugs,despitetheirdiverseinterpretationsofthelaw.Direction fromtheFDAmayclarifyhowcompaniescouldsharearangeofcomparativeeconomicinformationwithhealthplans. PHP105 HOWARECOVERAGEDECISIONSMADEINPUBLICLYFUNDEDHEALTHCARE PROGRAMSINLOW-ANDMIDDLE-INCOMECOUNTRIES?HornbergerJ,ShewadeA,GutierrezH CedarAssociatesLLC,MenloPark,CA,USAOBJECTIVES: Universalhealthcoverageforallpersonsworldwidewasthefocusof theFirstGlobalSymposiumonHealthSystemsResearch(Montreux,Switzerland; Nov2010).AspartofalargerinitiativesponsoredbyTheRockefellerFoundationto transformhealthsystems(accessible,affordable,equitable),westudiedhowcoveragedecisionsaremadebypubliclyfundedhealthcareprogramsinlow-and middle-incomecountries(LMICs). METHODS: Anonlinesurveywasadministered torepresentativesof 30healthcareprogramsinLMICs.In-depthcasestudies, basedonethnographicresearchmethodologies,wereconductedonasubsetof programstoaddressthefollowingresearchquestions:Whoiscoveredintheprogram?Whathealthconditionsandservicesarecovered?Whatarethecultural, political,ethical,scientic,andeconomicfactorsthatinuencecoveragedecisions?Whataretheadministrativeprocessesforcoveragedecisions?Whatevidenceisusedtodeterminecoverage?Whatprocessesareusedtomonitor outcomes? RESULTS: Maternalhealth,cardiovasculardiseases,andpreventing catastrophicillnessthatresultsinlargenancialburdentothepublictendtobethe highestpriorityconditionsforcoverage.Dentalcareandtreatmentforsubstance abusearetheleastcommonlycoveredconditions.Mostprogramsprovidecoverage fordrugs;policiesfordrugsarebasedmostlyonpublishedreportsofclinicalsafety andefcacy.Factorsinuencingdecisionsincludedlimitingthenancialriskfrom catastrophicillness,abilitytoscaleprograms,andequityissues.Evidenceusedfor coveragedecisionswasbasedonmedicalliterature,regionalandglobalhealth research,andcoveragepoliciesofotherprograms.Someprogramsundertakeeconomicappraisalsandcomparativeeffectivenessassessmentsindetermining policies. CONCLUSIONS: Thisethnographicsurveyonhowcoveragepoliciesare beingdevelopedprovidesopportunitiesforLMICstosharetheirexperienceswith eachother,andthusfurtherdevelopandrenetheirprogramstomeetgoalsof accessible,affordable,andequitableheathcareforall. HealthCareUse&PolicyStudies–ConceputalPapers PHP106 DISPARITYINACCESSTONOVELDIABETICAGENTSFORMEDICAREPARTD ELDERLYCOMPAREDTOCOMMERCIALINSUREDPATIENTS:INSIGHTINTO DOWNSTREAMEFFECTSOFCMSFORMULARYCOVERAGEGUIDELINESFOR PDP’SSchwartzEL1,DavisEA1,BelaziD2,HasanS2 1PriceSpective,SanDiego,CA,USA,2WoltersKluwerPharmaSolutions,Yardley,PA,USAA30VALUEINHEALTH14(2011)A1–A214

PAGE 39

BACKGROUND: Asageincreases,prevalenceoftype2diabetesintheU.S.rises dramaticallyasthepopulationapproachesandentersMedicareeligibility(CDC). Althoughensuringcategoryaccess,CMSformularyguidelinesforMedicarePartD (MPD)coveragedonottakeintoaccounttheeffectsofcost-sharingburdenon patientcompliance.Literaturedemonstratesthatpatientadherenceisreduced withhighercopaymentcostsandconsequently,thebenecialclinicalimpactsmay likelybeunrealizedformanypatients. OBJECTIVES: ToinvestigateaccesstodiabeticmedicationsforMPDpatientscomparedtocommerciallycoveredlives.Exploringcopaydifferentialsamongstthesepopulations,insightisgainedonhow MPDdiffersfromcommercialaccesstodiabetesmedications. METHODS: Analysis oftheWaltersKluwerPharmaSolutionsSourceLongitudinalPatientDatabase, samplingof26.7millioncommerciallivesand5millionMedicarePartDlivesin 2009.LowIncomeSubsidycoveredliveswereexcluded. RESULTS: Averagedrug copaymentformetforminandsulfonylureaforcommercialandMPDpatientsis $15and$19respectively.Averagedrugcopaymentforinsulinglargineincommercialis$17and$27forMPDpatients,forbrandedpioglitazone$31and$52,and exenatide$32and$68. CONCLUSIONS: Copaymentdifferentialsacrossthesepopulationsaresmallforgenerictherapiesandgrowlargerforbranded,noveldiabetic agents.Thisdatawouldsuggestabroader,moreinclusivereviewisneededto assesshowthenancialburdenfeltbyMPDdiabetespatientsaffectspatientcomplianceandoutcomes.Furtherinvestigationisneededtostudythepotentialvalue thatCMSwouldbenetfromre-evaluatingthecostsharingburdenforthispatient population. PHP107 BIOMARKERS:ACHANGINGPARADIGMFORDEVELOPMENTPardiniAT,SarafS,SparrowhawkK PriceSpectiveLtd.,London,UKThegeneralperceptionisthatpharmacompaniesgenerallyincorporatebiomarker (BM)developmentintotheirprocesseswhenadrugresponseisnotoptimalfollowingtheresultsofphase3studies.Usuallythisiswhenpoorerthanexpectedefcacyresultsareachieved.Inthisscenario,theBMdevelopmentenablestargeting ofanichepopulationthatisrepresentativeoftheresponders,thuseffectively increasingtheefcacymakingtheproductmoreattractivetopayersandhealthcareprofessionals.ThislatestageapproachtoBMdevelopmentalsotswiththe commonlyheldbeliefthatBMsarelinkedtoreducedmarketaccess(MA),lower marketsharesanddecreasedproductrevenues.InsuchasituationBMsareoften onlydevelopedretrospectivelytoovercomeaccessissues.OurobjectiveistodemonstratethatinvestmentintoBMsintheearlyphaseofdrugdevelopment(DD)is morecommerciallyattractive. METHODS: Threescenariosofdrugdevelopment weredened:1)thecurrent/traditionalmodel,2)wherebiomarkerdevelopment isincorporatedfromphaseIofdevelopmentand3)wherebiomarkerdevelopmentinvolvesaconcurrentPhaseIIIinvestigationorPhaseIVretrospective analysis.Thesescenarioswereanalyzedtodeterminetherelationshipbetween riskandrewardusingassumedcash-owcurvesandnet-presentvalueanalysis baseduponthosecurves.Ineachcase,theimplicationsofintegrationofBM developmentatvariousstagesandhowthisaffectsrisk-rewardwereassessed. RESULTSANDCONCLUSIONS: ThescenarioanalysisdemonstratesthatbyshiftinginvestmenttoearlierintheDDprocess,costsassociatedwithinvestmentheavyPhaseIIIwillbereduced.EarlyincorporationofBMsintoDDwillimprovethe commercialandhealthcarebenetsandthedrugwillhavethepotentialtobenet fromshortenedapprovaltime,earlyMAandhigherprice. PHP108 THECASEOFRAREDISEASEDRUGSBEFOREANDAFTERTHEINTRODUCTION OFPRICINGBODIES:LESSONSLEARNEDFROMBRAZILANDCANADA, IMPLICATIONSFORTHEUNITEDSTATESCostP,SnyderT,ZaidiQ HERONEvidenceDevelopmentLLC,Somerville,NJ,USAOBJECTIVES: Thisposterexaminesthepharmaceuticalpriceimplicationforrare diseaseproductsintwocountrieswhichrecentlydevelopedtechnologyassessmentandpricingprocesseswithalooktowardthepotentialimplicationsforthe UnitedStates. METHODS: Casestudiesarebuiltoutofexaminingpricesforthe MultipleSclerosisdrugsinterferonbeta-1aandnatalizumabinthecontextofBrazil,whiletheGaucher'sdiseaseproductsimigluceraseandmiglustatarestudiedin Canada.Ineachcase,abriefoverviewofthehealthsystemsisgiven,withspecic attentiontothepricingbodies.Thepricesfordrugswhichcametomarketbefore andaftertheadventofapricingbodyarecomparedrelativetoeachother.These differencesarethencomparedtothepricedifferentialintheUSandUKtodetermineiftheHTAbodywasinstrumentalinthispricingchange. RESULTS: InCanada miglustatis17.8%ofthecostforimiglucerasewhileintheUSitis38%,witha similartrendinthepriceofMSdrugsinBrazil.Tosomedegree,thelowerpriceis expectedasthedrugclassesaredifferent.However,thedisparitybetweena17.8% differentialanda38%differentialsuggeststhattheCanadianpricingbodyisused toapplydownwardpressureonthepriceofrarediseasedrugs. CONCLUSIONS: The pricedifferentialhasdistinctimplicationsfortheUSmarket,inwhichpayersmay looktowardsdevelopingatechnologyassessmentprocessusingcosteffectiveness researchtodrivedowncostsduetothecurrentenvironment.Asoneofthemost importantmarketsforpharmaceuticalprots,thishasconsiderableramications forindustryintermsofincomeandinnovationincentive. PHP109 ANOUTCOMESPROFILEREGISTRYFORESTABLISHINGABASELINEMATRIXIN COMPARATIVEEFFECTIVENESSSTUDIESINPREDICTIVEPHARMACOLOGYWheelerC NTELX,Inc,Vienna,VA,USAWeproposeanovelapplieddecisionanalyticssolutioninclinicaloutcomesanalysisforderivingoutcomestobeusedasbenchmarksindesigningappropriate therapiesinpersonalizedmedicineandpredictivepharmacology.Theefcacyof comparativeeffectivenessresearchinclinicalmedicineandpharmacologyislimitedbythelackofadenedsolutiontoderiveclinicaloutcomesacrossdiverse patientpopulationsandavarietyofdisparatedatasourcesthatcollectivelydene aclinicalproleatparticularpointintime.AnoutcomeattimeT1isdrivennotonly bystaticfactorssuchasrace,ethnicityandoccupation,thataregenerallytimeindependent,butalsobytheconditionproleandresultantoutcomeofthepatient'sconditionatT0.Oursolutionisanensembleanalyticalframeworkthat leveragesatemporalruleinductionalgorithmtocreatederivedoutcomesproles acrossthetimecontinuum.Itperformsanalysisonstructuredandunstructured datafromEMR/EHR,clinical,biological,biomarker,behavioralanddemographic datasourcesthatareintegratedintoacompositedatawarehouseviaourpropriety semanticresolutionandnaturallanguageprocessingalgorithms.Theoutcomes prolesreectanindexoraggregatescorefortheamalgamationofallavailable dataforaparticularpatientataparticulartime.Outcomesprolesfromthousands ofsamplesarecataloguedandnormalizedinaregistryandareusedtoestablisha baselinematrixforapplicationinhigherlevelstatisticalandpredictiveanalysesfor comparativeeffectivenessstudiesinpharmacology.Usingthisapproach,itispossibletodeterminebasedonavailabledataboththeappropriatetreatmenttoaffect adesiredoutcomeandthepredictedoutcomebasedonagiventreatmentatagiven time. PHP110 BIOSIMILARSLITERATUREREVIEW:THECURRENTLANDSCAPEAND IMPLICATIONSOFRECENTHEALTHCARELEGISLATIONFORTHEUNITED STATESMARKETMaieseBA,LeeEH,ToscaniMT JeffersonSchoolofPopulationHealth,Philadelphia,PA,USABiosimilarsrepresentanemergingareaofinterestforthepharmaceuticalindustry. Biosimilarsareessentially'generic'versionsofbranded'biologics,butarenot consideredidenticaltotheinnovatorbiologic.In2009,halfofthetoptenselling drugswerebiologics;thisproportionisexpectedtoreach80%by2015.Biosimilars areseenasacost-savingalternativetopayers,andasgenericdrugstakeoverthe market,manufacturersaredependingonbiologicstodrivegrowth.However,they willneedtoconsidertheinherentchallengesinthismarket.Thisliteraturereview wasundertakentoprovideasummaryofthecurrentstateofaffairswithbiosimilars,includingareviewoftherecenthealthcarelegislationandpoliciesincountriesthatalreadyhaveformalguidanceregardingbiosimilarsapproval.Signicant resourcesarerequiredtoparticipateinthebiosimilarsmarket,includingregulatoryexpertise,manufacturingcapabilities,andglobalmarketreach.Oneconcern topotentialmarketplayersistheuncertaintyregardingtheapprovalprocessfor biosimilarsintheUnitedStates(U.S.).ThePatientProtectionandAffordableCare Act(PPACA2009)authorizedtheFDAtodevelopanabbreviatedregulatorypathwayforbiosimilarapprovalbutguidancehasnotyetbeenissued.Itisimportantfor theU.S.tolearnfromothercountries.Thisreviewincludesasummaryofother countries'approachestobiosimilarsapproval.Forexample,since2006,theEuropeanMedicinesAgencyhashadextensiverequirementsforpre-clinicalandclinicaldatatodemonstratequality,safety,andefcacyoftheproductseekingapproval.Questionsremainincluding:Howwillthecompetitivelandscapelookas biosimilarsenterthemarket?Whatwillbethecomfortlevelwithsubstitutability ofbiosimilars?Willtherebepatentprotectionforthemanufacturingprocess?And, whatwilltheFDArequireintermsofclinicaltrialsandothersupportivedatato recognizebiosimilars?Theliteraturereviewwilladdressthesequestionsand more. PHP111 OVERINFLATIONOFTHEGENETICCONTRIBUTIONTOSCHIZOPHRENIA: IMPLICATIONSFORNOVELTHERAPEUTICSFlemingM,MartinCR UniversityoftheWestofScotland,Ayr,UKThebiologicalmodelofschizophreniaremainsthedominantmodelwithinmental healthservicesandhasapowerfulandenduringinuenceontheprevailingformat ofmentalhealthcaredeliverytopatientswiththediagnosis.Thereexistsalmost universalacceptanceofageneticcauseforschizophreniathoughinmanyinstancesthisconictsbothphilosophicallyandclinicallywithaperson-centred recoveryorientatedapproach.Areviewoftheunderpinningresearchthatsupports thegeneticargumentwasconducted.Appraisaloffamily,twinandadoptionstudiesuncoversseriousawsinthemethodologiesandstatisticalanalysesusedin studies.Theseawstendtoarticiallyinatetheperceivedgeneticcontributionto schizophreniaandmoreovermayalsoinvalidatemanyofthereportedstudyndings.Thereexistsanabsenceofareplicableandconsistentndingindicatinga cleargeneticpathwaytoschizophrenia.Noveltherapeuticapproachesaimedat neurotransmitterreceptorsiteabnormalitiesshouldnotthereforebediscouraged byanyfundamentalrefocusongenetherapyapproaches. PHP112 DEFINITIONALCRITERIAFORCHRONICFATIGUESYNDROME:ACRITICAL REVIEWChristleyY,DuffyT,MartinCR UniversityoftheWestofScotland,Ayr,UKChronicFatigueSyndrome(CFS)isanenigmaticandmisunderstoodclinicalentity. Abroadrangeofetiologicalmechanismshavebeensuggestedincludingendocrine, immune,infectious,muscularandneurologicalabnormalities.However,thecause remainselusivethusimpactingondevelopingmodelsofevidencedbasedthera-A31VALUEINHEALTH14(2011)A1–A214

PAGE 40

peuticinterventionandtrialmonitoringandoutcomeevaluation.Complicating thissituationisinconsistenciesinCFScasedenition.Themainobjectiveisto provideacriticalreviewofthesimilaritiesanddifferencesbetweenthevarying approachestoCFScasedenition.Theconictsandcontroversiesthathave emergedasaresultofthedifferingdenitionalcriterionforCFSarehighlighted andthepotentialimpactonfutureresearchisidentied.Acriticalreviewofthe mostfrequentlyusedcasedenitionsinCFSwasconducted.Therearecurrently vecasedenitionsofCFS;however,themostprominentisthe1994Centrefor DiseaseControlandPreventionCaseDenitions.However,primafacecomparative advantagesofthisdenitionareelusiveandindeed,ithasbeenwidelycriticized foritslackofspecicity.Counterintuitively,thereislittlecompellingevidenceto supporttheefcacyofanyofthecasedenitionshaveproducedevidencetodemonstratetheiraccuracyorprecisionatdeningcasesofCFS.Asummarydescriptionofthesymptomproleforeachofthecasedenitionsisprovided.TheinconsistenciesthathaveemergedinCFSresearchasaconsequenceofdiffering approachestocasedenitionarecontrastedanddiscussed.Clinicalandresearch implicationsarehighlighted. PHP113 IQWIGANDHIQA,WHATARETHEYGOODFOR?THEEVOLUTIONOFTHEHTA AGENCY:TIMEFROMCREATIONTOFIRSTASSESSMENTANDIMPACTFUL APPRAISALGreenMJ HERONEvidenceDevelopmentLLC,Somerville,NJ,USAOBJECTIVES: ToevaluatethetimespentfromthecreationorcharterofanHTA agencytoendpointsindicatingtheireffectiveness,suchaspublicationofassessmentsandevidenceoftheincorporationofassessmentintomeaningfulappraisal inuencingpatientaccesstohealthtechnologies. METHODS: Thisstudylooksat thecreationofHTAagencies(e.g.AHRQ,HIQA,IQWiG,PBAC,CADTHandNICE)and theirevolutionintermsofrolesinassessment(advisory,coordinating,decisionmaking)andtherelationshiptheyhavewithappraisal. RESULTS: Ithasbeendemonstratedthatthetimeittakesforanagencytogenerateassessmentsimpacting patientaccessvarieswidely.Forexample,inIreland,HIQAwascharteredinMay 2007,andentrustedwithperformingHTAassessments.In2008and2009,HIQAhas publishedonehealthtechnologyassessmentperyear,bothofwhichwerereceived andinturnimplementedbytheMinisterforHealthandChildren.Incomparison, NICEintheUKwasfoundedin1999,butitsappraisalswerenotsupportedby mandateuntil2005.Meanwhile,HTAsdrivenbyDAHTA@DIMDIinGermanyare knowntorarelyplayaroleinpricingandreimbursement. CONCLUSIONS: The evolutionofHTAbodieshasvariedfromcountrytocountry.However,evolutionin scopeandimpactmayprovideusefullessonsforcountrieswhereHTAisreceiving renewedemphasisorwhereappraisalisunderconsiderationforimplementation, especiallyasnewagenciesarecreatedandexistingagenciesevolve. PHP114 OPTIMIZINGTHEORGANIZATION;MIGRATINGHEALTHSERVICESRESEARCH OPERATIONSINTOTHECOLLABORATIVESCIENCECENTEROFEXCELLENCEBensonJ,EnterlineL Bristol-MyersSquibb,Plainsboro,NJ,USATheexecutionandmanagementofHealthServicesResearchprojectscanbean oneroustask.Oftenthereisnocentralizedbodyofknowledgewithinanorganizationaroundprocessandrequirements.Thisleadstolongexecutiontimelines, difcultieswithvendorsandultimatelyreducedproductivity.TheCollaborative ScienceCenterofExcellence(CSCoE)wasestablishedin2008atBristol-Myers Squibb(BMS).Thisgroupmanagestheglobaloperationsofawidevarietyofprograms,aportionofwhichincludesworldwideinvestigatorsponsoredresearch, non-clinicalresearch,expandedaccess,andriskevaluationandmitigationprograms.BeginningJune2009,operationalmanagementoftheentireUSHealthEconomicsandOutcomesResearch(HEOR)bookofworkwasmovedfromtheOR ScientistsintotheCSCoE.Thisincludedadministration,contractexecution,masterserviceagreementnegotiation,nancialmanagement,protocolwriting,AMCP dossierupdates,andinvoicetrackingandpayment.Withintherstyearover90 projectsweremigratedintotheCSCoE.BenetstheHealthServicesResearchgroup realizedincluded:1.Aconsolidated2010andplanned2011bookofwork;2.A reportablerepositoryofprojectinformation;3.HEORprotocolandAMCPdossier improvementthroughstandardizationofin-housescienticwriting;4.Expedited contractexecution;5.Innovativecost-sharing;6.Tieredandbatchedreviewof contractsreducedcorporatelegalhours;7.Rapidresponsetoorganizationalqueries.CentralizedprocessmanagementunlockedlatentvaluebyallowingORScientiststofocusonvalue-addedactivities,increasedorganizationaltransparency andagility,andmovedoperationstoalowercostenvironment. POSTERSESSIONI: DISEASE-SPECIFICSTUDIES CardiovascularDisorders–ClinicalOutcomesStudies PCV1 DOESROUTEOFADMINISTRATIONFORESTROGENHORMONETHERAPY IMPACTRISKOFVENOUSTHROMBOEMBOLISM:ESTRADIOLTRANSDERMAL SYSTEMVERSUSORALESTROGEN-ONLYHORMONETHERAPYLalibertŽF1,DeaK1,DuhMS2,KahlerKH3,RolliM3,LefebvreP1 1Grouped'analyse,LtŽe,Montreal,QC,Canada,2AnalysisGroup,Inc.,Boston,MA,USA,3NovartisPharmaceuticalsCorporation,EastHanover,NJ,USAOBJECTIVES: Toevaluatetheriskofdevelopingvenousthromboembolism(VTE) eventsassociatedwiththeuseofestradioltransdermalsystem(ETS;Vivelle-Dot¨) relativetooralestrogen-onlyhormonetherapy(HT)agents. METHODS: Ahealth insuranceclaimsanalysiswasconductedusingtheThomsonReutersMarketScan databasefromJanuary2002throughOctober2009.Patients 35yearsoldnewly initiatedonanETSororalestrogen-onlyHTwith 2dispensingswereanalyzed. VTEwasdenedas 1diagnosiscodefordeepveinthrombosisorpulmonary embolism.AsasecondaryoutcomeweassessedincidentVTEresultinginhospitalization.CohortsofETSandoralestrogen-onlyHTwerematched1:1basedon bothexactfactorandpropensityscorematching.Incidencerateratio(IRR)was usedtocomparetheratesofVTEbetweenthematchedcohorts.Remainingbaselineimbalancesfrommatchingwereincludedascovariatesinmultivariate adjustments. RESULTS: AmongthematchedETSandoralestrogen-onlyHTusers (27,018subjectsineachgroup),themean(SD)agesofthecohortswere48.9(7.1) years;ineachcohort6,044(22.4%)and1,788(6.6%)patientshadahysterectomyand anoophorectomyatbaseline,respectively.Themean(median)drugexposurefor theETSandoralestrogen-onlyHTcohortswas391(264)and401(272)days,respectively.Atotalof115ETSusersdevelopedVTEcomparedto164subjectsinthe estrogen-onlyHTcohort(unadjustedIRR:0.72;95%CI:0.57-0.91, P 0.006).After adjustments,ETSremainedstatisticallysignicantlyassociatedwithalowerincidence(33%reduction; P 0.0134)ofVTE.Theincidenceratereductionforhospitalization-relatedVTEeventsamongtheETSuserswasevenmorepronouncedwith theadjustedincidencebeing62%lowerforETSusersrelativetooralestrogen-only HTusers. CONCLUSIONS: Resultsofthislargepopulation-basedstudyshowedthat patientsreceivingETShadasignicantlylowerincidenceofVTEcomparedto patientsreceivingoralestrogen-onlyHT. PCV2 THERISKOFCARDIOVASCULAREVENTSASSOCIATEDWITHDIETARY CALCIUMANDVITAMINDSUPPLEMENTSINPATIENTSWITHOSTEOPOROSISChangJY1,NicholMB2 1UniversityofSouthernCalifornia,Alhambra,CA,USA,2UniversityofSouthernCalifornia,Los Angeles,CA,USAOBJECTIVES: CalciumandvitaminDsupplementshavebeenwidelyusedand recommendedforwomentopreventordelaytheonsetofosteoporosisandtherisk ofbonefractures.Otherbenetsincludetheimprovementofbloodpressureand lipidlevelsandaloweringofbodyweight.Intheory,thebenecialeffectsofcalciumandvitaminDsuggestimprovementsincardiovascularhealth.Recentpublicationssuggestthecontraryandalludetoincreaseserumcalciumasariskfactor foradversecardiovascularevents.Thisstudyexamineswhethertheexposureto thesesupplementsareassociatedwithcardiovascularevents. METHODS: The studywasbasedonCaliforniaMedicaid(Medi-Cal)fee-for-serviceadministrative claimsdatafromJanuary1995toDecember2002.Thestudypopulationconsistof patients 50yearswithrecordeddiagnosesofosteoporosisfollowedfromdiagnosesdatetotheendofeligibility.Patientswereexcludedforprioruseofthesupplementsordiagnosisofcardiovasculareventsordruginducedosteoporosis.Propensityscorematchingbasedonage,gender,elixhausercomorbiditiesandeligibility datacreatedcase(n 1594)andcontrolgroups(n 4782).Chi-squareanalysiswas conductedforcomparisonofthecardiovasculareventsdenedasICD9codesfor myocardialinfarctionandsearchabletermsof"cerebralinfarction,hemorrhage, ischemia"forstroke. RESULTS: Nostatisticallysignicantrelationshipwasfound betweenthestudygroupsforstroke(p 0.56)andmyocardialinfarction(p 0.54). Componentsofstrokeincludedcerebralarteryocclusion(p 0.94),precerebralarteryocclusion(p 0.27),intracerebralhemorrhage(p 0.23),andsubarachnoid hemorrhage(p 0.05).Theclinicalbenetsofthesupplementswereevidentwith subarachnoidhemorrhagewith0recordeddiagnosesinthecasegroupcompared to12recordeddiagnosisinthecontrolgroup;howeverstatisticalsignicancewas notestablished. CONCLUSIONS: TheuseofcalciumandvitaminDsupplementationyieldednorelationshiptotheriskofadversecardiovascularevents.Moreover, nobroadcardio-protectiveeffectscanbeconcludedfromthestudy. PCV3 RISKOFHOSPITALIZATIONSFORVENOUSTHROMBOEMBOLISMINATYPICAL VERSUSTYPICALANTIPSYCHOTICUSERSINANATIONALSAMPLEOF MEDICAREBENEFICIARIES:ACLAIMSDATAANALYSISDharmarajanSH,YangY,AthavaleAS,BentleyJP,NullKD,BanahanBF UniversityofMississippi,University,MS,USAOBJECTIVES: Toexaminethedifferencebetweentypicalandatypicalantipsychotic druguseintheriskofhospitalizationforvenousthromboembolism(VTE)inan elderlyMedicarepopulation. METHODS: Thisisaretrospectivecohortstudyusing 5%nationalsampleof2006-2007Medicareclaimsdata.Medicarebeneciarieswith continuousPartA,B,andDenrollmentin2006-2007andwhoinitiatedatypicalor typicalantipsychoticdrugtherapyinJuly2006-June2007wereincluded.Allstudy subjectswerefollowedforaperiodof180daysfromthedateofindexprescription. Atypicalandtypicaluserswerematchedonpropensityscore,calculatedusing pre-indexdemographics,clinicalcomorbidities,andmedicationuse.Aconditional logisticregressionmodelstratiedonthepropensityscore-matchedpairusingthe GreedymatchingalgorithmwasusedtocomparetheriskofhospitalizationforVTE innewusersofatypicalandtypicalantipsychoticdrugs.Sensitivityanalysisinthe unmatchedcohortwasperformedusingpropensityscoreasacontinuous,linear terminlogisticregression. RESULTS: Atotalof15,637newusersofatypicaland 2,337newusersoftypicalantipsychoticdrugswereidentied.Therewere472 (2.6%)individualswithahospitalizationforVTEduringfollow-up.417wereatypicaland55weretypicalantipsychoticusers.A1:1propensityscorematchyielded 2,333matchedpairs(4,666individuals).Inthematchedcohort,55typicaland64 atypicaldruguserswerehospitalizedforVTEinthefollowupperiod.Comparedto typicalantipsychoticusers,usersofatypicalantipsychoticswerelesslikelytohaveA32VALUEINHEALTH14(2011)A1–A214

PAGE 41

VTE-relatedhospitalizationsbuttheresultswerenotstatisticallysignicant(odds ratio:0.857;95%CI:0.596-1.233).Sensitivityanalysisresultsagreedwiththeprimaryndings(oddsratio:0.831;95%CI:0.622-1.110). CONCLUSIONS: Theriskof hospitalizationforVTEwasfoundtobesimilarforusersoftypicalandatypical antipsychoticmedicationinthiselderlyMedicarepopulation. PCV4 TREATMENTPATTERNS,CLINICALANDECONOMICBURDENOFVENOUS THROMBOEMBOLISMINABDOMINALSURGERYPATIENTSWangL,DysingerAH,DuJ,XieL,BaserO STATinMEDResearch,AnnArbor,MI,USAOBJECTIVES: Toexamineprophylaxisuseasitrelatestotheclinicalandeconomic burdenofvenousthromboembolism(VTE)inabdominalsurgerypatients. METHODS: Aretrospectivestudy(January1,2005toDecember31,2007)wasconductedusingasubsetoftheMarketScanHospitalDrugDatabaseanditslinked outpatientlesfromtheMarketScanCommercialandMedicareSupplemental database.Patients'demographics,clinicalanddischargestatuseswerecompared usingChi-squaretestingandstandardizeddifference.Risk-adjustedhealthcare visitsandcostswereestimatedusingtheGeneralLinearModel.Potentialrisk factorsforVTEeventswereselectedusingtheCoxProportionalHazardRegression Model. RESULTS: Inpatientswhounderwentabdominalsurgery(n 49,355), 24,473(49.59%)receivedanticoagulanttherapyduringtheirhospitalizationanduntil30daysafterdischarge.VTEeventsforpatientswhoreceivedanticoagulant prophylaxiswere4.63%versus7.02%forpatientswhodidnotreceiveanticoagulantprophylaxis.ComparedwithpatientswithoutVTEinthe6-monthfollow-up period,patientswithVTEweremorelikelytobeolder,havecomorbidconditions includingpreviousVTE,majorandminorbleeding,andcancer,andtohavehigher baselinehealthcarevisitsandcosts.Afterriskadjustmentforpre-speciedcovariates,inpatientcosts($16,677vs.$10,774),outpatientcosts($15,426vs.$7,424), pharmacycosts($2,541vs.$1,862),andreadmissionrates(0.23%vs.0.12%)were higherinpatientswhohadVTE.Inthemultivariateanalysis,appropriateanticoagulantprophylaxisusewassignicantlyassociatedwiththereducedriskofVTE (HR:0.515). CONCLUSIONS: TheVTEeventratewaslowerforpatientswhoreceived prophylaxiscomparedwiththosewhodidnot.Sincethehealthcarecostsofpatientswithaneventweresignicantlyhigherthanthoseofpatientswithoutan event,prophylaxisuseisassociatedwithlowerhealthcarecosts. PCV5 PREVALENCE,INCIDENCE,ANDOUTCOMESOFCRITICALLIMBISCHEMIAIN THEMEDICAREPOPULATIONINTHEUNITEDSTATESBaserO1,VerpillatP2,WangL1 1STATinMEDResearch,AnnArbor,MI,USA,2Sano-AventisGroup,Paris,FranceCriticallimbischemia(CLI)isasevereobstructionofthearteriesthatseriously decreasesbloodowtotheextremitiesresultinginamputationoftheaffectedlimb ifleftuntreated.Despitetheseverityofthedisease,thereisalackofdataon prevalence,incidenceandoutcomesofCLIintheUnitedStates. OBJECTIVES: We conductedalargepopulation-basedstudytodirectlyestimateprevalence,incidenceandoutcomesofCLIintheUnitedStates. METHODS: DatafromJanuary2006 toDecember2008wereextractedfromtheU.S.Medicaredataset.Weestimated1) age,gender,raceanddiabetes-specicCLIprevalenceandincidenceratesusing thedirectstandardizationmethod,and2)factorsassociatedwithoccurrenceof outcomes(legrevascularization,non-traumaticamputation,andmortality)among theCLIpopulation,usingtheCoxproportionalhazardregressionmodel. RESULTS: Atotalof68,074patientswereidentiedwitheligibleCLIICD-9codesin2007of whom44.53%werediabetic.Analyzing2yearsofdata,CLIprevalenceandannual incidenceratesinthiselderlypopulationwere0.23%(0.28%formaleand0.20%for femalepatients)and0.20%(0.23%and0.17%),respectively.Justaswithprevalence, incidenceincreasedsharplyamongbeneciariesaged65-69(0.13%)to 85(0.31%), wasaround2.3timeshigherinblackpatientscomparedtowhitepatients,and8.6 timeshigherindiabeticpatientscomparedtonon-diabeticpatients.Theoverall incidenceratesoflegrevascularizationandnon-traumaticamputationintheyear afterCLIdiagnosiswere29.7%and25.2%,respectively.Comparedtorevascularization,patientswhoareolder,male,black,andhavediabeteshadahigherprobabilityofamputation.30.3%ofthepatientsdiedwithintherstyearafterCLI diagnosis. CONCLUSIONS: ThisrstU.S.nationwide-basedstudyshowsthatprevalence,incidenceandoutcomesaredifferentaccordingtopatients'socio-demographiccharacteristicsandcomorbidities,suggestingthatCLIpatientmanagementvariesamongtheU.S.population. PCV6 ANALYSISOFMYOCARDIALINFARCTIONRELATEDCLINICALOUTCOMES, HEALTHCAREUTILIZATIONANDCOSTSOFPATIENTSWITHNON-VALVULAR ATRIALFIBRILLATIONWangL,XieL,BaserO STATinMEDResearch,AnnArbor,MI,USAOBJECTIVES: Toestimateclinicaloutcomes,healthcareutilizationandcostburden ofpatientswhosufferedamyocardialinfarctionduringthe180daysafterdiagnosisofnon-valvularatrialbrillation(NVAF)andcompareitwithpatientswhodid notsufferamyocardialinfarction. METHODS: Basedon2005-2007U.S.Medicare advantageinsuranceclaimles,patientsaged65yearsandolderwhohavehad twoormoreprimarydiagnosesforNVAFoccurringwithin30daysofoneanother wereselected.The180-dayfollow-upeventrates,healthcarefacilityuseandcosts forpatientswithamyocardialinfarctionandthosewithoutwerecompared.Risk adjustmentwasperformedusingthepropensityscorematchingmethodwiththe ProbChoicealgorithm. RESULTS: InpatientswhowereidentiedwithNVAF (n 18,575),258(1.39%)sufferedamyocardialinfarctionduringthe180daysafter NVAFdiagnosis.Patientswerenotsignicantlydifferentintermsofgender,region, andbaselinecomorbidconditions.Afterrisk-adjustmentforpre-speciedcovariates,mortality(10.08%vs.0.39%p 0.0001),outpatientemergencyroom(ER)visits (82.56%vs.48.06%p 0.0001),acutecoronarysyndrome(63vs.2/100personyears), ischemicstroke(31vs.4/100personyears),majorbleeding(10vs.1/100person years)andnon-majorclinicalrelevantbleeding(24vs.6/100personyears)wereall higherforpatientswhosufferedamyocardialinfarctioncomparedtothosewho didnot.Besidesinpatientcosts($26,646vs.$9,393),risk-adjustedoutpatientER costs($1,176vs.$863)werealsohigherformyocardialinfarctionpatients.The overallrisk-adjusteddifferenceinhealthcarecostsissignicant($36,584vs. $10,366p 0.0001). CONCLUSIONS: Mostoftheadverseeventsanalyzedwere higherforpatientswhosufferedamyocardialinfarctionafterNVAFrelativeto patientswhodidnot.Totalhealthcareutilizationandcostswerealsosignicantly increased. PCV7 ANALYSISOFTREATMENTPATTERNSANDCOSTSFORVENOUS THROMBOEMBOLISM,MAJORANDMINORBLEEDINGEVENTSINHOSPITALIZED MEDICALLYILLPATIENTSBaserO,XieL,DuJ,WangL STATinMEDResearch,AnnArbor,MI,USAOBJECTIVES: Toexaminetheprophylaxisuse,incidenceofVTE,majorbleeding, minorbleedingandassociatedeconomicburdenover90daysinhospitalizedmedically-illpatients. METHODS: Aretrospectivestudy(January1,2005toDecember 31,2007)wasconductedusingasubsetoftheMarketScanHospitalDrugDatabase anditslinkedoutpatientlesfromtheMarketScanCommercialandMedicare Supplementaldatabase.Eligiblepatientswereselectediftheywerecontinuously enrolledintheirhealthplanforatleast180dayspriortoand90daysfollowingthe indexhospitaldischargedate,whichishospitalizationwithamedicallyilldiagnosis.Prophylaxisuseisdenedastheadmissiondateofindexhospitalizationto30 daysafterindexhospitaldischargeandbeforethedateoftheirrstVTEevents. Patients'demographics,healthcarevisitsandcostswerecomparedusingChisquaretestingandstandardizeddifferences.Risk-adjustedtotalhealthcarecosts betweenpatientswitheventsandwithoutwereestimatedusingtheGeneralLinear Model. RESULTS: Inpatientswhowereidentiedasmedicallyill(n 12,077), 6,464(53.52%)receivedanticoagulanttherapyduringtheirhospitalizationanduntil 30daysafterdischarge.Comparedwithpatientswhodidnotreceiveanyanticoagulantprophylaxis,patientswhoreceivedanticoagulantprophylaxishadsignificantlylowerVTEevents(1.47%vs.3.58%,p 0.0001).Althoughtherewasnosignicantdifferenceinratesofmajorbleedingandminorbleeding,afterriskadjustmentforpre-speciedcovariates,patientswithoutcomeeventswere signicantlyassociatedwithhighertotalhealthcarecosts(VTE:$40,523vs.$17,698 p 0.0001;Majorbleeding:$27,430vs.$18,137p 0.0001;Minorbleeding:$25,696vs. $17,410p 0.0001). CONCLUSIONS: Despiteexistingguidelines,fewmedically-ill patientsarereceivinganticoagulantprophylaxis.AppropriateanticoagulantprophylaxisuseresultsinlowerVTEeventratesandtotalfollow-uphealthcarecosts inhospitalizedmedically-illpatients. PCV8 EVALUATIONOFASAFETYINITIATIVEANDOUTCOMESFORPATIENTSON STATIN-FIBRATECOMBINATIONTHERAPIESINTHEVETERANSAFFAIRS POPULATIONYiT,TranJN,BounthavongM VeteransAffairsSanDiegoHealthcareSystem,SanDiego,CA,USAOBJECTIVES: Duetothereportedhigherincidencesofadverseeventsinpatients oncombinationstatinsandbratescomparedtostatinsalone,theVeteransAffairs (VA)SanDiegoinitiatedaninterventiontoreducethenumberofpatientsoncombinationtherapies.Thisstudyaimedtoevaluatetheeffectivenessandsafetyof statin-bratecombinationtherapieswithintheVApopulation. METHODS: This wasaretrospectivecohortstudyofVAmedical,pharmacy,andlaboratorydata. PatientsonastatinandbratecombinationinJune2008wereselectedandstratiedbasedoncontinuationordiscontinuationofcombinationtherapybyJune 2009.Hyperlipidemicmeasures,safetymeasures,andadverseeventswereobtainedpre-andpost-intervention.Chi-squareandANOVAtestswereutilizedto testbetween-groupdifferencesandpairedt-testswereconductedtoanalyzewithin-groupdifferencesatpre-andpost-intervention.Repeatedmeasuresregressions wereusedtoassesslongitudinaldifferencesbetweengroupsovertime. RESULTS: Nodifferencesinratesofcardiovasculardiseasewerefoundbetweenpatientswho continuedanddiscontinuedcombinationtherapyatbaselineandoneyearlater. Noincidencesofrhabdomyolysisandpancreatitiswerereported.Comparedto thosewhocontinuedcombinationtherapy,thosewhodiscontinuedcombination therapyhadsignicantlylowertotalcholesterol(184versus172mg/dL,p 0.029), higherCK(171versus132U/L,p 0.006),andlowertriglyceridelevelspost-intervention.Therewerenosignicantlongitudinaldifferencesbetweengroupsover time.Withinthepatientswhodiscontinuedcombinationtherapy,ASTandALT weresignicantlyloweraftertheinterventionwhiletotalcholesterolandLDLlevelswerehigherpost-intervention.Withinthosewhocontinued,ALTwaslower whiletotalcholesterol,LDL,andCKwerehigherpost-intervention. CONCLUSIONS: Discontinuationofstatin-bratecombinationtherapydidnothaveasignicant impactonthemanagementofhyperlipidemiameasuresandnodifferencesin adverseeventswereobserved.Furtherstudiesshouldbedonetoassessthelongtermeffectsofstatin-bratecombinationtherapy.A33VALUEINHEALTH14(2011)A1–A214

PAGE 42

PCV10 THEEFFECTOFINTERACTIONSBETWEENCLOPIDOGRELANDPROTONPUMP INHIBITORSONADVERSECARDIOVASCULAROUTCOMESINCOMMERCIALLY INSUREDPATIENTSWITHACUTECORONARYSYNDROMEBhurkeS,MartinB,LiC,FranksA,BursacZ,SaidQ UniversityofArkansasforMedicalSciences,LittleRock,AR,USAOBJECTIVES: FollowingaFDAwarninginNovember2009,signicantcontroversy existsregardingtheoutcomesofpatientsco-medicatedwithclopidogreland omeprazoleafteracutecoronarysyndrome(ACS).Thisstudyexaminedtheeffect ofprotonpumpinhibitors(PPI)clopidogrelinteractionsonsubsequentACSemergencydepartmentandinpatientvisits. METHODS: Thiswasaretrospectivecohort studyofadministrativeclaimsdataforalargenationallydispersedgroupofcommerciallyinsuredsubjectsbetween2001and2008.Subjectsage 18yearswitha diagnosisofACSandatleastoneclopidogrelprescriptionwithin90daysafterthe diagnosiswereincluded.TheclopidogrelplusPPI(C PPI)groupwasdenedas subjectswithaminimumof7daysoverlapbetweenthePPIandclopidogrelprescriptions.Subjectswerefollowedfromtheirrstclopidogrelprescriptionuntil theyexperiencedare-hospitalizationorERvisitduetoACS,disenrolledorreached thestudyend.C PPIgroupwasmatched1:1withclopidogrelgroupusingpropensityscoringmethodswithcalipers.Coxproportionalhazardsregressionwasused toestimatetherelativeriskofanadversecardiovascularevent. RESULTS: Ofthe 10,101patientstakingclopidogrel,16.98%(n 1,716)wereprescribedaPPI.Propensitymatchingresultedin1,697patientpairs.Themeanagewas61.50yearswitha meanfollowupof259daysand69.64%weremales.13.20%(n 224)hadanACSrelatedre-hospitalizationorERvisitintheclopidogrelgroupversus16.32%(n 277) intheC PPIgroup.C PPIusewasnotassociatedwithasignicantlyincreasedrisk ofadverseoutcomes(HR 1.221;95%CI,0.984-1.517)comparedtoclopidogrelusers notco-medicatedwithaPPI. CONCLUSIONS: Concurrentuseofclopidogreland PPIstrendedtowardanon-signicantincreaseinriskofadversecardiovascular outcomesforACSpatients,whichsuggestscautionmaybewarrantedwhenprescribingaPPIwithclopidogrel.Futurestudiesshouldaccountfortimedependence ofexposure. PCV11 PNEUMONIAAFTERACUTEISCHEMICSTROKE:PREVALENCE,ASSOCIATED FACTORSANDOUTCOMESAl-JabiSW1,HassanY1,AbdAzizN1,LooiI2,ZyoudSH1 1UniversitiSainsMalaysia(USM),Pinang,Malaysia,2PulauPinangHospital,Pinang,MalaysiaOBJECTIVES: Pneumoniaisoneofthemostfrequentmedicalcomplicationsof acuteischemicstroke,oftenapparentearlyafterstrokeonset,anditisassociated withincreasedriskofdeathafterstrokeattack.Weaimedtoidentifyclinically usefulfactorsassociatedwithpneumonia,andtoexaminetheeffectofpneumonia onpatient'sfunctionaloutcomeatdischargeandonin-hospitalmortalityafterthe attack. METHODS: Itisanevaluationofpost-strokepneumoniacomplication amongischemicstrokepatientsattendingahospitalinMalaysiafromNovember1, 2008toApril30,2009.Dataincludeddemographicinformation,riskfactorsand clinicalcharacteristics.FunctionaloutcomeatdischargeasmeasuredbytheModiedBartherIndex(MBI)andin-hospitalmortalitywereassessed.Pooroutcome wasdenedasMBI 75.SPSSversion15wasusedfordataanalysis. RESULTS: A totalof256patientswerestudied,ofwhich33(12.9%)experiencedpneumonia complicationduringhospitalization.Thekeybaselinefactorsassociatedwiththe occurrenceofpost-strokepneumoniaweretotalanteriorcirculationinfarct(P 0.001),moderateandsevereGlasgowComaScale(P 0.001),atrialbrillation(P 0.035)andrenalimpairment(P 0.001).Moreover,24(60%)ofdeadcaseswere sufferingfrompneumoniaduringhospitalization(P 0.001).Additionally,after excludingthedeadcases,8(9.2%)ofpatientsdischargedwithpoorfunctional outcomewerehavingpneumoniaduringtheirhospitalstay(P 0.002). CONCLUSIONS: Pneumoniaisindependentlyassociatedwithischemicstrokepoor outcome.Identicationofmedicalhistoryandclinicalcharacteristicsonadmissioncanassistclinicianstoidentifypatientsathigherriskofdevelopingpoststrokepneumoniathushasteningtheinitiationofcertaininterventionstoimprove patientoutcome. PCV12 IMPACTOFPOTENTIALDRUG-DRUGINTERACTIONS(DDIS)ONHEALTH OUTCOMESANDCOSTTOMEDICAID:THEMONETARYBENFITSOFQUALITY HEALTHCAREAthavaleAS,BanahanBFI,MendoncaCM UniversityofMississippi,University,MS,USAOBJECTIVES: ToexaminetheimpactofpotentialDDIsonhealthoutcomesandthe associatedcosttotheMississippiMedicaidprogram. METHODS: Aretrospective matchedcohortstudywasconductedinMississippiMedicaidenrolleesforyears 2002-2004.EnrolleeswereclassiedasexposedtoapotentialDDIiftheobjectand precipitantdrugswerepossessedconcomitantly,withrstdayofoverlapbeingthe DDIeventdate.Exposedenrolleeswerematchedtoenrolleestakingobjectdrugs withoutapotentialDDIbasedondemographic,comorbidityandobjectdrug-relatedvariables.Controlswereassignedtheeventdateforthematchedexposed case.Conditionallogisticregressionwasusedtoanalyzetheeffectonhealthoutcomes(hospitalizationsandERvisitswithin30daysofpotentialDDI)andpaired t-testsforcoststoMississippiMedicaid. RESULTS: DDIswiththegreatesthealth andeconomiceffectsincluded:ACE/ARBsoddsratio(OR)forhospitalizations 1.75,ORforERvisits 1.39,differenceinaverageperpatienthospitalpayments ( APPHP) $124,anddifferenceinaverageperpatientERpayments( APPERP) $63);betablockersORforhospitalizations 1.52,ORforERvisits 1.24(n.s.), APPHP $78.13,and APPERP $43.67;clonidineORforhospitalizations 1.67,ORforERvisits 1.35, APPHP $50, APPHP $8.17;warfarinwithquinoloneORforhospitalizations 1.17(n.s.),ORforERvisits 1.27; APPHP $83.35, and APPERP $48.14);warfarinwiththyroidhormonesORforhospitalizations 2.47,ORforERvisits 1.96, APPHP $106.03,and APPERP $100.33. CONCLUSIONS: Basedonthesendings,MedicaidinterventionstrategiestoreducetheincidenceofpotentialDDIsshouldbeacost-effectivemethodforimprovinghealthcarequalityandthusapriorityforstateMedicaidprograms. PCV13 BURDENOFCOMORBIDITIESAMONGPATIENTSWITHATRIALFIBRILLATIONLaMoriJC1,GrossHJ2,PatelAA1,CrainM3,DiBonaventuraMD2,ModySH1,ScheinJ1 1Ortho-McNeilJanssenScienticAffairs,LLC,Raritan,NJ,USA,2KantarHealth,NewYork,NY, USA,3CentocorOrthoBiotechServices,LLC,VirginiaBeach,VA,USAOBJECTIVES: Atrialbrillation(AF)maymanifestwithcomorbidities.WeexaminedtheprevalenceofcomorbiditiesandgeneralmedicationuseamongAFpatientsinordertoassesstotaldiseaseburden METHODS: Datawereobtainedfrom the2009NationalHealthandWellnessSurvey(N 75,000),anannualcross-sectionalInternet-basedsurveyofadultsintheUnitedStates.Inadditiontodemographicsandmedicationuse,patientswithAFalsoreportedontheircomorbid conditions.Usingdemographicandpatientcharacteristics,aCHADS2score(an indexofstrokerisk)wascalculatedforeachpatient. RESULTS: Atotalof1297 patientsreportedadiagnosisofAF.Themeanagewas64.9years(SD12.2),and65% weremale.InadditiontoAF,thesepatientsreportedcomorbiditiesinvarious organsystems,including90%withacardiovascularcondition,62%withaurologicalcondition,42%witharespiratorycondition,and41%withagastrointestinal condition.SpeciccomorbidconditionsreportedinthisAFpatientpopulationincludedhypertensionin72%ofpatients,historyofmyocardialinfarctionin21%of patients,heartburn/gastroesophagealreuxdiseasein29%ofpatients,andallergies/hayfeverin29%ofpatients.ThemeanCharlsonComorbidityIndexscorewas 1.53forallpatients.Almosthalfofpatients(46%)hadaCHADS2scoreof 2.The percentageofpatientsreportingcurrentmedicationuseincluded:71%forAF,64% forhypertension,50%forhyperlipidemia,29%forarrhythmia,24%fordiabetes, and26%forgastrointestinalmedications.Overall,43%ofpatientswithAFwere usingananticoagulantmedication. CONCLUSIONS: Thisself-reportednational surveyidentiedAFpatientsashavingahighcomorbidityburden,withconditions affectingavarietyoforgansystems.Medicationsusedtotreatavarietyofconditionsarealsohighlyprevalentandshouldbetakenintoaccountinmanaging patientswithAF. PCV14 DYSPEPSIAANDDISEASEBURDENAMONGPATIENTSWITHATRIAL FIBRILLATION(AF)LaMoriJC1,GrossHJ2,PatelAA1,CrainM3,DiBonaventuraMD2,ModySH1,ScheinJ1 1Ortho-McNeilJanssenScienticAffairs,LLC,Raritan,NJ,USA,2KantarHealth,NewYork,NY, USA,3CentocorOrthoBiotechServices,LLC,VirginiaBeach,VA,USAOBJECTIVES: ManyagentsusedintreatingAFhavepotentialgastrointestinal(GI) tolerabilityissues.Treatment-relatedadverseGIeventsareacommonreasonfor noncompliancetotreatment.Thecurrentanalysisdescribestheprevalenceof dyspepsiainrelationtoanticoagulantuseamongAFpatients. METHODS: Data wereobtainedfromthe2009NationalHealthandWellnessSurvey(N 75,000),an annualcross-sectionalInternet-basedsurveyofUSadults.Respondentsanswered generaldemographicandhealth-relatedquestions.ACHADS2score(anindexof strokerisk)wascalculatedforeachpatientusingdemographicandclinical characteristics. RESULTS: Atotalof1297patients(1.7%)reportedadiagnosisofAF. Ofthesepatients,535(41%)alsoreportedaphysician-diagnosedGIcondition;449 ofthese(84%)wereconsistentwithdyspepsia(ulcers,abdominalbloating,abdominalpain,gastroesophagealreuxdisease/acidreux,orheartburn).Compared withAFpatientswithoutdyspepsia(n 848),thosewithdyspepsiawereyounger (mean62.9vs.66.0years, p 0.05)andmorelikelytobefemale(43%vs.31%, p 0.05).AFpatientswithdyspepsiawereinpoorerhealththanthosewithout dyspepsia,asevidencedbyahigherCHADS2score(1.9vs.1.4, p 0.05);thisdiffer encewasmorepronouncedinpatientsaged 65yearsandinthosewithCHADS2score 2.Despitethis,signicantlyfewerAFpatientswiththanwithoutdyspepsia weretakingaprescriptionmedicationtotreatAF(67%vs.73%, p 0.05)orananticoagulantforstrokeprevention(35%vs.47%, p 0.05). CONCLUSIONS: One-thirdof AFpatientsinthisanalysishaddyspepsia.ThesepatientsreportedagreaterdiseaseburdenandstrokeriskrelativetoAFpatientswithoutdyspepsia.Signicantly fewerAFpatientswiththanthosewithoutdyspepsiaweretakingaprescriptionto treatAForananticoagulantforstrokeprevention.PoorGItolerabilitymaybe signicantintheAFpopulationandshouldbeconsideredwhenadherenceto medicationiscritical,suchasinhigh-riskpopulations. PCV15 PREDICTORSOFTHECOMBINEDDIAGNOSISHYPERLIPIDEMIAAND HYPERTENSIO N-ANHANES2007-2008STUDYBhounsuleP,AbughoshS UniversityofHouston,Houston,TX,USAOBJECTIVES: Cardiovasculardisease(CVD)remainstheleadingcauseofdeathin theUnitedState(US).BothhyperlipidemiaandhypertensionareknownriskfactorsforCVD.TobaccouseaugmentsthisriskbyloweringtheHDLcholesterol, increasingbloodclotting,andacutelyelevatingbloodpressure.Theobjectiveof thisstudywastoexaminepredictorsofhavingbothco-morbiditiesincludingdemographiccharacteristicsandsmokingstatus METHODS: AretrospectivecrosssectionalstudywasconductedusingtheNHANES2007-2008database,astratied multistageprobabilitysampleoftheciviliannon-institutionalizedUSpopulation. TheoutcomevariablewasdenedasbeingdiagnosedwithbothhypertensionandA34VALUEINHEALTH14(2011)A1–A214

PAGE 43

hyperlipidemiabasedonthestandardizedinterviewconductedbytheNational CenterforHealthStatistics(NCHS)amongadults 18years.Descriptivestatistics andmultivariatelogisticregressionwereusedtodeterminepredictorshavingthe combineddiagnosisincludingthefollowingindependentvariables:gender,age, race,ethnicity,urban/ruralstatusandsmokingstatus.Statisticalanalysiswas performedusingSAS9.2. RESULTS: Thetotalsampleincluded33,994individuals andthecombineddiagnosisofhypertensionandhyperlipidemiawasreportedin 4.49%ofthesample.Ofthese,61.51%werefemalesand41%wereagedbetween 60-75years.Most(73.25%)ofthesepatientswerewhitesandmorethanhalf (54.75%)reportedtohavesmokedatleast100cigarettesintheirlifetime,while 34.01%werecurrentsmokers.Inthemultivariatemodel,signicantpredictorsof thecombineddiagnosisincludedfemalegender(OR 0.689;95%CI 0.5940.798; p .0001),age60-75yearsvs.youngerage(OR 1.627;95%CI 1.2962.042; p .0001),MexicanAmericanracevs.non-hispanics(OR 0.611;95%CI 0.490.76; p .0007)andcurrentsmokers.(OR 1.374;95%CI 1.08-1.75;p .0001). CONCLUSIONS: Ourndingsemphasizetheimportanceofdevelopingeffective interventionstrategiesthathelpsmokersquitandmayreduceprevalenceofhyperlipidemiawithhypertensionalongwiththeaccompanyingcardiovascularrisk. PCV16 MANAGINGGROWINGPOPULATIONOFTYPE-2DIABETESFROMCOMMUNITY TOTEACHINGHOSPITALSINCHINAZhangD1,JiL2,HuD2 1VitalStrategicResearchInstitute,Berwyn,PA,USA,2PekingUniversityPeople'sHospital, Beijing,ChinaOBJECTIVES: DiabetesprevalencehasgrownrapidlyinChina.Asthelargestepidemiologyandclinicaloutcomesassessmentprogram,ChinaCardiometabolic Registries(CCMR),hasbeendesignedtoestablishsystemicevaluationondisease progressionandinuencingfactors.TheNationwideAssessmentofCVDRiskFactors:BloodPressure,BloodLipid,andBloodGlucose,inChinesePatientswithtype 2Diabetes(T2D)(CCMR-3B)wasoneoftheCCMRstudiesandwasconductedto assesstheclinicaloutcomesofcurrenttreatmentpatterns. METHODS: Thiswasa crosssectional,observationalstudy.Patientswererecruitedfromcommunity(tier 1),regional(tier2),andteachinghospitals(tier3)fromall6majorgeographic regionsinChina,bycardiologists,nephrologistsandendocrinologist,andinternal medicine. RESULTS: Atotalof5099T2Dpatientswasincludedforthisanalysis. Acrossallhospitals,55%and34%ofthepatientshadhypertension(HTN)ordislipidemia(DYLP),respectively,while33%hadneither.Nodifferencewasfound betweenhospitalsregardingtheproportionofpatientsreachingtargetcontrol (HbA1C 7.0%,BP 130/80mmHg,LDL 2.6mmol/L)inthosewhoalsohaveHTNor DYLP(rangingfrom4.4%to6.6%fortheHTNgroupand6.9%to8.0%fortheDYLP group).HoweveratrendtowardabetterHbA1CcontrolinpatientswithoutHTNor DYLP(diabetesonly)inlargerhospitalswasseen(10.7%,11.2%,15.4%intier1,2and 3hospitals,respectively).Atrendtowardmoreuseofmultipleoralanti-diabetic drugsandinsulinwasalsoseeninlargerhospitals. CONCLUSIONS: BetterHbA1c controlseeninlargerhospitalswasassociatedwithmoreaggressiveuseofantidiabetictreatment.Thecontrolofcardiovascularriskfactors,bloodglucose,blood pressure,andbloodlipid,indiabetespatientswithHTNorDYLPremainstobe challenginginallhospitalsettings. PCV17 PREDICTORSOFTIMETODISCONTINUEBETA-BLOCKERFOLLOWINGACUTE MYOCARDIALINFARCTION:ANANALYSISOFTHEMEDICARE5%NATIONAL SAMPLEDATA2006-2007LokhandwalaT,YangY,ThumulaV,BentleyJP,StrumM,BanahanBF,NullK UniversityofMississippi,University,MS,USAOBJECTIVES: Tostudythepredictorsofbeta-blockertherapydiscontinuation amongpostmyocardialinfarction(MI)patientsenrolledinMedicare. METHODS: ThisisaretrospectivecohortstudyutilizingaMedicare5%nationalsampleclaims datafor2006-2007.MedicarebeneciarieswithcontinuousPartA,B,andDenrollmentin2006 2007,andwhowerehospitalizedforanacuteMIintherstsix monthsof2006wereidentiedusingavalidatedalgorithm,requiringahospitalizationepisode 3and 180dayswithanICD-9-CMof410.x1asprincipalorsecondarydiagnosis.Post-MIpatientswithalledprescriptionforbeta-blockerwithin 90daysofdischargewerefolloweduntiltheendofthestudyperiod.Timeto discontinuationwasdenedasdaysfrominitiationoftherapytoagapof 90days intherapy.SurvivalcurvesweremodeledusingtheKaplan-Meiertechnique,and potentialpredictorsoftherapydiscontinuation,includingdemographiccharacteristics,comorbidconditionsandconcomitantmedicationswereestimatedusing Coxproportionalhazardsregression. RESULTS: Ofthe2,505subjectswhometour inclusioncriteria,65.1%werefemales,83.3%wereCaucasian,meanage78.6( 8.2) years.About15%ofthemdiscontinuedtherapywithinsixmonthsandaround35% discontinuedwithinayear.Malesweremorelikelytodiscontinuetherapyascomparedtofemales(HR 1.166;[1.020-1.334]; p 0.0245)andCaucasians(HR 0.674; [0.557-0.815];p 0.0001)werelesslikelytodiscontinuetherapycomparedtoAfricanAmericans.TheresultsofCoxproportionalhazardsmodelshowsthatdiabetes (HR 1.224;[1.069-1.402]; p 0.0035),dyslipidemia(HR 0.675;[0.565-0.807]; p .0001),CHF(HR 1.234;[1.03-1.478];p 0.0223),cerebrovasculardisease (HR 1.206;[1.057-1.376];p 0.0054)andCOPD(HR 1.209;[1.061-1.377];p 0.0043) weresignicantpredictorsoftimetodiscontinuation.Patientsonconcomitant statintreatmentwerelesslikelytodiscontinuebeta-blockertherapy(HR 0.774; [0.667-0.899];p 0.0008). CONCLUSIONS: Manypatientsinitiatingbeta-blocker therapypostMIfailtoconsistentlyremainonthetreatment.Severaldemographic characteristicsandcomorbidconditionsareassociatedwiththisbehavior. PCV18 MEDICATIONADHERENCEANDHOSPITALIZATIONAMONGCHRONICHEART FAILUREMEDICAREBENEFICIARIESWITHANDWITHOUTDEPRESSIONChhabraP,DutcherSK,RattingerGB,ZuckermanIH,Simoni-WastilaL,GottleibSS, StuartB UniversityofMaryland,Baltimore,Baltimore,MD,USAOBJECTIVES: Adherencetochronicheartfailure(CHF)medicationsdecreasespreventablehospitalizations.Nevertheless,CHFrarelyoccurswithoutconcomitant conditions.Wesoughtto:1)evaluateimpactofcomorbiddepressiononCHFmedicationuseandadherenceamongMedicarebeneciarieswithCHF(CHF_MED_BEN);2)examinerelationshipsbetweenCHFmedicationadherenceandoutcomes;and3)determineifdepressionmodiesthisrelationship. METHODS: We employedacross-sectionaldesignof2006-2007CMSMedicareChronicCondition Warehouse(CCW)database.SamplewasrestrictedtoCHF_MED_BENwithMedicareA,BandDtoobserveallprescriptionandmedicalclaims.WeclassiedindividualswithCHFanddepressioniftheymettheCCWdenitions.Adherencemeasuresincludedanyuse(binary)andmedicationpossessionratio(MPR)for evidence-basedCHFmedications(EBM).Multivariableanalysesincludedaninteractiontermtotesteffectmodicationofdepressionontherelationshipbetween EBMadherenceandnumberofhospitalizations.Relativeratesofhospitalization arereportedwith95%CI. RESULTS: Atotalof151,924CHF_MED_BENmetinclusion criteria.Meanagewas75.9 /-12.0years;68%werefemale.25%haddepressionand theyweresignicantlyyounger(74.5vs.76.3years,p 0.001)andmorelikelytobe female(75%vs.66%,p 0.001)comparedtonon-depressedCHF_MED_BEN.AsignificantlylowerproportionofdepressedCHF_MED_BENreceivedEBM(75%vs. 79%,p 0.001),comparedtonon-depressedCHF_MED_BEN.MedianMPRwas0.92in bothgroups.AhigherproportionofdepressedCHF_MED_BENwashospitalized (76%vs.63%,p 0.001)andre-hospitalized(40.5%vs.38.6%,p 0.001)comparedto non-depressedCHF_MED_BEN.Inmultivariablemodels,depressionmodiedthe effectofadherenceonhospitalizations(interactiontermp 0.001):comparedto highadherence(MPR 0.9),pooradherence(MPR 0.5)wasassociatedwitha24% (95%CI1.20-1.28)increasedhospitalizationrateamongnon-depressedCHF_MED_BENanda45%(95%CI1.37-1.54)increasedhospitalizationrateamongdepressed CHF_MED_BEN. CONCLUSIONS: DepressionwasnotassociatedwithpoorEBMadherenceamongCHF_MED_BEN.PoorEBMadherencewasassociatedwithincreasedhospitalizationinbothdepressedandnon-depressedgroups,witha greatereffectamongdepressedCHF_MED_BEN. PCV19 ABAYESIANMULTIPLETREATMENTCOMPARISONOFPULMONARYARTERIAL HYPERTENSIONDRUGCLASSESBASEDONTHERISKOFMORTALITY REPORTEDINCLINICALTRIALSMutebiA,MaloneD UniversityofArizona,Tucson,AZ,USAOBJECTIVES: Tocomparetheperformanceofpulmonaryarterialhypertension (PAH)drugclassesbasedontheriskofmortalityreportedinclinicaltrials,and informarevisedPAHtreatmentalgorithm. METHODS: ThestudyusedBayesian AnalysisUsingGibbsSamplinginWindows(WinBUGS)andMonteCarloSimulationstoconductamultipletreatmentcomparisonofplaceboandthreePAHdrug classes;prostanoids(Treprostinil&Iloprost),endothelin-receptorantagonists (Bosentan,Sitaxsentan,&Ambrisentan),andphosphodiesterase(sildenal).Direct andindirectpairwiseoddsratios(OR)wereobtained.PublishedPAHstudies through2010wereidentiedfromMEDLINE(PubMed)databaseandanextended manualsearchwasalsoconductedbasedonreferencesfromidentiedstudies. Inclusionwasrestrictedtorandomizedcontrolledtrialslastingatleast12weeks andatmost16weeks,withsubjectshavingeitheridiopathicorassociatedPAH, andstudiesthatreportedmortalityasanendpoint.Studiesthatcomparedcombinationsofdrugswereexcludedfromtheanalysis.ResultsarereportedinORwith 95%credibleintervals. RESULTS: Intotal8studies(20treatmentarms,2,015subjectsenrolled)wereincludedintheanalysis.2studieswere3-armtrialsand1was a4-armtrial.Withplaceboasthereferenceclass,theORofmortalitywas1.00(0.06, 4.20)forprostanoids;1.14(0.15,4.69)forendothelin-receptorantagonistsand0.62 (0.05,2.83)forphosphodiesterase.Usingprostanoidsasthereferenceclass,theOR ofmortalitywas4.28(0.10,23.10)forendothelin-receptorantagonistsand2.40 (0.04,12.66)forphosphodiesterase.Usingendothelin-receptorantagonistsasthe referenceclass,theORofmortalitywas1.11(0.04,5.55)forphosphodiesterase. NoneoftheORswerestatisticallysignicant. CONCLUSIONS: Basedontheriskof mortalityreportedinclinicaltrials,thereisnostatisticallysignicantdifference amongPAHdrugclasses.Includingmorestudies/drugsandtheuseofdifferent PAHoutcomescouldinformamoredetailedcomparison. PCV20 PERSISTENCEWITHSTATINSANDPRIMARYPREVENTIONOF CARDIOVASCULAREVENTS:APOPULATION-BASEDCOHORTSTUDYChodickG,ShalevV,GoldsteinI,PorathA,SimahV MaccabiHealthcareServices,TelAviv,IsraelOBJECTIVES: Toevaluatetheassociationbetweenpersistentuseofstatinsandthe riskofacutecardiovasculareventsamongprimarypreventionpatientsincommunitysettings. METHODS: Apopulation-basedretrospectivecohortamong171,846 adultsaged45-75,withnohistoryofanycardiovasculardisease,whobeganstatin therapybetween1998and2009.Proportionofdayscovered(PDC)withstatinswas measuredbythenumberofdispensedprescriptionsduringfollow-upperiod. Studyendpointwasoccurrenceofamajorcardiovascularevent,whichcomprised myocardialinfarctionorperformanceofacardiacrevascularizationprocedure RESULTS: Thefullyadjustedsurvivalanalysisindicatedasignicantnegativeas-A35VALUEINHEALTH14(2011)A1–A214

PAGE 44

sociationbetweenpersistencewithstatinsandriskofincidentcardiacevents.The mostpersistentusers(PDC 80%)hadahazardratioof0.62(95%condence interval:0.58-0.67)comparedtonon-persistentusers(PDC 20%).Similarresults werefoundwhenanalyseswerelimitedtopatientswithmorethan5yearsof followup.Aninteractionanalysisbetweenpersistenceandotherfactorsdetected astrongerriskreductionamongdiabeticmales. CONCLUSIONS: Thislargeand unselectedcommunity-basedstudysupportstheresultsofseveralrandomized controlledtrialsregardingthebenecialeffectofpersistentstatintherapyagainst cardiaceventsamongprimarypreventionpatients. PCV21 EFFECTSOFAZILSARTANMEDOXOMILVERSUSVALSARTANAND OLMESARTANMEDOXOMILONTHEACHIEVEMENTOFSYSTOLICBLOOD PRESSUREGOALSAMONGHYPERTENSIVEPATIENTSWITHDIABETESSussmanM1,LeahyM2,FriedmanM1,MenzinJ1,NicholsC1,NeumannP3,SullivanSD4 1BostonHealthEconomics,Inc.,Waltham,MA,USA,2TakedaPharmaceuticalsInternational, Inc.,Deereld,IL,USA,3TuftsUniversitySchoolofMedicine,Boston,MA,USA,4Universityof Washington,Seattle,WA,USAOBJECTIVES: HealthcareEffectivenessDataandInformationSet(HEDIS)denes controlledhypertensionassystolic/diastolicBP(SBP/DBP) 130/80mmHgforpatientswithessentialhypertensionanddiabetes.Weestimatedthepercentageof diabetespatientswithuncontrolledessentialhypertensionwhowouldreachSBP goalswiththeangiotensinIIreceptorblockers(ARBs)azilsartanmedoxomil,valsartan,andolmesartanmedoxomil. METHODS: AMonteCarlosimulationmodel wascreatedtoestimatethenumberofpatientswithhypertension(SBP 130mm Hg)anddiabeteswhowouldachieveSBPgoalwhentreatedwithazilsartanmedoxomil,valsartan,orolmesartanmedoxomilfor12months.Acohortof100,000 hypotheticaldiabetespatientswithuncontrolledhypertensionwascreatedfrom NHANES19992006andassignedabaselineSBP.Follow-upSBPsweregeneratedby randomlysamplingfromthemeanandSDofthepercentagechangefrombaseline tonalvisitinsittingofceSBPsbyusingdatafromthediabetessubpopulation fromtheazilsartanmedoxomilclinicaltrialprogram.Mean SDrelativechangesin SBPwere 10.36% 10.41, 4.58% 11.04, 5.16% 11.82%(F-test P .001)forazilsartanmedoxomil,valsartan,andolmesartanmedoxomil,usingthepooledefcaciesacrossalldosages,respectively.Weassessedgoalattainmentassumingthat adherencewasalternativelyperfectandthat48%ofpatientsreceivinganyARB woulddiscontinuetreatment. RESULTS: PatientcharacteristicsbasedonNHANES dataweremean SDage56 13years,56%male,23%withpriorcardiovascular disease,baselineSBP151 19mmHg.Weestimatedthat41.0%ofpatientsreceiving azilsartanmedoxomilwouldachieveSBPgoalvs.26.8%forvalsartanand28.8%for olmesartanmedoxomil,assumingperfectadherence;accountingfornonadherence,21.2%,13.9%,and14.8%ofpatientswouldreachSBPgoals,respectively. CONCLUSIONS: Ourndingssuggestthatmorediabetespatientstreatedwithazilsartanmedoxomilthanwithvalsartanorolmesartanmedoxomilareexpectedto reachSBPgoal.FurtheranalysisshouldaddresswhetherthesedifferencesinSBP translateintobetterHEDISqualityscores. PCV22 METAREGRESSIONANALYSISTOINDIRECTLYCOMPARETHESAFETYAND EFFICACYOFDALTEPARINTOENOXAPARININPATIENTSWITHUNSTABLE CORONARYARTERYDISEASEDranitsarisG1,JelincicV1,ChoeY2 1AugmentiumPharmaConsulting,Toronto,ON,Canada,2Eisai,Inc.,WoodcliffLake,NJ,USAOBJECTIVES: Lowmolecularweightheparins(LWMH)areindicatedinunstable anginaandnon-Q-wavemyocardialinfacrtion.Eventhoughbothagentshave demonstratedsafetyandatleastcomparableefcacytounfractionatedheparin (UH),therehavebeennoheadtoheadtrialstoassesscomparativeeffectiveness. Metaanalytictechniqueswereusedtoperformanindirectstatisticalcomparison betweendalteparinandenoxaparininpatientswithunstablecoronaryarterydisease(UCAD). METHODS: AliteraturesearchwasconductedfromJanuary1980to November2009forrandomizedcontrolledtrialsevaluatingdalteparinorenoxaparinforthepreventionofmyocardialinfarctions(MI)anddeathinUCAD.Binary outcomes(e.g.MIs,death)fortheLMWHrelativetoUHwerestatisticallypooled usingaxedeffectsmodel.UsingUHasthecommoncontrol,anindirectstatistical comparisonbetweendalteparinandenoxaparinwasperformedusingmetaregressionanalysiswithactivedrug(dalteparinorenoxaparin)astheprimaryindependentvariable. RESULTS: SixUHcontrolledenoxaparin(n 4)anddalteparin(n 2) trialsmettheinclusioncriteria.Themetaanalysisofalltrialdatashowedthat patientstreatedwithenoxaparinordalteparinhadan11%relativeriskreduction forMIcomparedtoUH(RR 0.89,p 0.009).Thisbenetwasachievedwithouta signicantincreaseintheriskformajorbleeding(RR 0.98,p 0.85),thrombocytopenia(RR 1.14,p 0.35)ordeath(RR 0.95,p 0.68).Theindirectstatistical comparisonwasunabletondsignicantdifferencesbetweenenoxaparinand dalteparinintermsofMI(p 0.67),majorbleeds(p 0.79),thrombocytopenia (p 0.23)ordeath(p 0.10). CONCLUSIONS: Ourndingssuggestcomparablesafety andefcacybetweendalteparinandenoxaparinwhenusedtopreventMIand deathinpatientswithUCAD.Therefore,treatmentdecisionsshouldbebasedon otherconsiderations,suchaspatientorphysicianpreference,easeofadministrationandcost. PCV23 COMPARINGTHEEFFECTIVENESSOFROSUVASTATINANDATORVASTATININ PREVENTINGCARDIOVASCULAROUTCOMES:ESTIMATESUSINGTHE ARCHIMEDESMODELSchuetzCA1,vanHerickA1,AlperinP1,PeskinB1,HsiaJ2,GandhiSK2 1Archimedes,Inc.,SanFrancisco,CA,USA,2AstraZenecaPharmaceuticalsLP,Wilmington,DE,USAOBJECTIVES: Nomajorclinicaltrialhascomparedrosuvastatinwithatorvastatin inpreventingmajoradversecardiovascularevents(MACE).Thisstudyestimated theeffectivenessofrosuvastatin20mg(R20)versusatorvastatin40mg(A40)and 80mg(A80)inpreventingMACEinseveralhighercardiovascularriskpatientpopulationsusingsimulation. METHODS: Theextensivelyvalidatedandpublished Archimedesmodelwasusedtosimulatehead-to-headclinicaltrialsinseveral populations[10-yearFraminghamriskscore(FRS) 5%,FRS 20%,EURO-SCORE 5,Diabetes,secondaryprevention,andAcuteCoronarySyndrome(ACS)]toestimatetheoccurrenceofMACE(comprisingMI,stroke,andcardiovasculardeath) overtime.Patients(ages4570)basedontheNationalHealthandNutritionExaminationSurveywereenrolledintrialsimulations.Treatmentsweremodeledand validatedusingbiomarkerandoutcomesdatafrompublishedtrials. RESULTS: The numberofpatientsineacharmrangedfrom3,060to55,000,dependingonthetrial population.R20reducedMACEmorethanA40orA80inallscenarios,withhigher risksubgroupsshowinggreaterabsolutebenet.Inindividualtrialsimulations,the 5yearrelativerisk(RR)ofMACEforR20versusA40rangedfrom0.90to0.92. Similarly,the5yearRRofMACEforR20versusA80rangedfrom0.93to0.95.Similar estimateswereobservedat10and20years.The5-yearnumberneededtotreat (NNT)topreventoneMACEeventwithR20vs.A40andA80decreasedconsistently withincreasingbaselinecardiovascularrisk(forR20vs.A40,NNTof268forFRS 5%and55forACS). CONCLUSIONS: ThemodelestimatedthatR20lowerstherisk ofMACEmorethanA40orA80.Whilesimulationmodelscannotreplacecontrolled clinicaltrials,thisstudybridgesgapsintheevidenceandhelpsidentifycohorts thatwouldbenetmostfromtreatmentwithrosuvastatinratherthanatorvastatin. PCV24 EVALUATIONOFTHECLINICALFACTORSANDPREVENTIVEMEDICATIONS ASSOCIATEDWITHTHELENGTHOFHOSPITALSTAYAMONGISCHEMIC STROKEPATIENTSAl-JabiSW1,HassanY1,AbdAzizN1,LooiI2,ZyoudSH1 1UniversitiSainsMalaysia(USM),Pinang,Malaysia,2PulauPinangHospital,Pinang,MalaysiaOBJECTIVES: Lengthofhospitalstay(LOS)isamajorcostcomponentofhospital budgets.AccuratepredictionofLOShasbecomeincreasinglyimportantforhealth caresystems,andreducingtheLOShasthepotentialforlargesavingsinthepublic hospitalsystem.Thisstudyaimedtoassessthefactorsassociatedwithprolonged LOSofacuteischemicstroketakingintoconsiderationdemographic,riskfactors, andclinicalsignsthatcanbeassessedatthetimeofadmission.ParticularattentionispaidontheimpactofpreviousmedicationuseonLOS. METHODS: Aretrospectivecohortstudyofallacuteischemicstrokesurvivorsattendingahospitalin MalaysiafromMay1,2008toDecember31,2008.Longhospitalstaywasdenedas astaygreaterthanorequaltothemedianofLOS.Dataincludeddemographic information,clinicalinformation,riskfactors,andpreviousmedicationuse.SPSS version15wasusedfordataanalysis. RESULTS: Overall,363patientswerestudied. Themedian(interquartilerange)ofLOSwas69(45-111)hours.Theindependent factorsassociatedwithprolongedLOSwereahistoryofatrialbrillation( P 0.011), patientswithmoderateandsevereGlasgowComaScale( P 0.001),patientswith higherbodytemperature( P 0.015),patientswithhigherfastingorrandomblood glucose( P 0.004),andpatientswithoutprevioususeofangiotensinconverting enzymeinhibitormedication( P 0.027). CONCLUSIONS: Thisstudyprovidedscienticdataforthefactorsthatcouldhamperthedischarge,particularlybefore clinicianscanevaluatethemosteffective,efcient,andacceptablemethodsof managingpatientswithacuteischemicstroke.Moreover,thesevariablesarepotentiallypreventableortreatableatadmissiontimeandwouldbeidealtargetsto reducetheburdenofillnessandhealthcarecostsofischemicstroke. PCV25 MANAGEMENTOFACUTEISCHEMICSTROKEANDITSLONGTERMEVOLUTION INTHEUNITEDSTATESANDCANADARouleauA,GobauxV,GuilhaumeC,YiouA,MileaD LundbeckSAS,Issy-les-Moulineaux,FranceOBJECTIVES: Acuteischemicstroke(AIS)isamajorpublichealthconcernand amongtheleadingcausesofdeathanddisabilityinwesternsocieties.ThemanagementofAISanditslongtermevolutionintheUSAandCanadawasassessed throughaliteraturesearch. METHODS: Medlinewassearchedforthetimeperiod 1999-2010toidentifystrokecohortsandregistriescontainingrelevantdataonAIS managementand/orlongtermevolution.Subsequently,arankingprocesswas usedtoidentifyandselectthemostrelevantreferences. RESULTS: Atotalof680 referenceswereretrieved(581USAand99Canada).Publicationsfrommorethan80 distinctcohorts/registries/databaseswereanalyzedandanalselectionledtothe identicationof43publicationsfortheUSAand25forCanada.ISproportion amongallstrokesrangedfrom43%-90%inhospitalcohorts.24%-33%ofallstroke patientsarrivedatthehospitalwithin3hourofsymptomonsetandrt-PAtreatmentwasreceivedin1.1%-14%ofallAISpatients.Themajorityofpatients,i.e. 82%-100%underwentaCT/MRIscanning.Ingeneral,hospitalsettingandcare characteristicswerepoorlydocumentedandrelativelyfewqualityindicatorsof strokemanagementweretargeted.Forlongtermevolutionoutcomes,mortality waswelldocumented(5%-17%1month,17%-35%3months,30%-37%3years, 37%-54%5years).AISrecurrenceandcardiaceventsweredescribedinfew,mainly USA,studies.Therewasalackofdataondisability(modiedRankinScale)evolutionafter3monthsandstudiesondepressionwerescarce. CONCLUSIONS: Hospitalsetting,characteristicsofcare,andlongtermevolutionofdisabilityarepoorly documentedinAmericanandCanadianregistries.Newcohortandregistrystudies shouldspecicallyaimtogeneratereallifedataonischemicstrokecarethatinuenceshortandlong-termclinicalandassociatedeconomicoutcomes.A36VALUEINHEALTH14(2011)A1–A214

PAGE 45

PCV26 EFFECTSOFTHIAZOLIDINEDIONESONCARDIOVASCULAREVENTSIN PATIENTSWITHTYPE-2DIABETESMELLITUSAFTERDRUG-ELUTINGSTENT IMPLANTATIONYehSZ1,HsiehCF2,TsaiYW3,HuangWF2 1AstraZeneca,Taipei,Taiwan,2NationalYang-MingUniversity,Taipei,Taiwan,3Instituteof Health&WelfarePolicy,NationalYang-MingUniversity,Taipei,TaiwanOBJECTIVES: Tocompareclinicaloutcomesoftype2diabeticpatientswithand withoutTZDstherapyafterpatientsreceivingDES. METHODS: Weconducteda retrospectivecohortstudyusingtheNationalHealthInsuranceDatabase(NHIRD). Thetype2diabetesmellituspatientswereincludediftheyreceivedrstlimuselutingstentorpaclitaxel-elutingstentplacementandwereidentiedbypresence ofahospitalclaimduringtheperiodfromDecember1,2006,throughDecember31, 2007.Follow-updatawereavailablethroughDecember31,2008.Patientswere classiedintotwogroupsbasedontheantidiabeticagentstheytookfrompharmacyrecordsforuseofTZD(rosiglitazone,pioglitazone)ornon-TZDwithin3 monthsaftertheindexdateofhospitalization.Atotalof1,743patientswhoreceivedstentsduringthestudyperiodwereidentiedasthestudysubjects.Our measureofeffectivenesswastheprevalenceofdeath,myocardialinfarctionand repeatrevascularization,denedasanyPCI,whetherornotthepatientreceiveda stent,orcrossedovertoCABGwithinoneyearafterindexhospitalization. RESULTS: Therewere268patientsinTZDgroup,1,475patientsinnon-TZDgroup. Comparedwithnon-TZDgroup,therewerenosignicantdifferenceinadjusted hazardratioofdeath,myocardialinfarctionandrepeatrevascularizationbetween limus-elutingstentgroupandpaclitaxel-elutingstentgroup.Instratiedanalysis, patientswhoreceivedlimus-elutingstentwithhistoryofmyocardialinfarction andtreatedwithTZDswereassociatedwithahigherriskofmyocardialinfarction (HR 5.292,95%CI 1.02827.232). CONCLUSIONS: OurndingssuggestthatTZDs couldnotimproveclinicaloutcomesintype2diabetespatientsafterdrug-eluting stentimplantation.TZDsmaycontributetohigherriskofmyocardialinfarctionin patientwithlimus-elutingstentandhistoryofmyocardialinfarction.ForthepleiotropiceffectsofTZDs,balancebetweenbenetandriskforcardiovascularevents todifferentsubgroups,maybedifferent.Furtherstudiesarerequiredtoinvestigate thisrelationship. PCV27 ISANEMIAASSOCIATEDWITHHEARTFAILURE?NHANESDATABASE(20052006)ANALYSISAlsalmanAJ,AlhammadAM,AlhashemYN,AlHawajMA,VanTassellBW,HarpeSE VirginiaCommonwealthUniversity,Richmond,VA,USAOBJECTIVES: Anemiaisaprevalentpublichealthproblemassociatedwithincreasedriskofmorbidityandmortality.Itisalsoassociatedwithcardiovascular disorders(CVDs)includingheartfailure(HF).Inalargecommunitystudy,58%ofHF patientshadanemia.ThereisawidevariationinthereportedprevalenceofanemiaamongHFpatients(7%-50%).Thiscommunitybasedstudyinvestigated whetheranemiaisassociatedwithHFusingdatafromNationalHealthandNutritionExaminationSurvey(NHANES)whichisanationalsurveytoassessthehealth andnutritionalstatusintheUnitedStates. METHODS: NHANESdatafromthe 2005-2006administrationwerecombinedandanalyzedcross-sectionally.Descriptivestatisticsandunivariatecomparisonswereusedasappropriate.Weighted multivariablelogisticregressionwasusedtopredictandcomputeoddsratios(OR) and95%condenceintervals(CI)oftheassociationbetweenHFandanemia prevalence. RESULTS: Therewere953respondentswithanemia,and180respondentshadHF.Amongthosewithanemia,theproportionofAfricanAmericanswas signicantlyhigherthanCaucasians(46%vs.31%;p 0.001;respectively).Respondentswithanemiaweresignicantlymorelikelytohavecancer,otherCVDsstatus, andbepregnantthanthosewithoutanemia(p 0.001,p 0.001,andp 0.002,respectively).Inaddition,prevalenceofanemiawassignicantlyhigherinrespondentswithHFascomparedtothosewithotherCVDs(18%vs.4%;p 0.001).Approximately10%ofHFrespondentshadanemiacomparedtojust2%amongthose withoutHF(OR 5.09,CI 5.08-5.11).Thisassociationremainedstatisticallysignificantafteradjustingforage,race,pregnancystatus,cancer,smokingstatus,and otherCVDs(OR 5.8,CI 5.54-5.96). CONCLUSIONS: Accordingtothisstudy,the prevalenceofanemiaamongHFpatientsisrelativelyhigh.AfricanAmericansand thosewithHFaremorelikelytoexperienceanemia.TheassociationbetweenHF andanemiaremainedaftercontrollingforconfounders. PCV28 CARDIOVASCULARRISKFACTORSAMONGCHINESEANDWHITEADULTSIN THEUNITEDSTATESMcDonaldM,ZhouJ,RubinsteinE,MardekianJ Pzer,Inc.,NewYork,NY,USAOBJECTIVES: Currentprevalenceofcardiovascularriskfactorsarelackingamong Chineseadults.Thisstudycontributestoourknowledgebyexaminingcurrent nationalestimatesforhypertension,highcholesterol,diabetesandobesityamong Chineseandwhiteadultsaged20yearsandolder,bygenderandagegroup. METHODS: Cross-sectionalobservationalstudydesign.Analysisofadults20years andoldersurveyedintheNationalHealthInterviewSurveys2006-2008(Chinese,n 853,white,n 51,769). RESULTS: AmongUSChineseadults(2.2million),hypertensionandhighcholesterolareequallyprevalent(19%,age-standardized),but signicantlylowerthanthe27%age-standardizedprevalenceofeachcondition amongthepopulationofUSwhiteadults(176million).AmongChineseandwhite adultsaged65yearsandolder,hypertensionratesaresimilar,56%and59%,respectively,buthighcholesterolprevalenceissignicantlyloweramongChinese adults,33%vs.49%,respectively,drivenbythelowerrateofhighcholesterol amongolderChinesemen(27%)thanolderChinesewomen(40%).Overalldiabetes prevalenceissignicantlyloweramongChinesethanamongwhiteadults(5%vs. 7%),butratesamongolderadultsaresimilar,16%and17%,respectively.Obesity ratesforChineseadultsaged20-39,40-64,and65 are6%,8%,and7%,respectively. Obesityratesforwhiteadultsaged20-39,40-64,and65 are26%,34%,and27%, respectively(p 0.0001forallobesitydifferences). CONCLUSIONS: Overallagestandardizedprevalenceratesofhypertension,highcholesterolanddiabetesare loweramongChineseadults.However,amongolderadults,hypertensionanddiabetesratesaresimilarintheChineseandwhitepopulations.Obesityratesarelower amongChineseadultsofallagegroups.Effectiveapproachesarerecommendedto reduceselectedcardiovascularriskfactorsamongbothgroupsofolderadults. PCV29 IMPROVINGBLEEDINGRISKASSESSMENTFORANTICOAGULANTUSEIN ATRIALFIBRILLATIONLeeEH1,KwongWJ1,CascianoJ2,MartinB3 1DaiichiSankyo,Inc.,Parsippany,NJ,USA,2eMAXHealthSystems,LLC,WhitePlains,NY,USA,3UniversityofArkansasforMedicalSciences,LittleRock,AZ,USAOBJECTIVES: Anticoagulationisrecommendedforatrialbrillation(AF)patients withmoderate-highstrokeriskandlowriskofbleeding.Previousdatahaveshown thatbleedingratesincreaseinparallelwithstrokeriskbecausestrokeandbleeding riskfactorsoverlap.Theobjectiveofthisstudywastoassesswhetheradding ATRIAspecicbleedingriskvariablestotheCHADS2strokeriskvariableswill improvetheabilitytopredictmajorbleeds. METHODS: MedicalclaimsforAFpatients(ICD-9code427.31)withcontinuouseligibility12monthspriortoanAF diagnosisintheMarketScandatabasebetweenJanuary2003toDecember2007 wereanalyzed.DatabeforeanindexAFdiagnosiswereusedtoassessthepresence ofCHADS2(congestiveheartfailure,hypertension,age 75,diabetes,priorstroke/ transientischemicattack)andATRIA(anemia,renaldisease,age 75,priorbleeding,hypertension)riskfactors.ThePearsoncorrelationcoefcientbetweenthe CHADS2andATRIAriskscoresandCoxproportionalhazardsregressionscompar ingareducedmodel(CHADS2covariatesonly)andafullmodel(CHADS2 ATRIA covariates)topredictmajorbleedsafteranAFdiagnosiswereestimated.Likelihoodratiotestswereusedtotestfordifferencesbetweenthemodels. RESULTS: A totalof64,946AFpatientswereincludedintheanalysis(47% 75yearsofage,45% females).ATRIAandCHADS2scoreswerecorrelated(r 0.63,p 0.0001).CHADS2riskfactorspredictedmajorbleeding(LR 444.66, P 0.001).TheadditionofATRIA riskfactorsenhancedpredictivepower(LR 519.65, P 0.001)andresultedinasignicantimprovementinmodelt(P 0.001).Arangeofsensitivityanalysesrevealedsimilarndings. CONCLUSIONS: AddingATRIAbleedingriskfactorstothe CHADS2signicantlyimprovedthemodelt.Theseresultssuggestthatinforma tionprovidedbytheATRIAbleedingriskindeximprovesbleedingriskassessment andmaypotentiallyhelpoptimizeanticoagulationdecisions.Furtherresearchis neededtoassessimprovementsinmodeldiscrimination. PCV30 AMETA-ANALYSISOFRISKFACTORSTODEVELOPCARDIOVASCULAREVENTS ANDSTROKEINSUBJECTSEXPOSEDTOENVIROMENTALTOBACCOSMOKEVillasis-KeeverM1,Rend—n-MaciasM1,Escamilla_Nu–ezA1,Mould-QuevedoJF2, Muci–o-OrtegaE3 ,Galindo-SuarezRM3 1CentroMŽdicoNacionalSigloXXI,InstitutoMexicanodelSeguroSocial,MexicoCity,Mexico,2Pzer,NewYork,NY,USA,3Pzer,MexicoCity,MexicoOBJECTIVES: Passivesmokingisaccountableeachyearforapproximately53,000 and22,000deathsintheUnitedStatesandEurope.Theaimofthisstudywasto conductameta-analysisofcohortandcase-controlstudiestodetermineriskfactorstoshowcardiovasculareventsandstrokeinpassivesmokers. METHODS: A systematicreviewwasconductedinFebruary2010throughMedline,EMBASE,and theCochraneCollaboration.Selectioncriteriavariableswere:case-control,cohort andcross-sectionalstudieswereallstudiesmustincludeexposedandnotexposed subjectstoenvironmentaltobaccosmoke.Twoindependentreviewersidentied theabstracts,selectedthefullarticles,andextractedthedata.Oddsratios(OR)and weightedmeansdifferenceswerecalculated.RandomeffectsmodelswereemployedintheanalyseswitheachoftheestimatedvariableswithMeta-Analysisv.2 software. RESULTS: Atotalof821studieswerereviewedand39trialswerenally selected.18/39studieswerecase-controlstudies,16/39studieswerecohortstudies and5/39werecross-sectionalstudies;29/39assessedcardiovasculareventsand 18/39studiesassessedstrokeevents.Threedifferenttobaccosmokeexposure measurementsweredened:1)17/39studieswererelatedtospousesmokinghabits;2)18/39wereaccordingtohomeandclosedspacesbehaviorexposures;and3) 4/39wereclinicalmeasurements(serumlevelsofnicotineornicotinemetabolite). RelativeRisk(RR)todevelopacardiovasculareventforpassivesmokersincasecontrolstudies(1.466;CI95%1.301-1.651,p 0.001)washigherincomparisonto cross-sectionalstudies(OR1.264;1.203-1.327,p 0.001).RRtodevelopstrokeevents wasalsohigherforcohortstudies(RR1.253;1.038-1.513,p 0.019)comparedto case-controlstudies(OR1.442;1.280-1.625,p 0.001).Nomeaningfuldifferences werefoundbetweenmale/femalepopulations. CONCLUSIONS: Passivesmoking increasestherisktodevelopacardiovasculareventorstrokebyapproximately 30%.Healthcarepolicymakersshouldconsidertheseresultstoimplementpolicies thatenhanceexposurereductionofenvironmentaltobaccosmoke. PCV31 ASSOCIATIONBETWEENHEARTFAILUREANDIN-HOSPITALMORTALITYIN ST-ELEVATIONMYOCARDIALINFARCTIONPATIENTSTREATEDWITH PERCUTANEOUSCORONARYINTERVENTIONPhatakH1,PanX2,KawabataH1 1Bristol-MyersSquibb,Princeton,NJ,USA,2Bristol-MyersSquibb,NewHaven,CT,USAA37VALUEINHEALTH14(2011)A1–A214

PAGE 46

OBJECTIVES: Todetermineassociationbetweenheartfailure(HF)andin-hospital mortalityinST-elevationmyocardialinfarction(STEMI)patientstreatedwithPercutaneousCoronaryIntervention(PCI). METHODS: RetrospectiveanalysisofI3InVisionDataMart(2003-08).Adultenrollees( 18years)withprimarydiagnosisof STEMIwereselectediftheywerea)treatedusingPCIprocedureandb)hadcontinuousenrollmentforatleast6monthspriortotheSTEMI-relatedindexhospitalization(baseline).Enrolleeswereexcludediftheyunderwentcoronaryarterybypassgraftduringindexhospitalization.In-hospitalmoralitywasassessedbasedon analgorithmusingclaimsdata(orlackthereof)forSTEMIpatientsafterindex hospitalization.Pre-existingHFwasdenedaspresenceofaclaimwithrelevant HFdiagnosisatbaseline.NewonsetHFwasdenedaslistingofHFasacomorbidity inthereimbursementclaimsforindexhospitalizationwithoutsupportingevidenceforpre-existingHFatbaseline.Multivariatelogisticregressionwasusedto assessassociationbetweenHFstatusandin-hospitalmortality.Priorisignicance levelof0.05wasusedfortheseanalyses. RESULTS: Overall7261enrolleeswere includedintheanalysis.Outofthese,187(2.6%)hadHFatbaselineand1,075 (14.8%)experiencedHFforthersttime.Outof7261enrollees,351(4.8%)died duringindexhospitalization.Chancesofin-hospitalmortalityincreasedsignicantlywitheitherpre-existingHF(absoluterate(AR):9.6%,OddsRatio(OR):2.3,95% Waldcondence limits(CI):1.0-5.4),newonsetHF(AR:10.7%,OR 3.0,95%CI:2.33.8)or inclusionofpre-existingHFasamajorcomorbidity(AR:16.7%,OR 3.9,95% CI:2.2-6.9)inindexhospitalizationclaims. CONCLUSIONS: PresenceofHF,recordedaseitherapre-existingconditionorasamajorcomorbidcondition,in STEMIpatientsundergoingPCIwasassociatedwithsignicantincreaseinthe in-hospitalmortality. CardiovascularDisorders–CostStudies PCV32 BUDGETIMPACTANALYSISOFINCREASINGLMWH/FXIUTILIZATIONSchillingB1,PowersA2,FariaC2,ChoeY2,BroderM3,BentleyT4 1AspenMedAssets,Denver,CO,USA,2Eisai,Inc.,WoodcliffLake,NJ,USA,3Partnershipfor HealthAnalyticResearch,LLC,BeverlyHills,CA,USA,4PharLLC,BeverlyHills,CA,AlbaniaOBJECTIVES: Lowmolecularweightheparins(LMWH)andfactorXainhibitors(FXI) areusedtotreatandpreventvenousthromboembolicevents(VTE)andmyocardial infarctions(MI).AlthoughhospitalizedpatientshaveincreasedVTE/MIrisk,almost 60%donotreceiveappropriateLMWH/FXImedication.IncreasingLMWH/FXIutilizationwhilecontainingcostshasbecomeanimportantqualityimprovement goal.OurobjectivewastoestimatefromthehospitalperspectivetheannualinpatientcostsofincreasingLMWH/FXIutilization. METHODS: Wedevelopedabudget impactmodelincorporatingeventcostsofdeepveinthrombosis(DVT),pulmonary embolism(PE),andMI,andpharmacycostsfortwoLMWHproducts(dalteparin sodiuminjectionandenoxaparinsodiuminjection)andoneFXI(fondaparinux). InpatienteventcostswereestimatedfromAHRQdataat$10,000,$20,000,and $9,000forpreventableDVT,PE,andMIevents,respectively,andinpatientproduct costswereestimatedusing2010wholesaleacquisitioncostswithmarket-based estimatesofcurrentproductdiscounts.Changesinevent,pharmacy,andtotal hospitalcostswereestimatedforahypothetical500-bedhospitalinwhichLMWH/ FXIutilizationincreasedfrom60%to80%,andmarketsharefollowedtwoscenarios:maintainingcurrentLMWH/FXIshare(90%enoxaparin,0%dalteparin)orcompleteformularyinterchangefromenoxaparintodalteparin. RESULTS: Increasing LMWH/FXIutilizationfrom60%to80%withunchangedmarketsharewoulddecreasecostsofDVTby$153,000(5.5%reduction);PEby$25,000(6.3%);andMIby $12,000(2.4%).Pharmacycostswouldcorrespondingly increase by$308,000(33.3%) foranetincreaseintotalhospitalcosts(eventpluspharmacycosts)of$117,000 (2.5%).Alternatively,increasingutilizationwhilealsoshiftingproductmarket sharetogreaterdalteparinandlessenoxaparinusewouldreducethepharmacy costincreaseto$208,000(22.5%),foranetcostincreaseof$17,000(0.4%). CONCLUSIONS: HospitalscouldpotentiallyimprovetreatmentqualitybyincreasingappropriateLMWH/FXIutilization,andbyshiftingutilizationfromenoxaparin todalteparin. PCV33 USEOFSECONDARYDATASOURCESTOESTIMATEINPATIENTCOSTSAND PAYMENTSFORACUTECORONARYSYNDROMEOhsfeldtR1,BhandaryD2,FoxKM3,GandhiSK2 1TexasA&MHealthScienceCenter,CollegeStation,TX,USA,2AstraZenecaPharmaceuticalsLP, Wilmington,DE,USA,3StrategicHealthcareSolutions,LLC,Monkton,MD,USAOBJECTIVES: Differentcostcomponents(charges,costs,payments)arerequiredto estimatetheeconomicimpactofadrugtherapyoninpatientcarefromahospital andahealthplan(payer)perspective.Toestimatedifferentcostandpaymentsfor acutecoronarysyndrome(ACS)inpatientcarefromdifferentpayer(hospitalvs. healthplan)andbenetdesign[Fee-for-Service(FFS)vs.ProspectivePaymentSystem(PPS)]perspectives. METHODS: ACSdischargeswereidentiedusingdiagnosis-relatedgroup(DRG)codesusingtwodatasources:1)2008MarketScanadministrativeclaimsforhealthplanpayments,and2)2008HealthcareCostand UtilizationProjectNationwideInpatientSample(NIS)forhospitalchargedata. Admissionswereclassiedasmyocardialinfarction(MI),unstableangina(UA),percutaneoustransluminalintervention(PCI),coronaryarterybypassgraft(CABG),and stroke.Cost-to-chargeratioswereusedtoestimatecosttothehospital.MarketScan datawereusedtoprovidepaymentsfrompayersbasedondifferentbenetdesigns (PPS,FFS,Medicareandcommercialcoverage).Costandpaymentcomponentswere estimatedateachDRGlevelandforindividualACSevents. RESULTS: ForACSdischarges(NISn 109,903,MarketScann 85,962),anadmissionwithbothPCIand CABGincurredhighesthospitalcost/planpayment($43,867/$65,543)andUAincurredlowest($4,369/$5,576).Hospitalchargeswereconsistentlyhigherthanplan payments(e.g.,PCI:$52,256vs.$22,828),whereasestimatedhospitalcostswere consistentlylowerthanpayments(e.g.,PCI:$15,902hospitalcostvs.$22,828payments).Medicarepaymentswereconsistentlylowerthancommercialpayments (e.g.,PCI:$17,205vs.$22,828).Detailedcharges,costs,andpaymentsestimatesfor variousACSeventswillbepresented. CONCLUSIONS: TheseACSinpatientcost estimatesfromhospitalandhealthplanperspectivesandfordifferentbenet designswillfacilitateeconomicevaluationsofACSdrugtherapies.Theanalytic approachdemonstratedthefeasibilityandvalidityofusingdifferentsecondary datasourcestoestimateinpatientscostsforvariouspayertypesandbenetdesigns. PCV34 PROPHYLAXISUSEAND90-DAYCOSTSFORVENOUSTHROMBOEMBOLISM, MAJORANDMINORBLEEDINGEVENTANALYSISINHOSPITALIZEDMEDICALLY ILLPATIENTSBaserO1,XieL1,YuceH2,DuJ1,WangL1 1STATinMEDResearch,AnnArbor,MI,USA,2NewYorkCityCollegeof Technology-CUNY/STATinMEDResearch,NewYork,NY,USAOBJECTIVES: Toexaminetheprophylaxisuse,incidenceofVTE,majorbleeding, minorbleedingandassociatedeconomicburdenover90daysinhospitalizedmedically-illpatients. METHODS: Aretrospectivestudy(January01,2005toDecember 31,2007)wasconductedusingasubsetoftheMarketScanHospitalDrugDatabase anditslinkedoutpatientlesfromtheMarketScanCommercialandMedicare Supplementaldatabase.Eligiblepatientswereselectediftheywerecontinuously enrolledintheirhealthplanforatleast180dayspriortoand90daysfollowingthe indexhospitaldischargedate,whichishospitalizationwithamedicallyilldiagnosis.Prophylaxisuseisdenedastheadmissiondateofindexhospitalizationto30 daysafterindexhospitaldischargeandbeforethedateoftheirrstVTEevents. Patients'demographics,healthcarevisitsandcostswerecomparedusingChisquaretestingandstandardizeddifferences.Risk-adjustedtotalhealthcarecosts betweenpatientswitheventsandwithoutwereestimatedusingtheGeneralLinear Model. RESULTS: Inpatientswhowereidentiedasmedicallyill(n 12,077),6,464 (53.52%)receivedanticoagulanttherapyduringtheirhospitalizationanduntil30 daysafterdischarge.Comparedwithpatientswhodidnotreceiveanyanticoagulantprophylaxis,patientswhoreceivedanticoagulantprophylaxishadsignicantlylowerVTEevents(1.47%vs.3.58%,p 0.0001).Althoughtherewasnosignificantdifferenceinratesofmajorbleedingandminorbleeding,afterriskadjustmentforpre-speciedcovariates,patientswithoutcomeeventswere signicantlyassociatedwithhighertotalhealthcarecosts(VTE:$40,523vs.$17,698 p 0.0001;Majorbleeding:$27,430vs.$18,137p 0.0001;Minorbleeding:$25,696vs. $17,410p 0.0001). CONCLUSIONS: Despiteexistingguidelines,fewmedically-ill patientsarereceivinganticoagulantprophylaxis.AppropriateanticoagulantprophylaxisuseresultsinlowerVTEeventratesandtotalfollow-uphealthcarecosts inhospitalizedmedically-illpatients. PCV35 COMPARISONOFMAJORBLEEDINGRELATEDMORTALITY,HEALTHCARE UTILIZATIONANDCOSTSOFPATIENTSWITHNON-VALVULARATRIAL FIBRILLATIONBaserO,DuJ,XieL,BaserE STATinMEDResearch,AnnArbor,MI,USAOBJECTIVES: Tocomparemortality,healthcareutilizationandcostburdenofpatientswhosufferedmajorbleedingduringthe180daysafterdiagnosisofnonvalvularatrialbrillation(NVAF)withpatientswhodidnot. METHODS: Basedon the2005-2007U.S.Medicareadvantageinsuranceclaimles,patientsaged65years andolderwhohavehadtwoormoreprimarydiagnosesforNVAFoccurringwithin 30daysofoneanotherwereselected.The180-dayfollow-upmortality,healthcare facilityuseandcostsforpatientswithandwithoutmajorbleedingwerecompared. Riskadjustmentwasperformedusingthepropensityscorematchingmethodwith theProbChoicealgorithm. RESULTS: Outofthepatientswhowereidentiedwith NVAF(n 18,575),266(1.43%)sufferedamajorbleedingduringthe180daysafter NVAFdiagnosis.Patientswerenotsignicantlydifferentintermsofgender,region, andbaselinecomorbidconditions.Afterrisk-adjustmentforpre-speciedcovariates,mortality(9.77%vs.0.38%p 0.0001),outpatientemergencyroom(ER)visits (84.21%vs.43.23%p 0.0001),ischemicstroke(44vs.8/100personyears),myocardialinfarction(10vs.2/100personyears)andosteoporoticfracture(7vs.1/100 personyears)wereallhigherforpatientswhosufferedamajorbleedingcompared tothosewhodidnot.Besidesinpatientcosts($26,568vs.$8,929),risk-adjusted outpatientERcosts($1,280vs.$744)werealsohigherformajorbleedingpatients. Theoverallrisk-adjusteddifferenceinhealthcarecostsissignicant($35,691vs. $10,480p 0.0001). CONCLUSIONS: Mostoftheadverseeventsanalyzedwere higherforpatientswhosufferedamajorbleedingafterNVAFrelativetopatients whodidnot.Totalhealthcareutilizationandcostswerealsosignicantlyincreased. PCV36 EVALUATIONOFTHEHOSPITALRESOURCEUTILIZATIONASSOCIATEDWITH TOLVAPTANUSAGEAMONGHEARTFAILUREPATIENTSWITHHYPONATREMIA FROMTHEEVERESTTRIALDastaJF1,ChiongJR2,KimS3,LinJ4 1OhioStateUniversity,Columbus,OH,USA,2LomaLindaUniversity,LomaLinda,CA,USA,3OtsukaAmericaPharmaceutical,Inc.,Princeton,NJ,USA,4NovosysHealth,Flemington,NJ, USAOBJECTIVES: TolvaptanisanorallyadministeredselectivevasopressinV2-receptorantagonistforhyponatremiatreatment.TheEfcacyofVasopressinAntago-A38VALUEINHEALTH14(2011)A1–A214

PAGE 47

nisminHeartFailureOutcomeStudywithTolvaptan(EVEREST)trialshowedthat tolvaptan,combinedwithstandardtherapy,improvedmanyheartfailuresigns andsymptomswithoutseriousadverseevents.Thisstudyevaluatedthehospital resourceutilizationassociatedwithtolvaptanusageamongheartfailure(HF)patientswithhyponatremiabasedontheEVERESTtrial. METHODS: Acostoffset modelwasconstructedtoevaluatetheimpactoftolvaptanonhospitalresource utilizationamongHFpatients.TheHealthcareCostandUtilizationProject(HCUP) 2008NationwideInpatientSample(NIS)databasewasusedtoestimatehospitalizationlengthofstay(LOS)andhospitalcost,forHF-associateddiagnosisrelated grouphospitalizations(DRG)ofadultpatients(age 18years).EVERESTtrialdata forpatientswithhyponatremiawereusedtoestimatethereductionofLOSassociatedwithtolvaptanvs.placebo. RESULTS: AmongEVERESTtrialHFpatientswith hyponatremia( 135mEq/L),tolvaptanpatientshadashorterhospitalizationLOS thanplacebopatients(9.72vs.11.44days,respectively),witharelativeLOSreductionof15%.933,189HF-associatedDRGhospitalizationswereidentiedfromthe HCUPNISdatabase.ThemeanLOSwas4.8dayswithmeantotalhospitalcostsof $7,545,andmeandailyhospitalcostsof$1,562.Usinganinpatienttolvaptantreatmentdurationof3dayswithadailywholesaleacquisitioncostof$250,thecost offsetmodelestimatedaLOSreductionamongUSHFhospitalizationsof0.73days withahospitalcostreductionaveraging$1,134perHFadmission.Thecostneutral breakevenmeandurationoftolvaptaninpatienttherapyis4.54days. CONCLUSIONS: BasedontheEVERESTtrial,tolvaptanisassociatedwithashorter hospitalizationLOSthanplaceboamonghyponatremicHFpatients,resultinginan estimatedmeanhospitalcostreductionof$1,134peradmissionintheUS. PCV38 POTENTIALECONOMICIMPACTOFDYSPNEAASSOCIATEDWITHTICAGRELOR USEBonafedeM1,JingY2,GdovinBergesonJ2,LiffmannDK3,MakenbaevaD2,GrahamJP2, DeitelzweigS4 1ThomsonReuters,Andover,MA,USA,2Bristol-MyersSquibb,Plainsboro,NJ,USA,3Thomson Reuters,Washington,DC,USA,4OchsnerHealthSystem,NewOrleans,LA,USAOBJECTIVES: Anti-platelettherapyisrecommendedforpatientswithahistoryof acutecoronarysyndrome(ACS),includingaspirinandclopidogrelorprasugrel. Ticagrelorisanewanti-plateletagent,notyetavailableonthemarket;inclinical trialsticagrelorwasassociatedwithsignicantlyhigherratesofdyspneathan clopidogrel.ACSpatientspresentingwithdyspnearequireadditionalmedicalattentiontoruleoutpossibleheartfailurewhichmayleadtoincreasedcosts.This studyusedrealworlddatatoquantifythedirectmedicalcostsofdyspneaamong patientswithahistoryofACS. METHODS: Patientswithanemergencyroom(ER) visitfordyspnea(ICD-9-CM786.0x)in2008or2009wereidentiedusingtheMarketScanResearchDatabases¨.Patientswererequiredtohavesixmonthscontin uousmedicalenrollmentpriortotheERvisit,withahistoryofACS(ICD-9-CM 410.xx,411.1x).ProcedureutilizationandexpenditureswereevaluatedfortheER visitandassociatedoutpatientservices,aswellastheproportionofERvisitswhich ledtoaninpatientstay.Costswereadjustedto2009USdollars. RESULTS: Atotalof 8,433ERvisitsfordyspneawereidentied.Theaveragecostperdyspneaepisode was$6,958,ofwhich$1,621wasoutpatientcostsassociatedwiththeERvisit(SD $3,269).Alongwithphysicianservices,assessmentofdyspneaoftenincludedelectrocardiogram(71.3%),chestradiograph(75.9%)and,occasionally,aB-typenatriureticpeptide(BNP)test(14.9%)orchestCATscan(12.2%).Theseproceduresconstituted35.1%oftheaverageoutpatientcosts.Overaquarter(25.8%)ofdyspneaER visitsledtoaninpatientstay,withanaveragecostof$20,693perpatient. CONCLUSIONS: Dyspneaisasignicanteventassociatedwithhighmedicalresourceutilizationandhospitalcosts.Consideringthattheincreasedriskofdyspneaforticagreloriswell-documented,thesecostsmaybeimportanttohealth plandecisionmakerswhenevaluatingcostsassociatedwitheachanti-platelet therapy. PCV39 CHARACTERIZATIONOFCOSTSASSOCIATEDWITHSTROKEREHABILITATION: AREVIEWOFTHELITERATURESaccoP,JohnsonJ,VermaA NovartisPharmaceuticals,EastHanover,NJ,USAOBJECTIVES: Theeconomicburdenofstrokeisapproximately3%oftotalhealthcarecostsor0.3-1.8%ofGrossNationalProductindevelopedcountries.Identifying whatdrivesvalueandcostinpost-strokecarecanadvancedevelopmentofcosteffectiveinterventionsthatimprovelongtermoutcomes.Thisreviewcharacterizescostsassociatedwithstrokerehabilitation. METHODS: PUBMEDwassearched between2000and2010;searchtermsincludedstroke,rehabilitation,costs,pharmacoeconomicandprice.CountriesofinterestwereCanada,China,France,Germany,Italy,Japan,Spain,UK,andUS.Dataweresynthesizedqualitatively. RESULTS: Of941abstracts,28metinclusioncriteria.Manualsearchingofretrieved papersandwebsitesyielded4additionalsources.Themainfactordetermining costwasthecountryinwhichthestudywasconducted.Patientswithseverestroke areroutinelyconsideredforrehabilitation;thosewithmild-moderatestroke(60 75%ofstrokesurvivors)aretypicallydischargeddirectlytohome.Ofallstroke patients,only10-15%receiverehabilitation.Countriesspendingalargeproportion onlongtermcareforpatientspost-stroke,spentproportionallylessoninpatient care.Inpatientcostsaccountforanaverageof76%ofcostsintherstyearfollowingstroke.Rehabilitationcostsareoftenincludedininpatientcosts,withthemajorityofrehabilitationcostsoccurringinyear1(82%forpatientswithhemiparesis; 67%forpatientswithouthemiparesis).Evidencesuggeststhatstrokeunitcare improveslongtermoutcomesascomparedtoconventionaltreatment.Thereare fewdataonindirectcostsdespitetherecognitionthattheyarelikelytobe substantial. CONCLUSIONS: Coststudiesinstrokeareheterogeneous,lackspecic economicdataandrarelyincludelongtermrehabilitationcostofcare.Future studiesshouldaimtodifferentiate1)directcostsofvariousrehabilitationservices, 2)locationforprovisionofrehabilitationservices,and3)indirectcoststodemonstratecost-effectivenessandvalueofrehabilitationonlongtermoutcomes. PCV40 COSTANDPRACTICEOFTREATMENTOFACUTEISCHEMICSTROKEINREAL WORLDSETTINGSINRUSSIAVorobyevP,BezmelnitsynaL,BorisenkoO RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,RussiaOBJECTIVES: Toevaluatetypicalpracticeandcostoftreatmentpatientswithacute ischemicstrokeinleadingmedicalhospitalsinRussia,tocomparetreatmentpatternsandcostoftreatmentin"realworldsettings"andaccordingtonational standardofcareinstroke. METHODS: Retrospectiveanalysisofcaserecordsof patientswithacuteischemicstroke.Treatmentpatternsin"realworldsettings" andaccordingtonationalstandardofcareinstrokewerecompared.Directcostof treatmentwasestimatedfromstatehealthcaresystempointofview. RESULTS: Datafrom140caserecordswasretrieved.Signicantinconsistenciesinthelistof medicinesbetweenprotocolofcareandtypicalpracticehavebeenidentied.Particularly,thefrequencyprescriptionofanticonvulsantsandmyorelaxantsinreal worldsettingswaslowerthaninprotocol.Lowfrequencyofpsychological(0%-in typicalpractice,80%-inprotocol)andrehabilitationmethodsoftreatment(33%intypicalpractice,70-80%-inprotocol)wasrevealed.Patientsalsodidn'treceive prophylacticsoftromboembolism(6%-intypicalpractice,80%-inprotocol).Averagetotalcostoftreatmentofonepatientwithischemicstrokeintypicalpractice was905dollars(302dollarscostofdrugtherapy,187dollarscostofmedical diagnosticandtherapeuticservices,416dollarsaccommodationcosts).Average periodofinpatientstaywas20days.Averagetotalcostoftreatmentofonepatients withischemicstrokeaccordingtotheprotocolwas1446dollars(322dollarscost ofdrugtherapy,686dollarscostofdiagnosticandtherapeuticservices,438dollars-accommodationcosts). CONCLUSIONS: Signicantinconsistenciesofmanagementpatternsinpatientswithischemicstrokeintypicalpracticeandnational standardhavebeenrevealed.Directcostsofinpatientcareintypicalpracticeare approximatelytwicelowerthancostscalculatedfromnationalprotocolofcare. PCV41 ANESTIMATEOFSOCIETALMONETARYBENEFITSOFSIMVASTATININ CANADAThanhNX,ChuckA,OhinmaaA,JacobsP InstituteofHealthEconomics,Edmonton,AB,CanadaOBJECTIVES: Toestimatethesocietalmonetarybenetofsimvastatintothree separategroups:1)developingmanufacturer(DM),2)genericmanufacturer(GM) and3)society. METHODS: ThemonetarybenetsforDMandGMwereestimatedby calculatingannualrevenuesbasedonIMSsalesdatabetween2000and2009.We usedadynamicMarkovmodeltoestimatethemonetarybenetsforsocietyover thesametimeperiodintermsofcostavoidanceassociatedwithpreventedcardiovasculareventsincludingstrokeandmyocardialinfarctionandlostproductivity duetodisabilityandprematuredeathintheworkingpopulation.Inputestimates fordynamicpopulationswerederivedfromtheIMSsalesdata.Inputestimatesfor preventedadverseeventsandcostswerederivedfromsystematicreviewand meta-analysiswhenpossible.Inputestimatesforreductioninhealthcareutilizationanddisabilitydayswereestimatedfrommultivariateregressionsusingdatain the2005CanadianCommunityHealthSurvey.Allcostsandbenetswereexpressedin2010Canadiandollars. RESULTS: Thecumulativesocietalmonetary benetofsimvastatinoverthe10years(from2000to2009)wasestimatedat$4 billion.Ofthis,DMaccountedfor23%,GM31%,andsociety46%(healthcare34% andproductivitylosses12%).Ofnote,thetrendwasdifferentamongthethree parties.ThebenettoDMlevelledoutfrom2002whenthepharmaceuticalwasno longerunderpatentprotectionwhenthebenettoGMstartedtoincrease.The benettosocietyincreasedconsistentlyoverthestudyperiod. CONCLUSIONS: The resultsindicatethatsimvastatinisassociatedwithsignicantbenettobothsociety andindustry,includingDMandGM.Acomprehensiveevaluationofbenetsofpharmaceuticalinnovationshouldthereforeconsiderbenetstoallbeneciaries. PCV42 ADECISIONMODELINGAPPROACHTOEVALUATETHECOST-EFFECTIVENESS OFPRASUGRELVERSUSCLOPIDOGRELINPATIENTSWITHPLANNED PERCUTANEOUSCORONARYINTERVENTIONRaoS,LinFJ,OjoO,PatelV,YuS,ZhanL,TouchetteDR UniversityofIllinoisatChicago,Chicago,IL,USAOBJECTIVES: Toevaluatethecost-effectivenessofprasugrelversusclopidogrel,in combinationwithaspirin,inpatientsundergoingplannedpercutaneouscoronary intervention(PCI)fromthehealthcareprovider'sperspectiveintheUnitedStates. METHODS: Second-orderMonteCarlosimulationwasconductedusingTreeAge Pro(2009)followingtheISPORtaskforceguidelinesformodeling.Modelbranches includedPCItype(baremetalstentanddrugelutingstent),CYP2C19polymorphisms,andclinicaloutcomes.Modelinputssuchascosts(2009dollarvalue), age-adjustedqualityoflife,andprobabilitieswereidentiedthroughsystematic literaturereview.AllfuturecostsandQALYswerediscountedby5%.Lifeexpectancywasestimatedusingdecliningexponentialapproximationoflifeexpectancy (DEALE)method.Acceptabilitycurvewasplottedtodeterminethemostcost-effectivestrategyatvariouswillingness-to-pay(WTP)thresholds($0-$250,000/QALY). One-waysensitivityanalyseswereperformedtodetermineifthemodelwassensitivetovariationinprobabilities,costsanddisutilitiesassociatedwithmyocardial infarction(MI),strokeandmajorbleeding. RESULTS: ClopidogreltherapyresultedA39VALUEINHEALTH14(2011)A1–A214

PAGE 48

inlifetimecostsandutilitiesof$17,208and10.4124QALYscomparedwith$16,780 and10.4057QALYsforprasugreltherapy.TheICERforclopidogrelwas$63,840/ QALYs.Theacceptabilitycurveshowedthatprasugrelwasnotlikelycost-effective with 80%certaintyatanyWTPthreshold.One-waysensitivityanalyses(WTP decisionthreshold:$100,000/QALY)showedthatprasugrelisthemostcost-effectivestrategywhenprobabilityofMIisincreasedby 12%,probabilityofbleedingis decreasedby 24%,anddisutilityassociatedwithMIis 0.1634.WhenonlypatientswithvariantCYP2C19wereconsidered,theICERwasfoundtobe$2,313,333/ QALYforclopidogrel. CONCLUSIONS: Inconclusiveresultsindicatethatthereisno benetinprescribingonetherapyovertheotherfortheentirepatientpopulation. CYP2C19polymorphismshouldbegivenconsiderationduringthedecisionmaking process.Forthebase-casescenario,prasugreltherapywasthepreferredstrategyin patientswithvariantCYP2C19. PCV43 COST-EFFECTIVENESSANALYSISOFRIVAROXABANVERSUSDABIGATRAN ANDENOXAPARINFORTHEPREVENTIONOFVENOUSTHROMBOEMBOLISM AFTERTOTALHIPREPLACEMENTVorobyevP1,KrasnovaL2,BorisenkoO1,LukyantsevaD2,BashlakovaE2 1RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,Russia,2Moscow StateMedicalUniversitynamedafterI.M.Sechenov,Moscow,RussiaOBJECTIVES: Toevaluatethecost-effectivenessofrivaroxabancomparedwith dabigatranandenoxaparinfortheprophylaxisofvenousthromboembolismin patientsundergoingelectivetotalhipreplacement(THR)inthecontextof Russian healthcaresystem. METHODS: Adecision-treemodelonthechoiceofregimens forthromboprophylaxisafterTHRwasadoptedfromthemodel,developedby McCullaghet.al.(2009).Primaryoutcomeswasmortality,occurrenceofdistaland proximalDVT,ratesofsymptomaticPE.Incidenceofgastrointestinalbleeding, strokeanddeathwasalsoincludedintothemodel.Delphimethodwasusedto determinetypicalpracticeandcostofmanagementofDVTandPE.Itwasassumed thatpatientswithDVTweretreatedfor90days,patientswithPEfor180days.All patientsinthemodelreceivedthromboprophylaxiswithoneofthefollowingregimens:rivaroxabandoseof10mg/dayorallyfor31-39days(RECORD2);dabigatran doseof220mg/dayorallyfor28-35days(RE-NOVETE);enoxaparindoseof40mg/ daysubcutaneouslyfor10-14days(RECORD2).Incrementalcost-effectivenessratios(ICERs)werecalculated. RESULTS: Thecostofprophylaxiswithenoxaparin was6991USD,withdabigatran-7076USD,withrivaroxaban-7147USD.Although rivaroxabanhasmoreeffectivenessinpreventingDVT(0.016vs.0.082vs.0.045) andPE(0.0012vs.0.005vs.0.004)thanenoxaparinanddabigatrancorrespondingly. ICERtoprevent1caseofdeepveinthrombosisafterTHRinrivaroxabanversus enoxaparinwas23.6USD,andindabigatranversusenoxaparinwas22.9USD.ICER toprevent1caseofpulmonarythromboembolismafterTHRinrivaroxabanversus enoxaparinwas556.7USD,andindabigatranversusenoxaparinwas850.3USD. CONCLUSIONS: DespiteofhighercostofprophylaxisofDVTandPEwithrivaroxaban,comparingtoenoxaparinanddabigatran,prophylaxiswithrivaroxabanwas moreeffectivewithacceptableICERs. PCV44 CHANGINGCOST-EFFECTIVENESSEARLYINTHEPRODUCTLIFECYCLE:THE EXAMPLEOFCLOPIDOGRELBISULFATEvanderGoesDN1,WillkeRJ2,GarrisonL3 1UniversityofWashington,Seattle,WA,USA,2Pzer,Inc.,NewYork,NY,USA,3Universityof WashingtonDepartmentofPharmacy,Seattle,WA,USAOBJECTIVES: Newmedicinesarelaunchedbasedonalimitednumberofclinical trials.Payers,providers,andpatientsbasetheirutilizationdecisionsonthisinformation.Furthermore,projectedcost-effectivenessandrealworldcost-effectivenessaremovingtargetsthatmaydiffergreatly,especiallyearlyintheproductlife cycle.Weexaminethisdivergenceusingclopidogrel,anantiplateletmedicationfor preventingstrokesandheartattacks,asacasestudy. METHODS: UsingtheNationalAmbulatoryMedicalCareSurvey(NAMCS)datafrom1998to2008,weestimatechangesinthevolumeanddistributionofpatientprescriptionsbyage, gender,andrace.Wecombinethesetimetrendswithestimatesfromthe pharmacoeconomicsliteratureoncost-effectivenessmeasuredascostperqualityadjustedlifeyear(QALYs). RESULTS: From1998(followingapprovalinNovember 1997)to2001,theaverageageofclopidogrelpatientsdroppedfrom77to70.Over thesameperiod,thepercentageofpatientswhowereunder54yearsofageincreasedfrom0%to13%.Comparingthereal-worldpatientpopulationtoapivotal phase3trialalsorevealsalargedifferenceingendermix:72%maleinthetrialvs. 45%inNAMCSin1998.Similartrendswerefoundforrace:by2008,patientswere muchlesslikelytobewhitethaninthetrialonly82%versus95%.In1999,1.7 millionofcevisitsincludedaprescriptionforclopidogrel,7.5timesasmanyasin 1998.By2008,almost16millionprescriptionswerewritten1.5forevery100ofce visits.Adjustingfordemographicmix,theestimatedreal-worldCEimprovedbetween2000and2001by16%($25,000perQALYvs.$21,000,respectively).From1998 to2008,theCEratiofellby23%. CONCLUSIONS: Theseresultsdemonstratethat cost-effectivenessprojectedatlaunchmayprovideonlylimitedindicationoftheultimatereal-worldimpact,whichimprovedsubstantiallyovertime,inthisexample. PCV45 COST-EFFECTIVENESSOFCYP2C9ANDVKORC1GENOTYPE-GUIDED WARFARINANTICOAGULATIONCARE:THEIMPLEMENTATIONOFDISCRETE EVENTSIMULATIONMODELONTHENATURALHISTORYOFVENOUS THROMBOEMBOLISMTeschemakerR1,LawrenceW2,WutohAK3 1HowardUniversity,Washington,DC,USA,2AgencyforHealthcareResearchandQuality (AHRQ),Rockville,MD,USA,3HowardUniversity,Hyattsville,MD,USAOBJECTIVES: Toevaluateandcomparedthelong-termcostsandoutcomesoffour warfarintreatmentstrategiesofCYP2C9,VKORC1,bothCYP2C9andVKORC1genotype-guideddosing,andstandardwarfarindosingamongnonvalvularVTEpatientsinthesocietalperspective. METHODS: Adiscreteeventsimulationmodel wasusedtodepictpatients'healthstatesasthediseaseevolveswithtime,and captureditsassociatedcosts(2007U.S.dollars)andqualityoflife.DatawasextrapolatedwiththecriteriaofincludingVTEpatientsofage 18yearsonwarfarin withINRtargetof2-3.Probabilities,costsandhumanisticpropertieswereobtained fromtheliterature,HCUP(NIS&SEDD),andtheMedicareReimbursementScheduledatabases.Sensitivityanalysiswasperformedforuncertaintyparametersin themodel.Allcostsandbenetswerediscountedat3%. RESULTS: Therewasa signicantdifferenceintheprevalenceofbleedingcomplicationbetweenstandard anticoagulation(6.1%)andthegenotypeguidedofCYP2C9andVKORC1groups ( 5.8%).ThemeancostandQALYsperpatientswere$14,340and8.1251.The genotype-guidedwarfarinanticoagulationstrategiesprojectedhighercostand higherQALYs.However,consideringthethresholdof$100,000/QALY,VKORC1genotype-guidedwasindicatedtobecost-effectiveamongallstrategies.Sensitivity analysisdemonstrated25%ofthereplicationsofbothCYP2C9andVKORC1genotype-guidedstrategytobe $100,000/QALY. CONCLUSIONS: Thisstudyshowed thattestingformultiplegenotypesofCYP2C9andVKORC1toguidewarfarinanticoagulationtherapyisnotcosteffectiveinallpopulationandthatpatientwith higherriskofcomplicationsaremorelikelytobenetfromthisnewinnovation. Forthegenotype-guidedtesttobecost-effectiveinthepopulationwithVTE,the costofthetestwouldhavetobe $400orberestrictedtopatientathighriskfor bleedingcomplications(RR 5.8). PCV46 LONG-TERMCOSTSANDHEALTHOUTCOMESOFTREATINGACUTECORONARY SYNDROMEPATIENTSWITHTICAGRELORBASEDONTHEEULABEL-COSTEFFECTIVENESSANALYSISBASEDONTHEPLATOSTUDYHenrikssonM1,NikolicE2,JanzonM3,HauchO4,LevinL2,WallentinL5 1AstraZenecaNordic,SšdertŠlje,Sweden,2LinkšpingUniversity,Linkšping,Sweden,3Linkšping University,LinkšpingUniversityHospital,Linkšping,Sweden,4AstraZenecaPharmaceuticalsLP, Wilmington,DE,USA,5UppsalaUniversity,Uppsala,SwedenOBJECTIVES: ThePLATOtrialshowedthatinpatientswithacutecoronarysyndromes(ACS)treatmentwithticagrelorcomparedwithclopidogrelsignicantly reducedtherateofmyocardialinfarction(MI),stroke,ordeathfromvascular causeswithoutasignicantincreaseintherateofoverallmajorbleeding.Basedon clinicalandhealth-economicdatafromPLATOthepresentstudyevaluatesthe long-termcost-effectivenessoftreatingpatientswithticagrelorbasedontheEU label. METHODS: Atwo-partdecision-analyticmodel,comprisingaone-yeardecisiontreeandalong-termMarkovmodel,wasconstructedtoestimatelifetimecosts andQALYsoftreatingACSpatientsforoneyearwithticagrelorplusacetylsalicylic acid(ASA)comparedwithclopidogrelplusASA.Usingindividual-patientdatafrom PLATO,eventrates,health-carecosts(Swedishinbase-caseanalysis),andQALYs wereestimatedfortherstyear.Forthesecondyearonwards,necessaryassumptionsandexternaldatasourceswereutilizedtoextrapolatequality-adjustedsurvivalconditionalonwhetheranon-fatalMI,anon-fatalstrokeornoeventoccurred duringtherstyear.Aprobabilisticanalysiswasperformedandincrementalcosteffectivenessratiosarepresentedfromahealth-careperspectivein2010prices.A genericclopidogrelpriceof 0.17($0.23)perday,andaticagrelorpricerangeof 2.25 ($3.00)to 3.50($4.65)perdaywereapplied. RESULTS: Treatmentwithticagrelor wasassociatedwithaQALYgainof0.13comparedwithclopidogrel.Thecostper QALYgainedwithticagrelorwasintherangeof 2,350($3,110)to 5,700($7,550) comparedwithclopidogrel.Ticagrelorislikelytobecost-savingifproprietary clopidogrelpricesareapplied.Theresultswereconsistentinmajorsubgroups acrossthebroadACSpopulation. CONCLUSIONS: BasedonclinicalandhealtheconomicevidencefromthePLATOstudy,treatingabroadspectrumofACSpatientswithticagrelorforoneyearbasedontheEuropeanlabeliscost-effective comparedwithclopidogrel. PCV47 COST-EFFECTIVENESSOFFONDAPARINUXANDENOXAPARININPATIENTS WITHNONST-SEGMENTELEVATIONACUTECORONARYSYNDROMEIN BRAZILPepeC1,MachadoM2,OlimpioA2,CanellaM2,RamosR3 1MedInsightEvidncias,Campinas,SP,Brazil,2GlaxoSmithKlineBrazil,RiodeJaneiro,Brazil,3InstitutoDantePazzanesedeCardiologia,S‹oPaulo,SP,BrazilOBJECTIVES: Associateduseofantithrombotics,antiplateletsandinvasivestrategiesinpatientswithnonST-segmentelevationacutecoronarysyndromes(NSTEACS)reducescardiovascularevents,however,withanincreaseintheriskofbleeding.Clinicalstudiesshowedthatfondaparinuxisaseffectiveasenoxaparinin treatingpatientswithNSTE-ACS,butwithreducedriskofbleedingevents.The objectivewastodeterminethecost-effectivenessoffondaparinuxversusenoxaparininpatientswithNSTE-ACSinBrazilfromtheperspectiveoftheBrazilian MinistryofHealth(MoH). METHODS: Ananalyticdecisiontreemodelwasconductedtoestimatetheresultantcostsandconsequencesofthetargetedtherapies inpatientswithNSTE-ACS.ModelinputdataderivedfromtheOASIS-5study (N 20,078NSTE-ACSpatientsrandomizedtofondaparinuxorenoxaparin).The analyzedoutcomewasacompositeofcardiovascularevents(i.e.,death,acute myocardialinfarction,stroke,andmajorbleedings).Modeltimehorizonwas9,30, and180dayspost-NSTE-ACS.DirectcostsofNSTE-ACSeventsandtreatments werecomputed(i.e.,drugs,coronaryangiography,myocardialrevascularization, percutaneousintervention-PCI,hospitalizations,etc.).Costswereexpressedin 2010Braziliancurrency(1BRL 0.59USD).Univariateandmultivariate(MonteCarlo) analysestestedmodelrobustness. RESULTS: Atday9,theaveragecostperpatientA40VALUEINHEALTH14(2011)A1–A214

PAGE 49

treatedwas2,575forfondaparinuxand2,688forenoxaparin.Over65%oftotal costswereattributedtotheinvasivetreatment(PCIandrevascularization).Drug costs(in-hospitaltherapies)accountedfor10%(fondaparinux)and12%(enoxaparin)oftotalcosts.Theestimatedratesofcardiovasculareventswere7.3%and 9.0%forfondaparinuxandenoxaparin,respectively.Resultskeptunchangedon days30and180post-NSTE-ACS.Sensitivityanalysisconrmedbase-caseresults. CONCLUSIONS: Fondaparinuxwasdominantoverenoxaparin(lowercosts,better long-termbenets).Thebudgetimpactafter5yearsofanticoagulantsubstitution (at20%constantadoptionrateperyear)couldreach90millionBRLinsavingsfor theBrazilianMoHandhealthcaresystem. PCV48 COST-EFFECTIVENESSANALYSISOFANTIARRHYTHMICTHERAPIESFORTHE TREATMENTOFSUPRAVENTRICULARTACHYCARDIAANDSURGICALLY INDUCEDTACHYCARDIASANDHYPERTENSIONDexterF1,YangJ2,CarltonR3,BramleyT3,HarbG2 1UniversityofIowa,IowaCity,IA,USA,2BaxterHealthcareCorporation,NewProvidence,NJ, USA,3Xcenda,LLC.,PalmHarbor,FL,USAOBJECTIVES: Theobjectiveofthisanalysiswastoestimatethecosteffectivenessof commonlyusedantiarrhythmicagentsforthetreatmentofsupraventriculartachycardia(SVT)andintraoperative/postoperativetachycardiaandhypertension. METHODS: Adecisiontreemodelwasbuilttoexaminethecosteffectivenessof esmolol,metoprolol,diltiazemandamiodaroneforthetreatmentofSVTandintraoperativeandpostoperativetachycardiaandhypertensionfromahospitalperspective.Thedefaultpharmacycostsinthemodelwerebasedonpubliclyavailable wholesaleacquisitioncosts(WAC).Literaturebasedvalueswereusedfortherates andmedicalcostsofadversecardiaceventsincludingmyocardialinfarction, stroke,hypotension,bradycardia,andischemia.Theprimaryefcacyparameter, rateofsuccessfulheartratecontrol,wasbasedonliteraturevalues.Theoutcome wasthecostpersuccessfulheartratecontrolwithincrementalcosteffectiveness ratios(ICERs)calculated.Nodiscountingwasappliedduetotheshorttimeframeof theanalysis.Fortheprobabilisticsensitivityanalysis,aMonteCarlosimulation consistingof1,000simulationswasconductedtotestthejointuncertaintyofall modelingparameterssimultaneously. RESULTS: Thetotalcostoftherapywas $1,250.82,$2,630.19,$2,280.21,and$1,555.14foresmolol,metoprolol,diltiazemand amiodarone,respectively.Therateofsuccessfulheartratecontrolwas90%(esmolol),64%(metoprolol),90%(diltiazem)and74%(amiodarone).Thecostpersuccessfulheartratecontrolwas$1,389.80(esmolol),$4,109.67(metoprolol),$2,533.57 (diltiazem),and$2,101.54(amiodarone).TheICERofesmololdominatedmetoprolol,diltiazemandamiodarone.Intheprobabilisticsensitivityanalysis,esmolol wasthemostcost-effectiveantiarrhythmicin99.6%ofsimulations.One-waysensitivityanalysesshowedthemodelwasmostsensitivetothecostofhypotension andbradycardia. CONCLUSIONS: Inthismodel,esmololwastheleastcostlyand mosteffectiveantiarrhythmic.Esmololiscost-effectiveincomparisonwithmetoprolol,diltiazemandamiodaroneforthetreatmentofSVTandintraoperative/ postoperativetachycardiaandhypertension. PCV49 COST-EFFECTIVENESSANALYSISOFRIVAROXABANVERSUSDABIGATRAN ANDENOXAPARINFORTHEPREVENTIONOFVENOUSTHROMBOEMBOLISM AFTERTOTALKNEEREPLACEMENTVorobyevP1,KrasnovaL2,BorisenkoO1,LukyantsevaD2,BashlakovaE2 1RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,Russia,2Moscow StateMedicalUniversitynamedafterI.M.Sechenov,Moscow,RussiaOBJECTIVES: Patientsaftermajororthopedicsurgeryonthejointsofthelower extremitiesrequireaneffectivethromboprophylaxistopreventdeepveinthrombosis(DVT)andpulmonaryembolism(PE).Objectiveofthisstudywastoevaluate thecost-effectivenessofrivaroxabancomparedwithdabigatranandenoxaparin fortheprophylaxisofvenousthromboembolisminpatientsundergoingelective totalkneereplacement(TKR)inthecontextofRussianhealthcaresystem. METHODS: Adecision-treemodelonthechoiceofregimensforthromboprophylaxisafterTKRwasadoptedfromthemodel,developedbyMcCullaghet.al.(2009). Primaryoutcomeswasmortality,occurrenceofdistalandproximalDVT,ratesof symptomaticPE.Incidenceofgastrointestinalbleeding,strokeanddeathwasalso includedintothemodel.Delphimethodwasusedtodeterminetypicalpracticeand costofmanagementofDVTandPE.ItwasassumedthatpatientswithDVTwere treatedfor90days,patientswithPEfor180days.Allpatientsinthemodelreceive thromboprophylaxiswithoneofthefollowingregimens:rivaroxabandoseof10 mg/dayorallyfor10-14days(RECORD3);dabigatrandoseof220mg/dayorallyfor 12-15days(RE-MODEL);enoxaparindoseof40mg/daysubcutaneouslyfor10-14 days(RE-MODEL).Incrementalcost-effectivenessratios(ICERs)werecalculated. Analyseswasmadefromstatehealthcarepointofview. RESULTS: Thecostof prophylaxiswithrivaroxabanwas5621USD(dominanttechnology),withenoxaparin-5657USD,withdabigatran-5763USD.Rivaroxabanhasmoreeffectiveness inpreventingDVT(0.096vs.0.36vs.0.36)andPE(0.00vs.0.001vs.0.00)than enoxaparinanddabigatrancorrespondingly. CONCLUSIONS: Resultsofmodeling haveshownthatrivaroxabanisdominanttechnologyforpreventionofvenous thromboembolismaftertotalkneereplacementcomparingtoenoxaparinanddabigatraninthescopeofRussianhealthcaresystem. PCV50 COST-EFFECTIVENESSANALYSISCOMPARINGDABIGATRANANDADJUSTEDDOSEWARFARINFORSTROKEPREVENTIONINATRIALFIBRILLATIONZhaoY1,LimL1,ColemanCI2 1Yale-GrifnHospital,Derby,CT,USA,2UniversityofConnecticutSchoolofPharmacy,Storrs, CT,USAOBJECTIVES: Atrialbrillationhasbeenestimatedtoaffectasmanyas2.3million Americans,makingitthesecondmostcommoncardiovascularconditioninthe UnitedStates.Atrialbrillationhasbeenfoundtoincreasepatient'sriskofstroke by5-fold.Wesoughttocalculatetheprojectedtotaltreatmentcosts,quality-adjustedsurvivalandcost-effectivenessofdabigatranandadjusted-dosewarfarinfor strokepreventioninpatientswithatrialbrillation. METHODS: Thisthree-state Markovtransitionmodel(healthywithatrialbrillation,disability,anddeath)simulatedthetreatmentcosts,quality-adjustedsurvivalandcost-effectivenessofdabigatran150mgtwicedailyandadjusted-dosewarfarin(internationalnormalized ratioof2-3)forstrokepreventioninatrialbrillation.Ourbase-caseconsistedofa hypotheticalcohortof 65yearoldpatientswithatrialbrillation,amoderaterisk ofstroke(CHADS2 1)andnocontraindicationstoanticoagulationtherapy.The parametersusedinthemodelwereadoptedfromtheliteratureresearch,Costeffectivenesswascalculatedoverapatient'slifetimeandusingasocietalperspective(excludingindirectcosts).One-wayandthresholdsensitivityanalyseswere performedonallrelevantvariables. RESULTS: Themeanquality-adjustedlifeexpectancyofsimulatedpatientswas12.9and12.2yearsforthosereceivingdabigatranandwarfarin.Totallifetimetreatmentcostswere$146,649and$118,904.The incrementalcost-effectivenessratiowas$40,580.Uponone-waysensitivityanalysis,ourconclusionswerefoundtobesensitivetochangesindabigatrancostand thedifferentialefcacyofthetwostrategies.Thresholdsensitivityanalysisfurther revealedthatdailydabigatrancostsgreaterthan$13perdayanddifferentialefcacybetweenthetwostrategiesoflessthan0.15%peryearresultedinincremental cost-effectivenessratiosgreaterthan$50,000perquality-adjustedlifeyeargained. CONCLUSIONS: Ouranalysissuggestedthatdabigatraniscost-effectiveforstroke preventioninatrialbrillation;however,thisconclusionwassensitivetochanges indabigatrancostsandtheantithromboticefcacyofthetwotreatmentstrategies. PCV51 THEVALUEOFATORVASTATINOVERTHEPRODUCTLIFECYCLEGrabnerM1,JohnsonWR2,AbdulhalimAM1,KuznikA3,MullinsCD1 1UniversityofMarylandSchoolofPharmacy,Baltimore,MD,USA,2UniversityofMaryland SchoolofMedicine,Baltimore,MD,USA,3Pzer,Inc.,NewYorkCity,NY,USAOBJECTIVES: Thecost-effectivenessofadrugisoftenevaluatedatasinglepointin time,yetcosts,effects,orrelevantcomparatorsmaychangeovertheproductlife cycle.Thisstudymodelsthecost-effectivenessofatorvastatinfromproduct launchin1997through2011andintothefuture. METHODS: Wemodeltheyearly cost-effectivenessofatorvastatincomparedtoitsmajorcompetitorsimvastatin from1997to2030fromaUSpayerpointofview.Keyeventsincludetheentryof genericsimvastatininJune2006andtheexpectedentryofgenericatorvastatinin November2011.Estimatesforincrementalcosts(inUSD)andeffects(inQALYs)for primaryandsecondarypreventionofcardiovasculareventsaretakenfrompreviousliteratureandadjustedfordrugpricechangesovertime.TotalstatinuseestimatesarederivedfromNHANES.Sensitivityanalysisexaminesvariationinstudy parametersincludingdrugprices,indicationuse,anddiscountrates. RESULTS: Assumingincreasingstatinuseovertime(withameanof1mnewusersperyear) anda3%discountrate,thecumulativeincrementalcost-effectivenessratio(ICER) foratorvastatinvs.simvastatinrangesfromcostsavingatreleasetoamaximumof $45,066perQALYaftersixyearsofgenericsimvastatinin2012.Overthefullmodeledlifecycle(1997-2030),thecumulativeICERofatorvastatinis$20,331perQALY. Resultsweresimilarinsensitivityanalysis. CONCLUSIONS: TheICERofatorvastatinvariesacrosstheproductlifecycle,risingduringtheperiodbetweengeneric simvastatinentryandgenericatorvastatinentry,anddecliningafterwards.Over itslifecycle,atorvastatinisassociatedwithacumulativeICERof$20,331perQALY, withamaximumof$45,066perQALY. PCV52 COST-EFFECTIVNESSSTUDYOFCITICOLINEINPATIENTSWITHACUTE ISCHEMICSTROKEINRUSSIAVorobyevP,BezmelnitsynaL,BorisenkoO RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,RussiaOBJECTIVES: Cerebrovasculardiseasesareonthesecondplaceamongallcausesof deathinRussia.Neuroprotectivetherapyisoneofthemainapproachesoftherapy inpatientswithacuteischemicstrokeinRussiadespitetheabsenceofthisgroupof medicinesintherecommendationsofnationalstandardofcareinstroke(GOSTR 200).Theobjectiveofthisresearchwastoevaluatecost-effectivenessratioforthe useofciticolin(aneuroprotectiveagentwidelyusedinsomeEUcountries,South Korea,Russiaandsomeothercountries)inpatientstreatedaccordingtothenationalstandardofcareinstroke(GOSTR200). METHODS: Thedataontheefcacy ofciticolininpatientswithacuteischemicstrokewereextractedfrompooling analysisofclinicaltrials"Oralciticolininacutestroke"(DavalosA.etal.,2002)The clinicaleffectwasmeasuredasglobalrecoveryindex(proportionofpatientswith fullrecoveryduring3months).Costoftreatmentwithciticolinandcost-effectivenessratio(CER)werecalculatedfromthepointofviewoftheRussianstatehealth caresystem. RESULTS: Accordingtotheresultsoftheabovementionedpooling analysis,theuseofciticolineinstrokepatientswasassociatedwithsignicantly greaterrateofrecoverythanplacebo(OR,1.33;95%CI,1.10to1.62).Costsoftreatmentofacuteischemicstrokeaccordingtorecommendationsofthenationalstandardplusciticolinwas1715.5USDper3months.Costsofthetreatmentwithout citicolinwas1289USDper3months.CER(i.e.directcostsperonefullyrecovery patient)treatedwithciticolinandplacebowere6354USDand6384USDrespectively.IncrementalCERwas6264USD. CONCLUSIONS: Accordingtotheapplied modelatreatment,citicolinappearedtodemonstrateitsclinicallyefcacyand cost-effectivenessintreatmentofthepatientswithacuteischemicstroke,comparedtoplacebo.A41VALUEINHEALTH14(2011)A1–A214

PAGE 50

PCV53 COST-EFFECTIVENESSANALYSISCOMPARINGIVABRADINEWITHISOSORBIDE MONONITRATE,AMLODIPINE,DILTIAZEMANDVERAPAMIL,INTHE TREATMENTOFSTABLEANGINAPECTORISErgeneO1,ErolMK2,OtoA3,KucukogluS4,OzdemirO5,TanM6 1IzmirAtaturkTrainingandResearchHospital,Izmir,Turkey,2ErzurumAtaturkUniversity SchoolofMedicine,Erzurum,Turkey,3HacettepeUniversitySchoolofMedicine,Ankara,Turkey,4IstanbulUniversityInstituteofCardiology,Istanbul,Turkey,5YorumConsultingLtd.,Istanbul, Turkey,6ServierIlacveArastirmaA.S.,Istanbul,TurkeyOBJECTIVES: Ivabradinepreventsmyocardialischemiabydecreasingheartratein stableanginapectoris(SAP)patients.Thisanalysispresentsthecost-effectiveness analysisofivabradineinthetreatmentofSAP. METHODS: AMarkov-chainmodel, inwhich,patientswithSAPtreatedwithivabradineandcomparatorsarefollowed fortwentyyears,wasbuilt.Annualratesofrevascularization,othercardiovascular eventsandmortalitywerecalculated.Directcostsweretakenintoaccountfrom theperspectiveofsocialsecurityinstitution.Randomizedclinicalstudies,other comparativestudiesandmeta-analysisweretakenassourcesofinputs. RESULTS: Annualrevascularizationratewas5.0%forivabradineand6.5-7.0%forcomparators.Estimatednumberofrevascularizationswithin20yearswas120procedures per100patientswithivabradineascomparedto135-139withothers.Annualcardiovasculareventratewas7.5%forivabradineand10.4-10.9%forcomparators. Totalnumberofcardiovasculareventswithin20yearswas151eventsper100 patientswithivabradineascomparedto207-219withothers.Durationofeventfreelifeyearswas9.9yearswithivabradineand7.7-8.1yearswithothers.IncrementalCostEffectivenessRatioforivabradinedecreaseddowntocost-effectivenessthreshold(threetimesGDPpercapitaperoneevent-freelifeyear)withinthe rstyearversusisosorbidemononitrateandverapamil,andinthesecondyear versusamlodipineanddiltiazem.Annualmortalityratewas1.7%withivabradine, whileitwas2.2%withcomparators.Life-yearsgainedper20yearswithivabradine was0.8years(16.4vs15.7withivabradineandwithcomparators,respectively). CONCLUSIONS: Whensavingsprovidedbythedecreasesinrevascularizationand cardiovasculareventratesaretakenintoaccount,annualtreatmentcostdifferencebetweenivabradineandotherdrugsdecreases.LongerlifeyearsandeventfreelifeyearsprovidedbyivabradinemakesitacosteffectivechoiceforSAP. PCV54 EFFECTIVENESSOFDISEASEMANAGEMENTCAREINTHECASEOFHEART FAILUREUgiliwenezaB UniversityofLouisville,Louisville,KY,USAOBJECTIVES: Themainobjectiveofthisstudywastoevaluatethecosteffectivenessofthemanagedcareinthecaseofheartfailure.Also,theproportionofpatientswithahospitalfreeyearwascomparedusingmanagedcareandnon-managedcarethenon-managedcare. METHODS: Ameta-analysiswasusedtoestimate theeffectiveness(hospitalfreeyears).Costeffectivenesswasanalyzedintwoperspectives:theprogram'sandthepayer's.Intheprogram'sperspective,theliteraturewasusedtoestimatethecost.Inthepayer'sperspective,ThomsonReuter's MarketScandatawereusedtoestimatethecost. RESULTS: Itwasfoundthat,inthe program'sperspective,ahospitalfreeyear'scostwas$8,872.60whileinthepayer's perspectivebyusingmanagedcare;itsavedanaverageof$53,109.22.Theproportionofindividualswithahospitalfreeyearinmanagedcareandintheusualcare werenotfoundtobestatisticallysignicant. CONCLUSIONS: Eventhoughmanagedcaredidnothaveadifferenthospitalizationusagethanstandardcare,itwas foundtobecost-effectiveforthepayer. PCV55 COST-MINIMIZATIONANALYSISOFTHROMBOPROPHYLAXISFOLLOWING ELECTIVETOTALHIPARTHROPLASTYDelateT,ClarkN,ChoS,WittD KaiserPermanenteColorado,Aurora,CO,USAOBJECTIVES: Tomodelthecostsofthromboprophylaxisfollowingtotalhiparthroplasty(THA)witheitherlow-intensitywarfarin,enoxaparin,orfondaparinux. METHODS: Thiswasacost-minimizationanalysisofanticoagulantsexpectedto haveequivalenttolerabilityandeffectiveness.DatawereobtainedfromanHMO's jointreplacementregistry.PatientswereincludediftheyhadTHAbetweenAugust 2005andMarch2009,continuousenrollmentinthe180dayspriortoand90days followingTHA,utilizedlow-intensitywarfarinthromboprophylaxis,andhadnot beenchronicallyanticoagulatedpreviously.Patientsreceivedwarfarintherapysupervisedthroughapharmacist-runanticoagulationservicethatincludedmobile phlebotomyforINRmonitoring.IterativemodelingfromtheHMO'sperspective utilizedregistrydatatocalculatethecostsofthromboprophylaxiswithanticoagulantsat2010AWPminus35%,monitoringat$50perphlebotomy,$15perINR,and 0.1hourpharmacisttime(at$75/hr)perINR.Patienttrainingforanyanticoagulant usewasincludedat0.25hrpharmacisttime.Durationoftherapywasmodeledto ameanof35daysforinjectible(enoxaparinandfondaparinux)thromboprophylaxisforwhichmonitoringserviceswouldnotberequired.Sensitivityanalyses includedreducingAWPsby25%priortodiscountingandincreasingphlebotomy andINRsto$75and$25,respectively. RESULTS: 835THAsfor800patientswere included.Themeanageanddurationwarfarintherapywere66yearsand40days, respectively,and62%werefemales.Thebase-casethromboprophylaxiscostwas $556(95%CI $545-$567),$701(95%CI $693-$709),and$2677(95%CI $2646$2708)forwarfarin,enoxaparin,andfondaparinux,respectively.25%AWPreductionresultedinthromboprophylaxiscostsof$553(95%CI $542-$564),$564(95% CI $558-$571),and$2145(95%CI $2120-$2171)forwarfarin,enoxaparin,and fondaparinux,respectively.Increasingmonitoringcostsresultedinthromboprophylaxiscostof$756(95%CI $740-$771)forwarfarin. CONCLUSIONS: Warfarin thromboprophylaxishadthelowestcostbutissensitivetomonitoringcosts whereasinjectiblethromboprophylaxisissensitivetodrugacquisitioncosts. PCV56 COST-MINIMIZATIONANALYSISOFLANDIOLOLFORCTSCANNINGFORCVD SUSPICIOUSPATIENTSINJAPANIgarashiA1,FujitoK2,FukudaT3 1TokyoUniversityFacultyofPharmacy,Tokyo,Japan,2TokyoUniversityofPharmacyandLife Sciences,Hachioji,Japan,3TokyoUniversity,Tokyo,JapanOBJECTIVES: Toconductacost-minimizationanalysisoflandiololforCTdiagnosis ofcardiovasculardiseases(CVD)suspiciouspatientwithtachycardiainJapan. METHODS: Adecision-treemodelwasconstructedtoanalyzecostsfromhealthcarepayer'sperspective.Drugcostsanddiagnosiscosts,computertomography (CT)andcoronaryangiography(CAG),areadoptedtothemodel.LandiololisadministeredonlytoslowtheheartratetotakeCTimageappropriately.Sincesometrials provedthattherewasnodifferencebetweenlandiololandplacebointermsof efcacyandsafety,weconductedcost-minimizationanalysis.22.4%ofthosewho suspiciousforCVDarethoughttobetakebeta-blockers.SuccessrateforCTscanningforlandiololandplacebo,derivedfromdomestictrialdata,were81.4%(96/118, 77.8%-84.9%)and54.2%(64/118,49.7%-58.8%).PatientswhofailedtotakeCT imagewerethoughttotakeCAG.ThehealthcarecostwerederivedfromJapanese datacostsoflandiolol,CTimaging,CAGareJPY6,500(USD79,USD1 JPY82.7), JPY39,100(USD472)andJPY105,900(USD1,279),respectively.PositiverateforCAG, derivedfromdomestictrialdata,were37.1%(33/89,32.0%-42.2%).Varioussensitivityanalyses,bothunivariateandprobabilisticones,wereconducted. RESULTS: Inthebasecaseanalysis,expectedcostsperpatientforlandiololandplacebowere JPY82,600(USD998)andJPY85,218(USD1,029),respectively.Inbudgetimpactanalysis,49,000patientsareeligibleforlandiololanditcansaveJPY120mil. (USD1.45mil.)forwholepatients.Sensitivityanalysessuggestedtherobustnessof theresults. CONCLUSIONS: LandiololforCTdiagnosisofcardiovasculardiseases (CVD)suspiciouspatientwithtachycardiaisthoughttobecost-saving. PCV57 HEALTHCARERESOURCEUTILIZATIONAMONGT2DMPATIENTSWITHPREEXISTINGMACROVASCULARCONDITIONS:AMATCHEDCOHORTSTUDYQiuY1,FuAZ2 1Merck&Co.,Inc.,WhitehouseStation,NJ,USA,2ClevelandClinic,Cleveland,OH,USAOBJECTIVES: T2DMisachronicmetabolicdisordercharacterizedbyhyperglycemiaandassociatedwithsignicantmorbidity.Thismaybeparticularlysoamong patientswithpre-existingmacrovascularconditions(MVC).ThisstudywasundertakentodifferentiatehealthcareresourceutilizationbetweenT2DMpatientswith andwithoutpre-existingMVCinEurope. METHODS: Thisisamatchedcohort studybasedontheReal-LifeEffectivenessandCarePatternsofDiabetesManagement(RECAP-DM)study,amulti-center,observationalstudywithretrospective medicalchartreviewsofT2DMpatientsineightEuropeancountries.Included patientswereaged 30yearattimeofdiagnosisofT2DMandaddedaSUoraTZD tofailingmetforminmonotherapy(indexdate),andhadpre-existing(i.e.,with onsetdatepriortoindexdate)MVC.AcontrolcohortwithT2DMbutwithout pre-existingMVCwasidentiedusing1:1propensityscorematching.Logitmodels wereusedtoidentifytherelationshipbetweenpre-existingMVCandthelikelihood ofERadmission,receivingmedical/surgicalprocedures,andhospitalizationduring thestudyperiod.Negativebinomialmodelswereusedtopredictthenumberof ofcevisitsandlengthofhospitalstayperyearattributabletothepre-existing MVC. RESULTS: Ofthe453eligiblepatientswithpre-existingMVC(cases),64%were male,meanageandtimefromT2DMdiagnosiswas64.5(9.1)and6.2(5.3)years, respectively.HbA1cpriortotheindexdatewas8.0(1.2).Relativetocontrols,patientswithpre-existingMVCweresignicantlymorelikelytoreportERadmissions (OddsRatio3.1;95%CI1.8-5.2),receivingmedical/surgicalprocedures(OR2.72; 95%CI1.7-4.3),andhospitalizations(OR2.6;95%CI1.7-4.0)aftercontrollingfor otherpredictors.Similarly,pre-existingMVCincurred2.9additionalofcevisits peryear(p 0.001)and0.41daysofhospitalstayperyear(p 0.004). CONCLUSIONS: ThisstudyfoundthatT2DMpatientswithpre-existingMVCweremorelikelytouse varioustypesofhealthcareresources. PCV58 ASSOCIATIONBETWEENHEARTFAILUREANDRESOURCEUTILIZATIONINSTELEVATIONMYOCARDIALINFARCTIONPATIENTSTREATEDWITH PERCUTANEOUSCORONARYINTERVENTIONPanX1,KawabataH2,PhatakH2 1Bristol-MyersSquibb,NewHaven,CT,USA,2Bristol-MyersSquibb,Princeton,NJ,USAOBJECTIVES: Todetermineassociationbetweenheartfailure(HF)andresource utilizationinST-elevationmyocardialinfarction(STEMI)patientstreatedwithPercutaneousCoronaryIntervention(PCI). METHODS: RetrospectiveanalysisofI3InVisionDataMart(2003-08).Adultenrollees( 18years)withprimarydiagnosisof STEMIwereselectediftheywerea)treatedusingPCIprocedureandb)hadcontinuousenrollmentforatleast6monthsprior(baseline)totheSTEMI-relatedindex hospitalization.Enrolleeswereexcludediftheyunderwentcoronaryarterybypass graftduringindexhospitalization.Pre-existingHFwasdenedaspresenceofa claimcomprisingrelevantHFdiagnosisatbaseline.NewonsetHFwasdenedas listingofHFascomorbidityinthereimbursementclaimsforindexhospitalization withoutsupportingevidenceforpre-existingHFatbaseline.GeneralandgeneralizedlinearmodelswereusedtoassessimpactofHF(versusnoHF)onresource utilizationparametersincludinglengthofstay(LOS)andoncosts(log-transformed)associatedwithSTEMI-relatedindexhospitalization,respectively.PrioriA42VALUEINHEALTH14(2011)A1–A214

PAGE 51

signicancelevelof0.05wasusedfortheseanalyses. RESULTS: Overall7,261enrolleeswereincludedintheanalysis.Outofthese,187(2.6%)hadHFatbaselineand 1,075(14.8%)experiencedHFforthersttime.MeanLOSwas4.29( 7.5)daysbut increasedsignicantlyto8.8( 17.6)daysincaseofnewonsetHF( 5.2 0.2, p 0.001)andto9.7( 10.5)daysifpre-existingHFwasincludedasamajorcomorbidity( 6.0 0.7,p 0.001)duringindexhospitalization.Hospitalizationcosts (log-transformed)wereassociatedwithrecordingofnewonsetHF( 0.4 0.02, p 0.001)butnotwithreportingofpre-existingHF( 0.1 0.1,p 0.1451)asa comorbidconditionduringindexhospitalization. CONCLUSIONS: NewonsetofHF, recordedasamajorcomorbidcondition,inSTEMIpatientsundergoingPCIexhibitedsignicantindependentassociationwithresourceutilizationincludingincreasedlengthofstayandgreatercosts. CardiovascularDisorders–Patient-ReportedOutcomes&Preference-BasedStudies PCV59 PERSISTENCEANDCOMPLIANCEINHYPERTENSIVEPATIENTSTREATEDWITH FIXEDORUNFIXEDCOMBINATIONSOFANGIOTENSINRECEPTORBLOCKERS, AMPLODIPINEANDHYDROCHLOROTHIAZIDESandbergA1,KostevK2,EhlkenB3,HolzB2,OberdiekA1 1DaiichiSankyoEuropeGmbH,Munich,Germany,2IMSHealthGmbH&Co.,Frankfurt, Germany,3IMSHealthGmbH&Co.,Munich,GermanyOBJECTIVES: Toevaluatepersistenceandcomplianceofxeddosecombinations incomparisontounxeddosecombinationsofangiotensinreceptorblockers (ARBs)inpatientswithhypertensioninGermany. METHODS: Thisretrospective studyanalyzedprescriptiondatacollectedbygeneralpractitioners,usingalongitudinaldatabase,theGermanIMSDiseaseAnalyzer(DA).TheDAdatabasewas searchedforpatientswithhypertension(ICD-10codeI10)whowereinitiatedon doubleortriplecombinationsofARBswithhydrochlorothiazide(HCT)and/oramlodipine(AML)intheperiod09/2008-08/2009withafollow-upofatleast12months. Persistencewasdenedasproportionofpatientswhoremainedontheirinitially prescribedtherapyat1year.Thedifferencebetweenmeanpersistencevalues (days)werecalculatedbyusingmultipleregressionanalysesadjustedbyage,gender,region,insuranceandco-morbidity.Compliancewasmeasuredindirectly basedonthemedicationpossessionratiocalculatedasnumberofdayssupplied withintherellintervalinrelationtothenumberofdaysintherellinterval. RESULTS: Overall,17,310patientswereeligibleforanalysis:35.7%onaxeddouble ARBcombination,33.0%onanunxeddoubleARBcombination,25.9%onasemixedtripleARBcombination,and5.4%onanunxedtripleARBcombination. Twelvemonthsafterrstprescriptionpersistenceonunxedversussemi-xed ARB/HCT AMLtriplecombinationwas32.7%and46.7%respectively,unxedversussemi-xedARB/AML HCTtriplecombinationwas41.5%and49.0%respectively,unxedversusxedARB/HCTdoublecombination23,8%and41.1%respectively,unxedversusxedARB/AMLdoublecombination27.5%and44.4% respectively.Meandaysofpersistencewashigherinpatientsreceivingxedcombinations(225.7-242.9)comparedtounxedcombinations(163.6-228.0).Mean compliancevaluewithxedcombinationsrangedbetween72.0%-79.0%,withunxedcombinationsbetween71.5-76.2%. CONCLUSIONS: Thesereal-lifedataconrmspreviousstudyresults,thatxedcombinationsimprovepersistenceinpatientscomparedtounxedcombinationsbecauseofreducedpillburden. PCV60 MEDICATIONADHERENCE,PERSISTENCE,ANDHEALTHCARECOSTSFOR PATIENTSINITIATINGSTATINTHERAPYINAMANAGEDCAREPOPULATIONZhaoZ1,PengX1,BaeJP1,DungeyJ1,FariesDE1,SponsellerCA2,WetmoreS1,YuCY1, LeNarzLA1 1EliLillyandCompany,Indianapolis,IN,USA,2KowaPharmaceuticalsAmerica,Inc., Montgomery,AL,USAOBJECTIVES: Comparestatinadherence,persistence,andhealthcarecostsamong patientsinwhomatorvastatin(AS),simvastatin(SS),rosuvastatin(RS),orpravastatin(PS)wasnewlyprescribed. METHODS: Studypatients 18yearswereidentiedbetweenOctober1,2006-September30,2007,withaminimum12-monthpreand24-monthpost-indexhealthplaneligibilityfromamanagedcaredatabase.A 6-monthwashoutperiodfreeofanystatinusewasappliedtoidentifypatientsin whomstatintherapywasnewlyprescribed.Theindexdatewasdenedasthedate oftherststatinprescriptionandfourcohortswerecreatedbasedonindexstatins: AS,SS,RS,andPS.Thesecohortswerematchedaccordingtopropensityscores basedonpatientclinicalanddemographiccharacteristics.Outcomeswerecomparedamongmatchedcohortsduringthe2-yearpost-indexperiod,includingadherencetoanystatin(medicationpossessionratio 0.8),timetoanystatindiscontinuation,andhealthcarecosts. RESULTS: Aftermatching,6185patientswere obtainedineachcohortandthebaselinecharacteristicswerebalancedamong cohorts.Halfofthepatientswereadherenttotheirstatintherapyat12months (52.2%forSS,47.3%forAS,48.6%forRS,and49.7%forPS,P 0.01),withfurther declineat24months(43.1%forSS,38.0%forAS,39.2%forRS,and38.5%forPS, P 0.01).Mediantimetoanystatindiscontinuationwithin24monthswereless than1yearforallcohorts(SS:308days;AS:250days;RS:266days;PS:301days, P 0.01).Thecohortshadslightlydifferent2-yearhealthcarecosts(SS:$15,417;AS: $16,826;RS:$15,874;PS:$15,481,P 0.01)withthemostdifferenceattributableto thelowerdrugcosts(SS:$4460;AS:$5485;RS:$5424;PS:$4586,P 0.01). CONCLUSIONS: Patientshadlowadherenceandpersistenceratestostatintherapy.Patientsinitiatingdifferentstatinshadvaryingadherenceandpersistence proles,including2-yearhealthcarecosts. PCV61 THEIMPACTOFTHEMEDICAREDOUGHNUTHOLEONMEDICATIONUSE BEHAVIORSANDHEALTHCARERESOURCEUTILIZATIONAMONG BENEFICIARIESTAKINGSTATINMEDICATIONJoshiN,BanahanBFI, UniversityofMississippi,University,MS,USAOBJECTIVES: TheobjectivewastodeterminetheimpactofMedicarepartDdoughnutholeorthecoveragegap'onmedicationcompliancebehaviorandhealthcare resourceutilizationamongMedicarebeneciariestakingstatinmedications. METHODS: Thiswasaretrospectivecohortstudyusinga20075%Nationalsample ofMedicarebeneciaries.BeneciariesenrolledinthePartDprogramwithatleast 1prescriptionclaimforstatinwereincludedinthestudy.Patientswereclassied ascontinuers,switchersanddiscontinuersofstatinmedicationsafterreachingthe coveragegap.Theprimaryoutcomesofinterestwereaveragenumberofemergencyroom(ER)visits,hospitalizations,outofpocket(OoP)costsandnumberof prescriptionspreandpostcoveragegap. RESULTS: 131237patientsmettheinclusioncriteria.Allpatientsinthestudysamplewerediagnosedwithhyperlipidemia, and8,251patientswereprescribedwithstatinsasaprimarypreventiontreatment. Ofthebeneciariestakingstatins,65.8%werefemale;80.71%white;10.91%black andthemeanagewas71.9( 12.0)years.Overall,24.3%ofstatinusersreachedthe coveragegapin2007,with10.2%reachingthegapbyMarch,42.0%byJune,and 74.4%bySeptember.Within90daysafterreachingthegap,74.2%ofstatinusers continuedusingtheirstatinmedication,1.0%switchedtoadifferentstatinmedication,and24.8%discontinueduseofstatins.Afterreachingthecoveragegap,the averagenumberofERvisits/monthincreasedfrom0.24to0.25andtheaverageOoP costs/monthincreasedfrom$58.51to$104.03. CONCLUSIONS: Theresultssuggest anoverallnegativeimpactoftheMedicarepartDcoveragegaponenrolleehealth, adherencetostatinmedications,numberofhospitalizations,emergencydepartmentvisitsandoutofpocketexpensesincurredbybeneciaries.Thedisruptionof thecoveragegaponthequalityofcareisdemonstratedbythelargepercentageof statinusersdiscontinuingtherapy. PCV62 PERSISTENCEANDCOMPLIANCEINHYPERTENSIONTREATMENTWITH OLMESARTANMEDOXOMIL–ANALYSISOFREAL-LIFEPRESCRIPTIONDATASandbergA1,KostevK2,EhlkenB3,HolzB2,OberdiekA1 1DaiichiSankyoEuropeGmbH,Munich,Germany,2IMSHealthGmbH&Co.,Frankfurt, Germany,3IMSHealthGmbH&Co.,Munich,GermanyOBJECTIVES: Toevaluatetreatmentcomplianceandpersistenceinpatientsreceivingxedcombinationscomparedtopatientsreceivingunxedcombinationswith olmesartanmedoxomilforhypertensiontreatmentinGermany. METHODS: This retrospectivestudyanalyzedprescriptiondatacollectedbygeneralpractitioners, usingalongitudinaldatabase,theGermanIMSDiseaseAnalyzer(DA).TheDA databasewassearchedforpatientswithhypertension(ICD-10codeI10)whowere initiatedondoubleortriplecombinationsofolmesartanmedoxomilwithhydrochlorothiazideand/oramlodipineintheperiod09/2008-08/2009withafollow-upof atleast12months.Persistencewasdenedasproportionofpatientswhoremainedontheirinitiallyprescribedtherapyat1year.Thedifferencebetween meanpersistencevalues(days)werecalculatedbyusingmultipleregressionanalysesadjustedbyage,gender,region,insuranceandco-morbidity.Compliancewas measuredindirectlybasedonthemedicationpossessionratiocalculatedasnumberofdayssuppliedwithintherellintervalinrelationtothenumberofdaysinthe rellinterval. RESULTS: Intotal,7,360patientswereeligibleforanalysis,58.0% weretreatedwithaxedorsemi-xedcombinationofolmesartanand42.0%with adualortripleunxedcombinationofolmesartan.After12monthsfollowingthe rstprescription,xedorsemi-xedcombinationshadahigherrateofpersistence (43.4%-50.2%)comparedtodualortripleunxedcombinations(25.0%-35.8%).Mean durationofpersistencewashigherinpatientsreceivingdualortriplexedor semi-xedcombinations(222.5-242.7days)comparedtopatientsreceivingunxed combinations(183.6-205.0days).Meancompliancevaluewitholmesartaninall combinationsrangedbetween70.0% 80.0%. CONCLUSIONS: Overall,persistence isincreasedinpatientsreceivingxedcombinationsofolmesartanincomparison tofreecombinations.Thesereal-lifedataconrmspreviousstudyresults,that persistenceofxedcombinationsincomparisontounxedcombinationsisimprovedbecauseofreducedpillburden. PCV63 THEIMPACTOFCARDIOVASCULARDISEASEHOSPITALIZATIONSONCHANGES INPATIENTADHERENCETOANTIHYPERTENISVEDRUGSINAMEDICAID POPULATIONYanX,ShayaFT UniversityofMarylandSchoolofPharmacy,Baltimore,MD,USAOBJECTIVES: Theimpactofhealthbehaviors(e.g.,medicationadherence)onoutcomes(e.g.,hospitalizations)hasbeenwidelyexploredinliterature,butfewstudieshaveevaluatedthereverseassociation.Weinvestigatedtheimpactofcardiovasculardisease(CVD)hospitalizationsonchangesinadherenceto antihypertensive(AH)drugs. METHODS: MarylandMedicaidclaimsdatafromJanuary2001toJune2006wereanalyzed.TheindexdateforpatientswithaCVD hospitalizationwastheadmissiondateoftherstobservedCVD-relatedhospitalization.Matchedcontrolshadthesameindexdateasindexpatients.Proportionof dayscovered(PDC)wasusedtomeasuremedicationadherence(adherence: PDC 80%).PatientswithandwithoutaCVDhospitalizationwerematched1:1,on therstpharmacyclaim( 6month)ofanyAHdrugsrecordedinthedatabase,age ( 5years),gender,race,numberofdrugstakeninthe6-monthpre-indexperiod ( 1drug),andadherenceratetoAHdrugsinthe6-monthpre-indexperiod( 5%A43VALUEINHEALTH14(2011)A1–A214

PAGE 52

PDC).TheoutcomewasadherencetoAHdrugsinthe6-monthpost-indexperiod. LogisticregressionanalysiswasconductedtoexploretheimpactofCVDhospitalizationsonchangesinadherencetoAHdrugs. RESULTS: Therewere1332patients withAHdrugs.PatientswithaCVDhospitalizationwere2.9times(95%Condence Interval:2.1-3.9)morelikelytobeadherentinthe6-monthpost-indexperiodcomparedtocontrolpatients.AmongpatientswithaCVDhospitalization,theproportionofpatientswhowerenon-adherenttoAHdrugsinthe6-monthpost-index periodwas30.6%. CONCLUSIONS: PatientadherencetoAHdrugsimprovedaftera CVDhospitalization,buttherewasstillasubstantialproportionofpatientswho werenon-adherentafterthathospitalization.Counselingpatientsonmedication adherenceduringtheirhospitalizationmaybeaneffectivewayforimprovingtheir adherencefollowingdischarge. PCV64 ADHERENCETOMEDICATIONSWITHONCE-A-DAY(QD)ANDTWICE-A-DAY (BID)DOSINGFORMULATIONSINACUTECORONARYSYNDROME(ACS) PATIENTSHessG1,BhandaryD2,FonsecaE3,KumarD3,PotterD2,SmithD2,GandhiSK2 1SDI,LeonardDavisInstitute,UniversityofPenn,PlymouthMeeting,PA,USA,2AstraZeneca PharmaceuticalsLP,Wilmington,DE,USA,3SDI,PlymouthMeeting,PA,USAOBJECTIVES: Toestimatepatientadherencewithonce-a-day(QD)vs.twice-a-day (BID)chronicmedicationsfollowinghospitaldischargeforACS. METHODS: Aretrospectivecohortstudyofpatientsdischargedbetween1/1/2007and4/30/2009 withanACSdiagnosiswasperformedusingalargehospitalandpharmacyclaims dataset.TwochronicmedicationsdispensedforQDandBIDutilization,carvedilol andmetformin,wereanalyzedforadherencemeasures[persistency,daysontherapy,compliance(medicationpossessionratio,MPR),total#ofdispensedprescriptions,gap(days)betweenrells]overa12monthpost-indexperiod.Includedpatientshadrstdispensedprescriptionofcarvedilolormetforminwithin60daysof discharge(indexprescription)andhadRxactivityfor any drug 12monthspostindex.Persistencewasdenedaspercentageofpatientswithoutatherapylapse of 30daysfromlastdispensedday'ssupply. RESULTS: PersistencywithcarvedilolQDvs.BID(N 168vs.2086)at6monthswas44.0%vs43.7%andat12months was24.4%vs.25.5%.PersistencywithmetforminQDvs.BID(N 136vs.614)at6 monthswas50.7%vs53.7%andat12monthswas28.7vs.35.0%.Theaveragedays ontherapyforcarvedilolQDvs.BIDat6monthswas120.5vs.121.9andat12 monthswas196.7vs.203.0.AveragedaysontherapyformetforminQDvs.BIDat6 monthswas123.6vs.136.2andat12monthswas206.1vs.237.7.Compliance(MPR) withQDvs.BIDcarvedilolat12monthswas84.2%vs80.7%andformetforminwas 77.6%vs81.6%.AdditionaladherencemetricswereconsistentforQDvs.BID dosing. CONCLUSIONS: InACSpatients,noclinicallymeaningfuldifferenceson adherencemeasureswereobservedbetweenQDversusBIDdosingformulations overa12monthfollow-upperiod.ResultsindicatepotentialopportunitiestoimprovepersistencywithchronictherapiesinACSpatients. PCV65 NEWSTATINUSERS’PERSISTENCEANDADHERENCE:BOTHARECRITICAL CONCEPTSINTHECOMPREHENSIVECHARACTERIZATIONOFMEDICATION EXPOSURESlejkoJF1,ValuckRJ1,HoPM2,AndersonHD1,NairKV1,CampbellJD1 1UniversityofColoradoSchoolofPharmacy,Aurora,CO,USA,2UniversityofColorado,Denver SchoolofMedicineandUSDepartmentofVeteransAffairs,Denver,CO,USAOBJECTIVES: JusticationfortheUseofStatinsinPrevention:anInterventionTrial EvaluatingRosuvastatin(JUPITER)demonstratedastatinbenetforprimaryprevention.However,real-worldpatientsmaynotexhibitmedicationpersistenceand adherenceseeninthetrial.Wedescribedpersistenceandadherenceofrst-time statinusers. METHODS: A10%randomsampleoftheIMSLifeLinkHealthPlan ClaimsDatabasewasusedtoobtainprescriptionclaimsrecordsforadult( 18 years)rst-timestatinuserswithcontinuoushealthplaneligibility12months priorand32monthsaftertheindexstatinprescriptionbetweenJuly1,1997and December31,2008.Persistenceandadherenceweremeasuredduringthe24 monthsafterstatininitiation.Patientswerepersistentifgapsinstatinusedidnot exceed180consecutivedays.Adherencewasmeasuredasthemedicationpossessionratio(MPR)duringtheperiodofpersistence.Persistencegroupswerecategorizedasshort'( 9months),intermediate'(9-16months)andlong'(17 months) andcomparedusingANOVA. RESULTS: Among26,530newstatinusers,themean lengthofpersistencewas17months.Theproportionsineachpersistencecategory were:21%short',12%intermediate'and68%long'.32%werepersistentfor24 monthsormore,ascomparedto75%ofJUPITERpatientstakingmedicationafter themedian1.9yearstudyperiod.MeanMPRoftheintermediate'andshort'persistencecategoriesweresimilar(0.70vs.0.69,P .15),butlowerthantheoverall meanMPRof0.80.MeanMPRwasgreatestinthelong'persistencecategory(0.85, P .0001)andwashigherthantheoverallmeanMPR. CONCLUSIONS: Persistence andadherencemeasuretwodifferentbutcriticalconcepts:thelengthoftime patientsusestatinsandtheiradherencetothestatinregimenduringthatperiod, whichwefoundtovary.Extrapolatingtheprimarypreventionbenetsofstatins mustaccountforbothmeasures,astheydifferfromclinicaltrialstopractice. PCV66 ADHERENCERATESAMONGHEALTHPLANMEMBERSSTARTINGGENERIC VERSUSBRANDSTATINTHERAPYChenQ1,KachurSG2,FatoduH2,GarberH2 1JohnsHopkinsBloomsbergSchoolofPublicHealth,Baltimore,MD,USA,2JohnsHopkins HealthCare,GlenBurnie,MD,USAOBJECTIVES: Severalstudieshavedemonstratedthathigherpatientout-ofpocket costmayresultinlowermedicationadherence.Thepurposeofthisstudywasto measureMedicationPossessionRatio(MPR)amongpatientsnewly-prescribeda brandorgenericstatinmedicationinamanagedMedicaidplanandacommercial healthplan. METHODS: Weconductedaretrospectiveanalysisusingpharmacy claimsdatatoidentifypatientswhowerenewtostatintreatment(nopharmacy claimforalipid-loweringmedicationintheprevious12months).Patientswere categorizedbasedontheirindexmedication.WeusedaRobustlinearregression modeltodeterminepredictorsofadherence. RESULTS: Atotalof738commercial patientsand2175Medicaidpatientswereincluded.SixtypercentofMedicaidpatientsand49%ofcommercialpatientsinitiatedtherapywithagenericmedication. Averagepatientout-of-pocketcostforcommercialplanpatientswas$9/monthfor genericand$15/monthforbrand.Medicaidpatientshadnocopaymentforgeneric orbrandmedications.Inthecommercialplan,therewasnosignicantdifference inMPRbetweenpatientswhoinitiatedtherapywithgenericorbrandstatins(MPR 0.75vs.0.73,respectively).IntheMedicaidplan,MPRwassignicantlyhigher amongpatientswhostartedongenericmedications(0.69vs.0.63).Inrobustlinear regression,MPRwassignicantlyrelatedtoage,numberofcomorbiditiesandgenericuse.AfterweadjustedMPRforageandcomorbidities,MPRremainedsignificantlyhigherintheMedicaidgenericgroup. CONCLUSIONS: Medicaidpatients prescribedagenericstatinasinitialtherapyweremoreadherentthanthoseprescribedabrand,despitehavingnocopaymentforgenericorbrandmedications. Thisdifferencewasnotpresentamongcommercialplanpatientswhohadahigher costshareforbrandmedications.Thissuggeststhatadditionalresearchisneeded toidentifynon-nancialbarrierstoadherence. PCV67 EXAMININGMEDICATIONADHERENCEAMONGTRICAREBENEFICIARIES RECEIVINGSTATINTHERAPYFORSECONDARYPREVENTIONOFCORONARY HEARTDISEASEINMILITARYTREATMENTFACILITIESINTHEUNITEDSTATESNwokejiE1,YargerS1,TriceS2,ChaoS3,DevineJ2,PotykR2,GutkeG3,BonnemaA3 1DepartmentofDefense/GDIT,FortSamHouston,TX,USA,2DepartmentofDefense,FortSam Houston,TX,USA,3AirForceMedicalSupportAgency,SanAntonio,TX,USAOBJECTIVES: Toexaminestatinadherenceandpersistenceamongpatientsreceivingtreatmentforsecondarypreventionofcoronaryheartdisease(CHD)atUS MilitaryTreatmentFacilities(MTF)withintheDepartmentofDefense(DoD). METHODS: RetrospectivecohortstudyutilizingtheDoDMilitaryHealthSystem databasetoexamine21,053TRICAREbeneciariesbetween18-75yearsofage, receivingmedicalservicesforaprimaryCHDeventatanMTFbetweenJanuary1, 2004andDecember31,2008.DrugadherencewasmeasuredusingtheMedication PossessionRatio(MPR)at6,12and18months.Persistencewasmeasuredasdurationofstatin-therapybasedon 35-dayrellgap.Covariatesincludedage,gender, comorbidities,drug-switchinganddosagetitration.Logisticregressionwasconductedtoassesspredictorsofadherence[95%Condence-Intervals]. RESULTS: TheCHDcohort(N 21,053)was74%malewithameanageof57.4(SD 8.8)years. OverallmeanMPRwas89%(SD 22)atmonthsix(M6),84%(SD 25)at12-months (M12),and81%(SD 26.4)after18-months(M18).Approximately80%ofpatients wereadherent(MPR 80%)withstatin-therapyatM6whichdeclinedto71%atM12, and69%attheendofM18(p .001).Oldermalepatientswithhyperlipidemiawere moreadherent.Adjustingforcovariates,patientsweremorelikelytobeadherent atM6,M12,andM18thatswitchedstatins(OR 1.87[1.65-2.13],1.3[1.24-1.46],and 1.07[1.00-1.15])orhadatleastonetitrationadjustment(OR 2.99[2.62-3.40],1.86 [1.72-2.01],and1.54[1.44-1.65])comparedtopatientswithnotherapeuticadjustments.Overallmeanpersistencetostatinswas322days.Patientsexperiencing 35-dayrellgapincreasedfrom34%to53%to63%at6,12,and18-months respectively. CONCLUSIONS: Thisstudyshowedstatinadherencewashighamong DoDpatientsreceivingmedicalcareatMTFsforsecondarypreventionofCHD duringtherst6-months.Adherenceandpersistence,however,declinedbymonth 12.Improvedpatientadherencewasassociatedwithclosermonitoringofprescribedtherapy,asseenamongpatientswhoweretitratedorswitchedstatins. PCV68 FACTORSASSOCIATEDWITHNON-ADMINISTRATIONOFORDERED PHARMACOLOGICVENOUSTHROMBOEMBOLISMPROPHYLAXISDOSESShermockKM1,SharV2,StreiffMB3,LavalleeDM4,PintoBL1,HautE3,KrausP5,ErdL5, ElderS5 ,RandC3 1TheJohnsHopkinsMedicalInstitutions,Baltimore,MD,USA,2PhiladelphiaCollegeofPharmacy, Philadelphia,PA,USA,3TheJohnsHopkinsUniversitySchoolofMedicine,Baltimore,MD,USA,4CenterforMedicalTechnologyandPolicy,Baltimore,MD,USA,5TheJohnsHopkinsHospital, Baltimore,MD,USAOBJECTIVES: Arecentstudybyourteamindicatedthatapproximately13%of prescribedvenousthromboembolism(VTE)prophylacticdosesarenotadministered.Thegoalofthisstudywastodeterminedocumentedreasonsanddescribe thedistributionpatternsfornon-administeredVTEprophylaxisdoses. METHODS: Weconductedaretrospectivereviewofelectronicmedicationadministrationrecordsusingourcomputerizedphysicianorderentrysystem.Thestudyincluded hospitalizedpatientsaged18yearsorolderwhowereorderedpharmacologicVTE prophylaxisfromDecember1,2007throughJune30,2008. RESULTS: Atotalof 108,533VTEprophylaxisdoseswereorderedfor8,607patients.12.8%ofordered doseswerenotadministered.Non-administrationratesvariedbypatientoor from4.8%to33.9%.Approximately53%ofdosesthatwerenotadministered(6.8% ofalldoses)weredocumentedaspatientrefusalthemostcommondocumented reasonfornon-administrationofVTEprophylaxisbyawidemargin.Patientrefusalsvariedgreatlybynursingunit,rangingfromlessthan1%to19%ofordered doses.Patientrefusalswerehighlyconcentratedoncertainnursingunits;5nursingunitsaccountedfortwo-thirdsofallrefuseddoses.Asmallnumberofpatients accountedforthevastmajorityofrefuseddoses;11%ofpatientswhorefusedmore than1doseaccountedfornearly87%ofallrefuseddoses. CONCLUSIONS: PatientA44VALUEINHEALTH14(2011)A1–A214

PAGE 53

refusalofVTEprophylaxisdosesisthemostcommondocumentedreasonthese dosesarenotadministered.Asmallnumberofpatientsandnursingunitsaccount foravastmajorityofrefuseddoses.Wehaveinitiatedamixed-methodsinvestigationtodeterminewhatpatient,nursingunit,andcultureofcare-levelfactors mayexplainrefusalofthesedoses. PCV69 IMPACTOFCOMMUNITYPHARMACIST-LEDCOUNSELINGONIMPROVING MEDICATIONADHRERENCEJiangJZ,KhandelwalN,RudkinK,DuncanI Walgreens,Deereld,IL,USAOBJECTIVES: Toassesstheimpactofpharmacist-ledface-to-facecounselingon prescriptionrstrellrateinpatientswhoarenewtostatinsorthyroid medications. METHODS: Aretrospectivepre-postwithcase-controlcohortstudy designwasemployed.InMay2010,twocommunitypharmaciesimplemented face-to-facecounselingbytrainedpharmacistsatthesiteofcarewhenpatients initiallylledtheirstatinorthyroidprescriptions.Patients'rstrellrate,dened aspercentageofpatientswhorelledtheirprescriptionswithinatwo-week'speriodafterexhaustionoftheirinitialll,wasassessedastheprimaryoutcome measure.Similarcomputationswereperformedforpatients'wholledtheirstudy medicationsatthetwostores,12-monthspriortotheinterventionprogram.Additionally,twocontrolpharmacystoreswereselectedbasedonpharmacytype,operationalyearsofstores,prescriptionvolume,andcorrespondingpopulationcharacteristics.Rellratesinthepreandpost-periodswerealsocomputedforcontrol stores.Achi-squaretestwasperformedtoinvestigatetherelationshipbetween face-to-faceinterventionandpatients'rstrellrate. RESULTS: Teststoresthat implementedface-to-facecounseling,conductedinterventionson81newtotherapypatientsinthepostperiod,whileatotalof76(newtotherapypatients)were includedinthepre-period.Atotalof73and81newtotherapypatientswere selectedfromcontrolstoresinthepreandpost-periodrespectively.Forteststores, chi-squareanalysesshowedasignicantincreaseintherellratesinthepostperiodascomparedtothepre-period(Pre-period:55.7%VsPost-period:70.4% p 0.01),whereascontrolstoresdidnotshowanysignicantincreaseacrosstwo periods(Pre-period:56.2%VsPost-period:61.7%,p 0.05). CONCLUSIONS: Pharmacist-ledface-to-facecounselinghadapositiveimpactonpatient'srellbehavior. Thestudyshowedanearlyimpactoftheprogram;additionallargerpilotstudies andlongerfollow-upperiodareneededtobetterunderstandthefullbenetofthe program. PCV70 VARIATIONSINTHEIMPACTOFILLNESSPERCEPTIONSANDMEDICATION BELIEFSONMEDICATIONCOMPLIANCEOFELDERLYVERSUSGERIATRIC HYPERTENSIVECOHORTS:ACOMPARATIVEANALYSISRajpuraJR1,NayakR2 1PurdueUniversity,WestLafayette,IN,USA,2St.John'sUniversity,Jamaica,NY,USAOBJECTIVES: Tomeasureandcomparetheextentofmedicationcompliancein elderlyandgeriatrichypertensivecohorts.Toevaluatetheroleofillnessperceptionsandinvestigatehowmedicationbeliefsanddimensionsofillnessperceptions,mightimpactmedicationcomplianceoutcomeinhypertension. METHODS: Across-sectionalresearchdesign,utilizingconveniencesamplingstrategiesand self-administeredsurveysofelderlyhypertensiveresidentslivinginNewYorkCity seniorcarecenters(N 117).MedicationcompliancewasmeasuredusingMorisky's test;medicationbeliefsweremeasuredusingBeliefsaboutMedicationQuestionnaire(BMQ),andillnessperceptionsusingBriefIllnessPerceptionsQuestionnaire (BIPQ). RESULTS: MajoriMajorityofthepopulationwaswhite(75%).Therewereno signicantdifferencesobserveddemographicallybetweentheelderly(55-65years) andgeriatric(65 years)samplesexceptwithrespecttocollegeeducation(53%vs. 47%).About46%oftheelderlysamplewasnoncompliantwithhypertensionmedications,comparedto63%whowere65yearsandolder(geriatric).BIPQcomponent scorescorrelatedpositivelywithmedicationcompliancescoreacrossbothelderly andgeriatricsamples(r 0.134;p 0.496andr 0.447;p 0.000,respectively).The scoreonSpecicNecessityBeliefsdimensionassociatedpositivelywithMorisky's Testscoreacrossboththecohorts(r 0.027,p 0.892andr 0.297;p 0.020), whereasGeneralHarmandGeneralOveruseBeliefsaboutmedicationcorrelated negativelywithmedicationcompliance.Amajorityoftheparticipantsrank-orderedstressasthenumberonecauseoftheirillness(49.1%),followedbylifestyle' (43.8%)andhereditaryfactors'(7.1%). CONCLUSIONS: Medicationnoncompliance seemstoworsenwithanincreaseinage.Strongernecessitybeliefsandmore favorableperceptionsabouttheillnessareobservedingeriatricscomparedtoelderly.Thestudyunderscorestheimportanceofincorporatingpatientperceptions aboutillnessandmedicationbeliefsintodiseasemanagementframeworksthat seektobuildmedicationcomplianceinhypertension. PCV71 EVALUATINGMEDICATIONADHERENCEWITHCHRONICCARDIOVASCULAR THERAPIESAMONGUNINSUREDWORKINGADULTS:RESULTSFROMTHE TEXASDEMONSTRATIONTOMAINTAININDEPENDENCEANDEMPLOYMENT (DMIE)STUDYNwokejiE1,BohmanT1,ReedB2,WallischL1,OstermeyerB2,StonerD3,SpenceR1 1UniversityofTexasatAustin,Austin,TX,USA,2BaylorCollegeofMedicine,Houston,TX,USA,3TexasDepartmentofStateHealthServices,Austin,TX,USAOBJECTIVES: Toexaminepatternsofmedicationadherenceandpersistencewith antihypertensiveorstatintherapiesamonguninsuredworking-adultsenrolledin theTexasDMIEStudy. METHODS: Thisstudywasarandomizedcontrolledtrial (RCT)conductedtoassesstheeffectsofanintervention(i.e.,improvedmedical servicesandsubsidizeddrug-copay)amonguninsuredworking-adultparticipants age21-65years,receivingmedicalservicesintheHarrisCountyHealthDistrict. Cardiovascularmedicationadherenceandpersistenceweremeasuredseparately forparticipantsprescribedeitherstatinorantihypertensive(i.e.,angiotensin-converting-enzyme-inhibitororangiotensin-receptor-blocker)therapyduringan18monthperiod.Medicationadherencewascalculatedastheproportionofdays covered(PDC).Participantswere"adherent"whenPDC 80%.Persistencewas measuredasa 35-daygapbetweendrugrells.Logisticregressionandcoxproportionalhazardsmodelsevaluatedpotentialpredictorsofadherenceand persistence. RESULTS: Forparticipantsonstatintherapy(N 449),ahighermean PDC(p 0.01)wasobservedintheinterventioncohort(0.61[SD 0.27])versus control(0.51[SD .30]).Switchingstatinsandtitrationadjustmentsweresignificantpredictorsofadherence.Adjustingforcovariates,theintervention (OR 1.28[CI95%,1.01,2.68])wasmorelikelytobeadherenttostatinsversuscontrol. Meanpersistencewithstatinsintheinterventionwas230-dayscomparedto171daysforcontrolgroupparticipants.Theinterventionwaslesslikelytoexperience arellgap(OR 0.69[CI95%,0.51,0.92])versuscontrol.Forantihypertensivedrugs (N 636),meanPDCintheinterventionwas0.65(SD 0.27)versus0.60((SD 0.28) forcontrol(p 0.02).Predictorsofadherenceincludedrace/ethnicity,polypharmacy,anddepression.Theinterventiongroup(OR 1.32[CI95%,1.07,2.88])was morelikelytobeadherentversuscontrol.Meanpersistencewithantihypertensivesintheinterventionwas291-daysversus257-daysforcontrol.Nosignicant differenceinpersistenceseenbetweeninterventionandcontrol. CONCLUSIONS: ThisRCTshowedimprovedadherencetostatinandantihypertensivemedications amongtheinterventioncohort.Findingsfromthisstudywouldbeusefulinguiding ongoingeffortstooptimizecardiovascular-drugadherencetargetedtowarduninsuredworking-adults. PCV72 THERISKOFSTROKEANDPREVENTATIVESTEPSAMONGPATIENTS DIAGNOSEDWITHATRIALFIBRILLATIONINTHEUNITEDSTATESGuptaS1,WagnerJS2,GorenA2,DiBonaventuraM2 1KantarHealth,Princeton,NJ,USA,2KantarHealth,NewYork,NY,USAOBJECTIVES: ThisstudyexaminedstrokeriskandpreventionamongUSpatients diagnosedwithatrialbrillation(AF). METHODS: DatawereextractedfromtheUS 2010NationalHealthandWellnessSurvey,anannual,Internet-based,cross-sectionaldatabaseof75,000adults(age 18).StrokeriskwasassessedwithCHADS2, determinedbypresenceofcongestiveheartfailure,hypertension,diabetesmellitus,oldage(75years ),andhistoryofstrokeortransientischemicattack.Low(CHADS2 0),moderate-(1),andhigh-(2 )riskpatientsreportedwhichsteps,if any,theytooktopreventstroke:lowfatorlowsodiumdiet,regularexercise, smokingcessation,weightloss,bloodpressureorcholesterolreduction,oruseof babyaspirin,aggrenox,warfarin,clopidogrel,orothermedications.High-riskpatientswerecomparedwithlowandmoderate-riskpatients,adjustingforsociodemographicandhealthcharacteristicsusinglogisticregressionmodels. RESULTS: Among1,350respondentsdiagnosedwithAF(projectedUSprevalenceof 1.49%),308(22.8%)werelow,473(35%)moderate,and569(42.2%)highinriskof stroke.High-vs.low-riskpatientsweresignicantlymorelikelytotakeanypreventativesteps(60.5%vs.34.1%,respectively;OR 2.75, p 0.001 ).Furthermore, high-riskpatientsweresignicantlymorelikelytotakestepstolowertheirblood pressure(OR 8.33, p 0.001 ),cholesterol(OR 2.46, p 0.007 ),useclopidogrel (OR 7.46, p 0.027 )orotherstroke-prophylacticmedications(OR 3.32, p 0.003 ) comparedwithlow-riskpatients.Nodifferenceswereobservedonlifestylemodicationsuchasdiet,exerciseorsmokingcessation. CONCLUSIONS: Theseresults suggest,USadultsathighriskofstrokearemorelikelytotakestepstolowerhigh bloodpressureandcholesterol,andtousemedicationstopreventstroke,comparedwithlow-riskindividuals.Nosignicantdifferencesemergedondiet,regular exercise,smokingcessationandaspirinuse.Higherriskcorrelatedwithhigher prevention,yetthereremainsanunmetneedforincreasedtargetedtreatmentand lifestylechangesforhigh-riskAFpatients. PCV73 CONSTRUCTVALIDITYOFHEALTHUTILITIESINDEX(HUI)JAPANESEVERSION: CROSS-SECTIONALSTUDYFORSTROKEINJAPANNotoS1,UemuraT2,IzumiR1,MoriwakiK1 1NiigataUniversityofHealthandWelfare,Niigata,Japan,2KyorinUniversity,Tokyo,JapanOBJECTIVES: ToassessthemeasurementpropertiesandvalidityofHealthUtilities IndexMark3(HUI3)JapaneseversionforstrokepatientsinJapan. METHODS: The HUI3wasadministered553patientswithstrokewhowereadmittedinsevenhospitals.ProxiesofthepatientscompletedJapaneseversionoftheHUI3questionnairesatstartphaseofrehabilitation.Patientswerecategorizedusingmodied Rankinscale(MRS).Theconstructvaliditywasassessedbyanalyzingthedegreeto whichlowerscoreontheHUI3scorescorrelatedpositivelywithMRS. RESULTS: Meanglobalscorewasasfollows:MRS1 0.62(SD 0.22),MRS2 0.48(SD 0.28), MRS3 0.27(SD 0.24),MRS4 0.00(SD 0.18),MRS5 0.23(SD 0.11).Worst meansinglescorewasambulation(0.29)and,subsequentlycognition(0.49),and dexterity(0.58). CONCLUSIONS: OurresultsindicatetheproductivefutureforJapaneseHUI3useandshowedprecisemeasuringpropertiesoftheHUI3Japanese versioninassessinghealthstatusforstrokepatients. PCV74 ASSESSMENTOFKNOWLEDGEABOUTHYPERTENSION(HTN)AMONG HYPERTENSIVEPATIENTSINSOUTHERNPUNJAB,PAKISTANAtifM1,AsifM2,SaleemF1,AhmadN1,SaqibN3 1UniversitiSainsMalaysia(USM),Pinang,Penang,Malaysia,2TheIslamiaUniversityof Bahawalpur,Bahawalpur,Punjab,Pakistan,3MinistryofHealth,Karachi,Sindh,PakistanA45VALUEINHEALTH14(2011)A1–A214

PAGE 54

OBJECTIVES: Toassessknowledgetowardsetiology,treatmentandmanagement ofHTNamonghypertensivepatientsinSouthernPunjab,Pakistan. METHODS: A convenientsampleof400HTNpatientsvisitingpublichospitalinBahawalpurwas selectedforthestudy.Apre-validated,selfadministeredquestionnairewasused fordatacollection.AnalysesweredonebySPSS16. RESULTS: Atotalof89.9%ofthe patientsweremaleswithameanageof35.69yearswith240(60%)havingbachelor levelofeducation.Atotalof180(45%)oftheparticipantshadincomebelowPakistanrupees20,000permonth.296(74%)patientsknewthevaluesofnormalBlood Pressure.160(40%)knewthecausesandsymptomsofHTN.188(47%)ofpatients hadhypertensivepatientsintheirfamilies;308(77.75%)thoughtthatthereshould belifestylemodicationsincaseofHTNandalsostatedthatHTNiscurable. Majorityofpatients216(54%)didnotmonitortheirbloodpressureregularlyand weredependingonsignsandsymptoms.Atotalof139patients(34.78%)became hypertensiveaftertheageof30yearswhile130(32.61%)gotthediseaseafterthe ageof20.Allpatientswerereceivingallopathictreatmentbutonly130(30%)were followingspecialdietprogram.Asmallnumber100(25%)ofpatientswerereceivingcounselingservicesfromhealthcareprofessionalswhereas316(79%)were dependinguponopinionsfromfriends,peersandpatientsalreadysufferingfrom HTN. CONCLUSIONS: ThestudyresultsidentiedastrongneedforeducationprogramregardingHTNatthecommunitylevel.Theinteractionbetweenpatientsand healthcareprovidersshouldbeincreasedinordertohavemaximumtherapeutic outcome.Furtherstudiesarerecommendedtoidentifythecontributingfactorsin thetreatmentandmanagementofHTN. PCV75 EFFECTOFACEIANDARBONHEALTHRELATEDQUALITYOFLIFEIN PATIENTSWITHHYPERTENSIONANDDIABETESJohnsonML,ParikhR,RajanSS,AparasuR UniversityofHouston,Houston,TX,USAOBJECTIVES: Toassesstheeffectofangiotensinconvertingenzymeinhibitors (ACEI)andangiotensinreceptorblockers(ARB)onhealthrelatedqualityoflife (HRQoL)inpatientswithhypertensionanddiabetes. METHODS: MedicalExpenditurePanelSurvey(MEPS)2002-2007wasusedforthestudy.EachpanelinMEPS followedpatientsforabout2yearswithdatacollectedinveroundsof4to5 monthseach.HRQoLwasmeasuredusingthephysicalandmentalcomponent scoresoftheSF-12qualityoflifeinstrument.BaselineHRQoLwasmeasuredatthe endofround2.IncidentutilizationforACEIandARBwasdeterminedduringround 3or4,andoutcomeHRQoLwasmeasuredattheendofround4.Analysiswas performedontwopopulations:patientsutilizingeitherACEIorARBvs.neither; andpatientsutilizingARBvs.ACEI.MultiplelinearregressionmodelswereperformedtotesttheuniqueeffectofACEIandARBonthephysicalandmental componentofHRQoLwhilecontrollingforbaselineHRQoL,comorbidities,comedications,demographicandsocio-economicfactors. RESULTS: Overall60%ofpatientsreceivedeitherACEIorARB.Ofthe20%ofthesewhowereincidentusersof ACEI/ARB,ACEIwasutilizedby65%ofthepatients.Multiplelinearregression modelsdidnotrevealanystatisticallysignicanteffectofusingACEIorARBvs neitheronthephysical(p 0.27)ormental(p 0.47)componentofHRQoL.Similarly,usingACEIversusARBdidnothaveanystatisticallysignicanteffectonthe physical(p 0.20)andmental(p 0.66)componentsofHRQoL. CONCLUSIONS: ACEIandARB,nowregardedasqualityofcaremeasuresforpatientswithconcomitanthypertensionanddiabetes,didnotseemtoaffectpatients'HRQoL.More researchisneededtoevaluatethelongtermconsequencesofACEIandARBuse. PCV76 PATIENTSYMPTOMEXPERIENCESFOLLOWINGAMYOCARDIALINFARCTIONGwaltneyC1,MartinML2,FalveyH3 1BrownUniversity,Providence,RI,USA,2HealthResearchAssociates,Inc.,Seattle,WA,USA,3NovartisPharmaAG,Basel,SwitzerlandOBJECTIVES: Cardiovasculardiseaseculminatinginmyocardialinfarction(MI)remainsprevalentinsocietyandisresponsibleforsignicantmorbidityandmortality.PatientsmayexperienceavarietyofdeleteriouseffectsfollowinganMI.An enhancedunderstandingofpatients'post-MIexperiencesmayleadtoimproved assessmentsofpost-MItreatmentoutcomesand,ultimately,improvedtherapies. METHODS: Participantswere38patientswhohadexperiencedanMI(withor withoutSTsegmentelevation)withinthepast6months(n 17)orinthepast6-12 months(n 21).One-hourin-personinterviewswereconductedusingasemistructuredinterviewguide.InterviewresponseswerecodedusingATLAS.tisoftware.Thepercentofpatientsmentioningaspecicpost-MIexperiencewasusedto identifythemostrelevantandgeneralizablepatientexperiences. RESULTS: The averageageofthepatientswas53(range 34-87,63%male).Feelingtired,themost commonlyreportedpost-MIexperience,wasreportedby76%ofpatients.Within thistirednesscategory,patientsreportedexperiencingalossofphysicalenergy (66%),mentalenergy(34%),andmotivation(26%).Amajorityofpatientsalsoreporteddyspnea(58%).Additionally,patientsreportedemotionaleffects,including worry(61%)anddepression(58%),aswellasanimpactonrelationships(58%)and socialactivities(50%).TimesincetheMIhadlittleeffectonpatientexperiences. CONCLUSIONS: PatientsreportawidevarietyofnegativeexperiencesafteranMI. Alossofphysicalenergyisonlyonecomponentofthetirednessexperiencedby patients;lossofmentalenergyandlossofmotivationarealsocommon.These experiencesmaybedirectlyrelatedtotheeffectsoftheMI,butmayalsobesecondarytotheexperienceofdepressionandworry.Furtherrenementoftheconceptsassociatedwiththepatient'sexperiencefollowinganMImayyieldnovel endpointsforuseinclinicaltrialsandbettertherapies. PCV77 HEALTH-RELATEDQUALITYOFLIFEINPATIENTSWITHHYPERTENSIONIN KOREAJeongMH1,KimDS2,ChangHJ3,HongGR4,ParkSW5,KoSK6,ParkHJ6 1ChonnamNationalUniversityHospital,Gwangju,SouthKorea,2InjeUniversityPusanPaik Hospital,Busan,SouthKorea,3SevranceHospital,YonseiUniversity,Seoul,SouthKorea,4YeungnamUniversityMedicalCenter,Daegu,SouthKorea,5SamsungMedicalCenter,Seoul, SouthKorea,6PzerPharmaceuticalsKoreaLtd.,Seoul,SouthKoreaOBJECTIVES: Inchronicconditionssuchashypertension,health-relatedqualityof life(HRQoL)isanespeciallyimportantoutcome,giventheirlifelongnatureandthe needfordailyself-management.WeaimedtoassessHRQoLamongpeoplewith hypertensioninKorea. METHODS: Aprospective,observationalstudywascarried outonhypertensivepatientsfrom21medicalcentersthroughoutKorea.Atotalof 273hypertensivepatients45-75year-oldmalesor55-75year-oldfemaleswere assessedforHealth-RelatedQoL(HRQoL).HRQoLwasassessedon8dimensionsof theSF-36questionnaire. RESULTS: Patientswithhypertensionwere60 7.6years old(44.3%female),hadhypertensionfor5 5.2years.Intotal,39.2%hadhypertensionandhyperlipidemia;12.8%hadhypertensionanddiabetes.HRQoLphysical componentmeanscores(PCS)weredeterminedas49.7 7.2,andmentalcomponentmeanscores(MCS)weredeterminedas50.3 9.3.Womenhadconsistently lowerscoresonalldimensionsofSF-36,indicatingpoorerHRQoL.Therewereno signicantdifferencesbyage,durationofhypertensioninanyoftheSFcomponents.HRQoLMCSofpatientswithcombinedhypertensionanddiabeteswere lowerthanthatoftheothergroup(p 0.021).Inthehypertensivepatientswith diabetesrevealedpoorermentalfunctioningcomparedwiththosewithhyperlipidemia(HRQoLMCS:47.0 9.8,52.0 8.8,respectively).Thescalescores(0-100scoring)inalleightdimensionsofSF-36were81.2 21.7inphysicalfunctioning, 86.2 23.1inrole-physical,79.6 23.7inbodilypain,58.2 18.8ingeneralhealth, 62.2 21.0invitality,89.4 18.3insocialfunction,87.8 22.0inrole-emotional, 74.0 20.1inmentalhealth. CONCLUSIONS: Hypertensionseemstocomparably impairHRQoLinKorea.Thecoexistenceofhypertensionanddiabetes,inparticular,hasamorenegativeeffectonHRQoL.ThendingssuggestthatQoLinhypertensivepatientshassimilartrendcomparedtotheresultsfromnationalhealth surveyaimingatresidentsinKorea(Utilityweightforhypertension:0.77from2005 KNHANES). PCV78 HEALTHRELATEDQUALITYOFLIFEINHYPERTENSIVEPATIENTSVISITING PUBLICHOSPITALSOFQUETTA,PAKISTANSaleemF1,HassaliMA1,ShaeAA1,BashirS2 1UniversitiSainsMalaysia(USM),Pinang,Malaysia,2UniversityofSargodha,Sargodha,Punjab, PakistanOBJECTIVES: ToassesstheHealthrelatedqualityoflife(HRQoL)inhypertensive patientsattendingpublichospitalsinQuetta,Pakistan. METHODS: EuropeanQualityofLifescale(EQ-5D)wasusedforassessmentofHRQoL.Descriptivestatistics wereusedtodescribedemographicanddiseasecharacteristicsofthepatients.Chi squaretestwasusedtodetermineassociationamongvariableswith(p 0.05)consideredassignicant.HRQoLwasscoredusingvaluesderivedfromtheUnited Kingdomgeneralpopulationsurvey.AllanalyseswereperformedusingSPSSversion16.0. RESULTS: 385hypertensivepatientswereenrolledforthestudywith meanageof39.02 6.596years.Majority186(48.3%)werecategorizedinagegroup of28-37years.265(68.85%)weremaleswith3.01 0.939yearsofhistoryofhypertension.ThemeanEQ-5Ddescriptivescorewas0.4674 0.28444andEQ-Vasscore was63.97 6.621indicatinglowerhealthrelatedqualityoflife.Education,occupationanddurationofdiseasehadsignicantrelationwithHRQoL.Bonferroniadjustmentwasdonetondintergroupsignicance.Amongeducation,group1had signicantrelationwithgroup3,4,and5(wheregroup1 illiterate,3,4and5being Primary,FA/FSc,andBA/BScrespectively).Inoccupation,group3wasfoundsignicantwithgroup0and2(wheregroup3 PrivateJob,group0 joblessand group2 Governmentofcial).Fordurationofdisease,group4wasfoundsignificantwithgroup1and2(wheregroup4 morethan5years,group1 lessthan1 yearandgroup2 morethan1yearbutlessthan3years). CONCLUSIONS: Lackof healthcareprofessionalsandlittleinteractionwithpatientsisoneofthemajor healthproblem.Patientstherefore,donothaveadequateknowledgeabouttheir conditionswhichinturnaffecttheirHRQoL.Toovercomethisissue,paramedical staffcanbeutilizedforcounselingthepatients. CardiovascularDisorders–HealthCareUse&PolicyStudies PCV79 IMPACTOFDISEASEMANAGEMENTPROGRAM“DETODOCORAZON”OF MUTUALSERHMOTOCONTROLCARDIOVASCULARRISK,COHORT2004-2009, COLOMBIAPazJJ1,GranadosCE2,MoralesC1,RodriguezN2,BuitragoLA2,GueteA3,CaceresHA4, RamirezP4 ,PardoR2 1MutualserEPSS,Cartagena,Colombia,2UniversidadNacionaldeColombia,Bogot‡,Colombia,3PzerS.A.,Cartagena,Colombia,4PzerS.A.,Bogot‡,ColombiaOBJECTIVES: EvaluatetheimpactoncardiovascularriskcontrolanditsdeterminantsinpatientsenrolledintheDiseaseManagement(DM)Program"DeTodo Coraz—n". METHODS: TheobjectivetheDMprogramistomodifycardiovascular risk(CVR)classiedaccordingtoAdultTreatmentPanelIII(ATPIII)methodology, theprogramincludesdirectpharmacologicalintervention,integratedwithpromotionandpreventionactivities.Thisprogrambeganin2004and27,596patients correspondingtoadynamiccohorthavereceiveditsbenetsfromituptoDecember2009.Theeffectivenessoftheprogramwasassessedthroughtheanalysisof trendsintheCVRmarkervariablesandgoalsaccomplishments.Thestudypopu-A46VALUEINHEALTH14(2011)A1–A214

PAGE 55

lationcorrespondstoparticipatingsubjectssince2004whoreceivedcontinuous annualfollow-upvisitsuntil2009.Thetrendanalysisofthedifferentriskfactors overtimewasperformed,aswellastheestimateoftheannualCVRusingATPIII categories(High,moderateandlowerrisk). RESULTS: 3,870subjectswereincluded intheassessmentwithameanageof62years(SD 11.52)andthemajorityof subjectsarewomen(73%).Theresultsshowedareductionof30%insmoking.In addition,overthepast5years,bloodpressure(BP)diminishedconsistently throughriskgroups(meanreductionof14mmHg);LDLcholesterollevelswere reducedwithinhighandmoderateriskgroups(meanreductionof25mg/dl)and thespeedofkidneydamageprogressiondecreased.Complementaryanalyseswith anextendedcohortcontainingover8,000patientsexhibitedsimilarresults. CONCLUSIONS: CardiovascularRiskDMProgram"DeTodoCoraz—n"showeda meaningfulimpacttomodifyriskfactorsinColombianpatientssuchassmoking, BP,lipidproleandkidneyrisk.Theresultsofawidercohortsupportedthatobservedpositiveoutcomestrendsweregeneratedbythisprogram. PCV80 MEDICATIONTHERAPYMANAGEMENTIMPROVESHEALTHCAREUTILIZATION ANDCOSTSFOREMPLOYERSPintoSL,ParthaG,JaniaA UniversityofToledo,Toledo,OH,USAOBJECTIVES: Todeterminethehealthcareutilizationandpotentialcostsavingsfor anemployer-sponsoredmedicationtherapymanagement(MTM)program. METHODS: Aprospective,pre-postlongitudinalstudy.Employeesincludedinthe studywereLucasCountyemployeesandtheirdependentswhohaddiabetes,hypertension,hyperlipidemia,oracombinationofthethree.TheMTMserviceswere providedbyindependentpharmacistsfromsevensitesinNorthwestOhio.Variablesmeasuredincludedsocialmeasures(alcoholconsumption,tobaccoconsumption,exercise,caffeineconsumption)andprocessmeasures(specialtyphysicianvisits,sickdays,emergencyroomvisits,ushot,hypoglycemicevents,and self-monitoringofbloodglucose).Differenceindisease-relatedspending,pre-and postjoiningMTMprogramwereanalyzedforemployeeswhoimproved,didnot improveorshowednegativeimprovementattheirnalrecordedvisitforprocess measures.DatawasanalyzedusingSPSSv17.0andMicrosoftExcel. RESULTS: Therewasanoveralltrendofimprovementinboththesocialandprocessmeasuresbeingmeasured.Allthreespecialtyphysicianvisitshaveincreased:podiatristvisitsby24%,ophthalmologistvisitsby41%,anddentistvisitsby26%.Average costsavingsforemployeeswhoimprovedormaintainedappropriateutilization, rangedfrom$931.67to$1437.77peremployeepertwoyears.Approximately90%of employeeseithertooklessorhadthesameamountofsickdays.Employeeswho hadfewersickdayssavedalmost$1230.79peremployeewhileemployeeswho tookmoresickdaysspentapproximately$2146.51peremployee.Employeeutilizationofushotshavebeensubstantial,with70%gettingaushotatleastonce overthe24months.Alcoholandtobaccoconsumptiondecreasedby50%and55%, respectively.Caffeineusedecreasedby26.47%.Reportedexerciseincreasedby 39%. CONCLUSIONS: Interactionwiththepharmacisthashadapositiveimpacton employees'outcomesrelatedtosocialandprocessmeasures,whichmightalsobe helpfulincausingreductionincostburdentotheemployer. PCV81 SUSTAININGCLINICALOUTCOMESOVER24MONTHS:ISMEDICATION THERAPYMANAGEMENT(MTM)THEANSWER?PintoSL1,JaniaAW1,CoehrsB2,RiepenhoffC2 1UniversityofToledo,Toledo,OH,USA,2ThePharmacyCounter,LLC,Toledo,OH,USAOBJECTIVES: Toassesstheimpactofanemployer-sponsored,pharmacist-providedMTMprogramonclinicaloutcomesinpatientswithdiabetesandhypertensionovera24monthperiod. METHODS: Aprospective,intent-to-treat,pre-post longitudinalstudy.PatientswereLucasCountyemployeesanddependentswith diabetes,hypertension,oracombination.TheMTMserviceswereprovidedby independentpharmacistsinNorthwestOhioatsevensites.ADAandJNC-VII guidelineswereusedtodesigninterventionsandsetpatientgoals.DatawasanalyzedusingSPSSv17.0.Friedmantestwasusedtocomparereadingsatbaseline,12, and24months. RESULTS: Todate623patientshaveenrolledintheMTMprogram, 41withdiabetesonly,224withdiabetesandoneormorecomorbidities,85with hypertensiononly,and101withhypertensionwithcomorbidities.Atotalof190 patientswithA1Cvaluesand412patientswithbloodpressurevalueswereanalyzed.MeanA1Cvaluesincreasedfrom7.14 1.53to7.28 1.52forallpatients.For uncontrolledpatients,A1Cdecreasedfrom8.40 1.57to8.03 1.63.Diabetes patientsdecreasedsystolicbloodpressure(SBP)from133.59 19.57to129.65 16.04anddiastolicbloodpressure(DBP)from81.95 10.38to79.49 8.72.UncontrolledpatientsdecreasedSBPfrom147.65 16.62to134.02 14.99andDBPdecreasedfrom87.65 6.66to82.31 8.31.HypertensivepatientsdecreasedSBP from134.42 17.28to129.86 14.13andDBPfrom84.76 10.97to81.76 10.10. UncontrolledhypertensivepatientsdecreasedSBPfrom151.89 11.10to134.95 12.56andDBPfrom98.27 6.24to85.71 11.86. CONCLUSIONS: Thepharmacists' effortshavebeenshowntohelpdecreasemeanA1CandBPinpatients.Theimprovementsinthesevalueshelptodecreasetheriskforrelatedcomplicationsand additionalemployercosts. PCV82 EVALUATIONOFRELATIONSHIPBETWEENBLOODPRESSURECONTROL AMONGHYPERTENSIVEPATIENTSANDINTEGRATIONOFSERVICESWITHIN PHYSICIANORGANIZATIONSSmalarzA1,FraserK1,WongK2,WuN1,WogenJ3,BoulangerL1 1UnitedBioSourceCorporation,Lexington,MA,USA,2NovartisPharmaceuticalsCorporation, EastHanover,NJ,USA,3MedMentisConsulting,LLC,Towaco,NJ,USAOBJECTIVES: Toassesstherelationshipbetweenintegrationofservicesandthe managementofhypertensivepatientswithinselectphysicianorganizationsinthe U.S. METHODS: AmodiedversionofthePhysicianPracticeConnectionReadiness Survey(PPC-RS),developedbytheNationalCommitteeforQualityAssurance (NCQA),wasadministeredtochiefmedicalofcersat28U.S.-basedphysicianorganizationsin2010.Theintegrationscoreincludedstructural,functional,andnancialsub-domainswithpossiblescoresof0to100;ahigherscoreindicating greaterintegrationofservices.Demographicandclinicaldatawerecollectedfrom arandomsampleof300hypertensiveadultsselectedforchartreviewateach participatingorganization.Sitecharacteristicsandintegrationscoreswerereported.Hierarchicalregressionmodelswereestimatedtoassesstheassociation betweenanorganization'sintegrationscoreandpatients'outcomes. RESULTS: Of the28participatingsites,mosthadelectronicmedicalrecordsfunctions(71.4%) andhadmorethan50staffmembers(78.6%).Theaverageintegrationscorewas46, withmeanscoresof64,33and42forthestructural,nancial,andfunctional sub-domains,respectively.Integrationscoresweregenerallyhigherinsiteswith 50 physiciansascomparedtositeswith 50physicians(52.8vs.36.6,respectively),andwerealsohigherinsitesownedbyhospitals.Amongthe8,400patients reviewed,approximately60.5%hadcontrolledBP.ApositivecorrelationwasobservedbetweenintegrationscoresandcontrolledBP(Pearsoncoefcient 0.39, p 0.05).Additionally,patientBPcontrolwasbetterinorganizationswithintegrationscoresinthehighestquartile(64.3%)thaninsitesinthelowestquartile (56.8%).Althoughthepointestimateswerenotsignicantinthehierarchical model,regressionanalysessuggestedthathigherintegrationscoreswereassociatedwithbetterBPcontroloutcomes. CONCLUSIONS: Thisstudyprovidessome evidencethatbetterintegrationofserviceinphysicianorganizationsisassociated withbetterpatientoutcomesasreportedviathePPC-RS. PCV83 GAPSANDUNMETNEEDSINANTIPLATELETTHERAPIESFORACUTE CORONARYSYNDROME(ACS)ANDCHRONICCORONARYHEARTDISEASE (CHD)LianJ1,AzmiS2,NavaratnamP3 1NovartisPharmaceuticalsCorporation,EastHanover,NJ,USA,2AzmiBurhaniConsultingSdn Bhd,PetalingJaya,Selangor,Malaysia,3DataMedSolutionsLLC,Hilliard,OH,USAOBJECTIVES: Antiplatelettherapieswithaspirin,ADPreceptorinhibitors,andglycoproteinIIb/IIIainhibitorsplayanimportantroleinmanagementofacutecoronarysyndromesandchroniccoronaryheartdisease.Anassessmentoftheliteraturewasconductedtoidentifyevidencegapsandcriticalunmetneedsin antiplatelettherapyinthemanagementofACSandchronicCHD. METHODS: A targetedliteraturesearchwasconductedonantiplatelettherapywithafocuson ADPreceptorinhibitorsusingPubMed,MedlineandGoogleScholar.Parallel searcheswereinitiatedfortheACSandchronicCHD.Additionalreferenceswere identiedandincludedbasedoncitationsfromreviewedarticles.Atotalof179 articleswerereviewed.Articletypesconsideredinthestudyincludedreviewarticles,clinicaltrials,observationalstudies,andeconomicevaluations. RESULTS: The reviewidentiedgapsandunmetneedsindrugtherapyforACSandchronicCHD usingestablished(aspirin,clopidogrel,andglycoproteinIIb/IIIaagents)andnewer drugs(prasugrelandticagrelor).Keyevidentiarygapswereidentiedontheimpact ofclopidogrelresistance,antiplateletassociatedmajor,minorandnuisancebleedingrisk,impactonhumanisticoutcomes,longtermcost-effectiveness,impacton elderlypatientsandotherhighrisksubpopulations.Ourassessmentrevealedthat currentantiplatelettherapiesareassociatedwithsignicantdownsides,suchas increasedbleedingrisksaswellasotheruntowardeffects. CONCLUSIONS: Our assessmentrevealedimportantgapsandareasofunmetneedassociatedwith currentandnewantiplatelettherapiesinthemanagementofACSandchronic CHD.Furtherstudiesshouldbeconductedtoaddresstheseevidentiarygaps,while unmetpatientneedsshouldbetakenintoconsiderationinfuturedrugdevelopment. PCV84 EFFECTIVENESSOFAMEDICALEDUCATIONINTERVENTIONTOTREAT HYPERTENSIONINPRIMARYCAREMartinezValverdeS1,ReyesMoralesH2,CastroRiosA3,PerezCuevasR3, KlunderKlunderM1 ,SalinasEscuderoG1 1HospitalInfantildeMŽxicoFedericoG—mez,MexicoCity,D.F.,Mexico,2InstitutoNacionalde SaludPœblica,Morelos,Mexico,3MexicanInstituteforSocialSecurity(IMSS),MexicoCity,D.F., MexicoOBJECTIVES: ToreporttheeffectivenessofaCMEinterventiontoimproveappropriatecareforhypertension,whichinturnhasapositiveeffectonbloodpressure controlofhypertensivepatientsinprimarycareclinics. METHODS: Asecondary dataanalysiswascarriedoutusingdataofhypertensivepatientstreatedbyfamily doctorswhoparticipatedintheCMEintervention.Theevaluationwasdesignedas apre-/post-interventionstudywithcontrolgroupinsixprimarycareclinics.The effectoftheCMEinterventionwasanalysedusingmultiplelogisticregression modellinginwhichthedependentvariablewasuncontrolledbloodpressureinthe postinterventionpatientmeasurement.Themodelwasadjustedbyconsidering thefamilydoctorsasaclusterunit.Thisallowedcorrectingtheeffectofincluding severalpatientswhoreceivedtreatmentbythesamedoctor.Selectionofregressionvariableswascarriedoutbycomparing2valuesoftheloglikelihoodinthe models;thiswascarriedoutbyusingthebackwardstepwiseprocedure.Goodnessof-tofthemodelwasevaluatedusingtheHosmerLemeshowtest. RESULTS: AftertheCMEintervention,thenetreductionofuncontrolledbloodpressurebetweenstagesintheinterventiongroupwas10.3%.Themodelresultswerethat beingtreatedbyafamilydoctorwhoparticipatedintheCMEinterventionreduced by53%theprobabilityoflackofcontrolofbloodpressure;receivingdietaryrecom-A47VALUEINHEALTH14(2011)A1–A214

PAGE 56

mendationsreduced57%theprobabilityofuncontrolledbloodpressure.Having uncontrolledbloodpressureatthebaselinestageincreasedtheprobabilityoflack ofcontrolin166%,andpereachunitofincreaseinbodymassindexthelackof controlincreased7%. CONCLUSIONS: CMEinterventionimprovedthemedicaldecision-makingprocesstomanagehypertension,thusincreasingtheprobabilityof hypertensivepatientstohavebloodpressureundercontrol. PCV85 TREATMENTFORDEPRESSIONINWOMENWITHHYPERTENSIONAtkinsE,SambamoorthiU WestVirginiaUniversity,Morgantown,WV,USAOBJECTIVES: Theobjectiveofthestudywastoexamineratesofdepressiontreatmentandvariationsindepressiontreatmentbydemographic,socioeconomic,accesstocare,healthstatus,andlife-stylecharacteristicsamongwomenwith hypertension. METHODS: Thestudydesignwascross-sectional,usingdataon from2006and2007MedicalExpenditurePanelSurvey(MEPS),large-scalesurveys offamiliesandindividualstoprovidenationallyrepresentativeestimatesofhealth careuseandexpenditures.Thestudyincluded1304womenaged22andolderwith hypertensionanddepression,identiedfromMEPSmedicalconditionles.Antidepressantswereidentiedfromprescriptiondrugreportsandpsychotherapywas identiedfromoutpatientvisitsles.Depressiontreatmentpatternsbydemographic,socioeconomic,healthcareaccessandhealthstatuswereanalyzedusing chi-squaretests,logisticregressionandmultinomiallogisticregressions.AllanalysisaccountedforthecomplexdesignoftheMEPSusingSAS9.2. RESULTS: Inour studysample,23.9%hadnodepressiontreatment,56.8%hadantidepressantuse only,and19.3%hadpsychotherapywithorwithoutantidepressants.Among womenwithhypertensionanddepression,AfricanAmericans(AOR 0.47),Latina (AOR 0.46),anduninsured(AOR 0.39)weresignicantlylesslikelytoreportany treatmentfordepressioncomparedtoWhitesandthosewithprivateinsurance. Comparedtonotreatment,psychotherapywithorwithoutantidepressantswas lesslikelyamongthosewithlessthanhighschooleducationandmorelikely amongwomenreportingfair/poormentalhealth. CONCLUSIONS: Nearlyonequarterofwomenwithhypertensiondidnothavetreatmentfordepression.Disparitiesindepressiontreatmentbyrace/ethnicity,healthinsurance,andeducation statuswerenoted.Furtherstudiesneedtoexplorereasonsfornotreceivingdepressiontreatmentandwhethersuchlackoftreatmentfordepressionisassociatedwithpoorhealthoutcomesinthesewomen. PCV86 INSURANCESTATUSANDTHEUSEOFANGIOTENSINIIRECEPTORBLOCKERS (ARBS)ANDANGIOTENSIN-CONVERTINGENZYME(ACE)INHIBITORSAMONG POST-MYOCARDIALINFARCTION(MI)PATIENTSKaisaengN,ZhangJ VirginiaCommonwealthUniversity,Richmond,VA,USAOBJECTIVES: ToevaluatetheeffectofinsurancestatusontheuseofARBsandACE inhibitorsamongpost-MyocardialInfarction(MI)patients. METHODS: The2007 full-yearconsolidateddatafromtheMedicalExpenditurePanelSurvey(MEPS), nationallyrepresentativesurvey,wasusedandlinkedwiththeprescribedmedicinedata.Across-sectionalsurveydataanalysiswasperformedtoevaluatetheuse ofARBsandACEinhibitorsinpatientsdiagnosedwithMyocardialInfarction(MI) anditsrelationshipwithpatients'insurancestatus.TheoutcomemeasuresincludedthebinaryvariablesofACEinhibitorandARBusage.Atotalof464post-MI patientswereincludedintheanalysesbyconductingmultivariatelogisticregression,controllingforsocio-economicanddemographicvariables. RESULTS: Outof 464patientswithMI,67(14.44%)usedonlyARBs,244(52.59%)usedonlyACEinhibitors,15(3.23%)usedbothdrugs,and138(29.74%)usedneitherARBsnorACE inhibitors.PatientswithMedicarecoverageonlyandpatientswithoutinsurance werelesslikelytouseARBs,comparedtopatientswithprivateinsurance(adjusted OR0.26,p 0.018;adjustedOR0.03,p 0.064,respectively).Therewasnosignicant differencebetweenhealthinsurancestatusanduseofACEinhibitors.Furthermore,healthinsurancestatuswasnotsignicantlydifferentamongpatientswho usedneitherARBsnorACEinhibitors.AmongMIpatientswhousedbothARBsand ACEinhibitors,patientswithMedicarecoverageonlywerelesslikelytouseboth drugs(adjustedOR0.02,p 0.001),relativetopatientswithprivateinsurance.Additionally,patientswitheducationalattainmentofhigherthancollegedegrees weremorelikelytousebothdrugsthanthosewhodidnotnishcollegedegrees (adjustedOR19.61,p 0.042). CONCLUSIONS: HealthinsurancesignicantlyaffectedtheusageofmoreexpensiveARBsbutnotlessexpensiveACEinhibitors. Policymakersneedtobeawareofthemoralhazardarisingfromtheinsurance coverageindruguse. PCV87 FACTORSASSOCIATEDWITHSELECTIONSOFSTATINSAMONGELDERLY PATIENTSZhaoZ1,PengX1,BaeJP1,SponsellerCA2,LeNarzLA1 1EliLillyandCompany,Indianapolis,IN,USA,2KowaPharmaceuticalsAmerica,Inc., Montgomery,AL,USAOBJECTIVES: Toassessdemographicandclinicalfactorsassociatedwithstatin selectionamongelderlypatients. METHODS: Aretrospectivecohortstudywas conductedtoexaminepredictorsofstatinselectionamongpatientsaged365years, whoinitiatedpravastatin(PS)vs.simvastatin(SS),atorvastatin(AS),orrosuvastatin(RS)between1/1/2007and12/31/2007.Indexstatinusewasdenedastherst statinclaimfollowingatleast90daysofnostatinaccess.MultiplelogisticregressionmodelswereemployedtoassesspredictivefactorsofPSinitiationversus otherstatininitiations. RESULTS: Of96,450statinusersidentied,therewere8,165 PSinitiators,38,099ASinitiators,11,320RSinitiators,and38,866SSinitiators. Comparedtootherstatinusers,ahigherpercentageofPSinitiatorswereaged75-85 (PS:42.3%,SS:41.0%,AS:41.1%,RS:38.0%, P 0.01)andmorelikelytobefemale(PS: 56.6%,SS:51.3%,AS:51.5%,RS:55.7%, P 0.01).PSinitiatorsweremorelikelytohave atrialbrillation(PS:10.0%,SS:9.4%,AS:9.6%,RS:8.6%, P 0.01)andtakewarfarin (PS:11.7%,SS:9.8%,AS:10.4%,RS:9.9%, P 0.01)andtriazoles(PS:3.1%,SS:2.4%,AS: 2.6%,RS:2.8%, P 0.01)inthebaselineperiod.AhigherpercentageofPSinitiators tookmorethan3uniquemedications90-daypriortotheindexdate(PS:85.4%,SS: 81.8%,AS:83.0%,RS:83.9%, P 0.01).Aftercontrollingfordemographicandclinical characteristics,useofwarfarinwasassociatedwithinitiatingPScomparedwithSS, AS,andRS.OtherpredictorsofPSinitiationincludedpriorhistoryofatrialbrillation,HIVinfectionanduseofcalciumchannelblockers,ezetimibe,andfenobratesoverthe1-yearpre-indexperiodcomparedwithASandSS. CONCLUSIONS: PatientprolesweredifferentamongPSuserscomparedtootherstatinusers. SelectedcomorbiditiesandprioruseofcertainmedicationsweresignicantpredictorsofPSinitiationamongacohortofMedicarepatients. PCV88 PATTERNSOFSTATINPRESCRIPTIONAMONGPRIVATELYINSURED COMMERCIALANDMEDICAREPATIENTSBaeJP1,PengX1,ZhaoZ1,ZagarA1,SponsellerCA2,LeNarzLA1 1EliLillyandCompany,Indianapolis,IN,USA,2KowaPharmaceuticalsAmerica,Inc., Montgomery,AL,USAOBJECTIVES: In2008,therewerecurrently6statinsavailableintheU.S.marketand morethan193millionstatinprescriptionswerewritten.Wesoughttoexaminethe patternsofstatinuseamongprivatelyinsuredcommercialandMedicarepatients intheU.S. METHODS: AretrospectiveanalysiswasperformedusingMarketScan CommercialandMedicaredatafromJan.2008toDec.2008.Twocohortswere created:1)thecommercialcohort(CC)and2)theMedicarecohort(MC).Statin scriptsofminimum30daysupplywereextractedfromthepharmacyclaimsdata andindividualdemographicandclinicalinformationwerelinkedfromtheirmedicalandadministrativerecords.Prescriptionpatternswereexaminedandanalyses byageandgenderwereperformed. RESULTS: Therewere18millionand9.8million statinprescriptionslledbytheCCandMC,respectively.Inbothcohorts,simvastatinaccountedforthelargestshareofstatinprescriptions(MC:44.1%;CC:44.0%), followedbyatorvastatin(MC:31.6%;CC:31.1%),rosuvastatin(MC:7.8%;CC:11.2%), pravastatin(MC:8.1%;CC:6.5%),lovastatin(MC:7.4%;CC:6.6%)anduvastatin (MC:0.9%;CC:0.6%).Themajorityofstatinprescriptionswereforgenerics(MC: 59.6%;CC:57.1%).Theaverageannualnumberofstatinprescriptionperuserwas 9.3fortheMCand7.6fortheCC.IntheCC,58.1%ofprescriptionswerelledby male,while50.4%ofprescriptionswerelledbymaleintheMC.Diabeticpatients accountedfor35.2%ofprescriptionsintheMC,and26.8%intheCC.Milligram dosingdistributions(mean;median;mode)wereforsimvastatin(MC:32.1;20;20 CC:32.9;20;40);atorvastatin(MC:24.3;20;10CC:24.7;20;10);rosuvastatin(MC:12.7; 10;10CC:13.1;10;10);andpravastatin(MC:37.6;40;40CC:38.2;40;40). CONCLUSIONS: Thisanalysissuggestsdifferentstatinusepatternsbyage,gender,andinsurance. Mostfrequentlydispenseddoseswere20mgforsimvastatin,10mgforatorvastatinandrosuvastatin,and40mgforpravastatin. PCV89 COMPARISONOFDOSINGPATTERNSAMONGPATIENTSINITIATINGSTATIN THERAPYINAMANAGEDCAREPOPULATIONZhaoZ1,PengX1,BaeJP1,DungeyJ1,FariesDE1,SponsellerCA2,WetmoreS1,YuCY1, LeNarzLA1 1EliLillyandCompany,Indianapolis,IN,USA,2KowaPharmaceuticalsAmerica,Inc., Montgomery,AL,USAOBJECTIVES: Tocomparedosingpatternsamongpatientsinwhomatorvastatin (AS),simvastatin(SS),rosuvastatin(RS),orpravastatin(PS)wasnewlyprescribed. METHODS: UsingalargeUSmanagedcaredatabase,studypatientsaged 18who werenewlyprescribedAS,SS,RS,orPSbetween10/01/2006-09/30/2007wereidentiedwithaminimum12monthspre-and24monthspost-indexhealthplan eligibility.A6-monthwashoutperiodfreeofuseofanystatinwasappliedtoidentifypatientsinwhomstatintherapywasnewlyprescribed.Theindexdatewas denedasthedateoftherstprescriptionllforstatinsand4cohortswerecreated basedonindexstatins:AS,SS,RS,orPS.Patientswerefollowedupto24months aftertheindexdate,andtheaveragedailydosefor12prescriptionsperperson, average/mediandailydoseineachoftherst12prescriptions,andtheaverage dailystatincostswerecomparedamong4cohorts. RESULTS: Atotalof142,692 patientswereidentied,comprisedof47,972ASpatients,21,475RSpatients,62,954 SSpatients,and10,291PSpatients.Theaveragedailydosesofallprescriptions were21.3mgforASpatients,10.2mgforRSpatients,29.2mgforSSpatients,and 34.6mgforPSpatients.Theaverage/mediandailydoseforthe1stand12thprescriptionswere20.8mg/20mgand21.2mg/20mgforASpatients,10.0mg/10mgand 10.5mg/10mgforRSpatients,28.4mg/20mgand28.7mg/20mgforSSpatients,and 32.8mg/40mgand35.9mg/40mgforPSpatients.Theaveragedailystatincostsfor the1stand12thprescriptionswere$2.9and$3.2forASpatients,$2.7and$3.1for RSpatients,$1.5and$0.8forSSpatients,and$1.4and$0.9forPSpatients. CONCLUSIONS: PatientsinitiatingAS,RS,SS,andPSexperiencedlittleescalation oftheirstatindailydosingoverthe24-monthfollow-upperiod. PCV90 FACTORSASSOCIATEDWITHSELECTIONSOFSTATINSAMONGPATIENTSINA LARGEEMPLOYER-BASEDCLAIMSDATABASEZhaoZ1,PengX1,BaeJP1,SponsellerCA2,LeNarzLA1 1EliLillyandCompany,Indianapolis,IN,USA,2KowaPharmaceuticalsAmerica,Inc., Montgomery,AL,USAA48VALUEINHEALTH14(2011)A1–A214

PAGE 57

OBJECTIVES: Toassessdemographicandclinicalfactorsassociatedwithstatin selectionamongpatientsinalargeemployer-basedclaimsdatabase. METHODS: Thisstudyexaminedpredictorsofstatinselectionamongstatinuserswhowere18 to64yearsoldandinitiatedpravastatin(PS)vs.simvastatin(SS),atorvastatin(AS), orrosuvastatin(RS)between1/1/2007and12/31/2007.Indexstatinusewasdened astherststatinclaimfollowingatleast90daysofnostatinaccess.Multiple logisticregressionmodelswereemployedtoassesspredictivefactorsofPSinitiationversusotherstatininitiations. RESULTS: Ofthetotal1,336,433statinusers identied,therewere14,122PSinitiators,66,156ASinitiators,27,062RSinitiators, and77,180SSinitiators.Comparedtootherstatinusers,patientsinitiatingPSwere older(PS:54.0,SS:53.3,AS:53.9,RS:53.5, P 0.01)andmorelikelytobefemale(PS: 52.3%,SS:47.6%,AS:45.4%,RS:47.0%, P 0.01).AhigherpercentageofPSinitiators haddiabetes(PS:20.1%,SS:17.9%,AS:17.9%,RS:18.3%, P 0.01),liverdisease(PS: 2.7%,SS:1.9%,AS:2.1%,RS:2.4%, P 0.01),humanimmunodeciencyvirusinfection(HIV)(PS:0.7%,SS:0.1%,AS:0.3%,RS:0.3%, P 0.01),andhypertension(PS: 45.1%,SS:40.8%,AS:41.1%,RS:43.9%, P 0.01)inthebaselineperiod.Aftercontrollingfordemographicandclinicalcharacteristics,olderage,femalegender,history ofdiabetes,liverdisease,HIV,andhypertensionweresignicantpredictorsofPS initiation.OtherpredictorsofPSinitiationincludeduseofcalciumchannelblockers,ezetimibe,fenobrates,andproteaseinhibitorsoverthe1-yearpre-indexperiod.Primarycarephysician(PCP)astheprescriberwaspositivelyassociatedwith initiatingPScomparedtoRSorSS. CONCLUSIONS: Theanalysesindicatethatolder age,femalegender,presenceofselectedcomorbidities,prioruseofcertainmedications,andPCPasprescriberarepredictorsofPSinitiationamongworkingage, commerciallyinsuredpatients. PCV91 BETABLOCKERSANDDIURETICSUSAGEINCROATIAANDSLOVENIADURING ANINE-YEARPERIODVitezicD1,KucanM2,VitezicM3,KovacevicM1,MrsicPelcicJ4 1UniversityofRijekaMedicalSchoolandUniversityHospitalCentreRijeka,Rijeka,Croatia,2JADRANGalenskilaboratorij,Rijeka,Croatia,3UniversityofRijekaMedicalSchooland OrthopedicHospitalLovran,Lovran,Croatia,4UniversityofRijeka,Rijeka,CroatiaOBJECTIVES: Cardiovasculardiseasesarethemajorhealthproblemandadequate therapy,whichincludesusageofbetablockers(ATC,C07)anddiuretics(ATC,C03), leadstoreductionofcardiovascularmorbidityandmortality.Theaimofourstudy wastoidentifyandanalyzechangesintheusageofthesedrugsinCroatiaand Sloveniafrom2000-2008andtoidentifytherateofthegenericdrugsusageaswell astheaveragepricefor1DDD. METHODS: Thedataconcerningconsumptionhave beenobtainedfromtheInternationalMedicalStatisticsdatabaseforCroatiaand theyarepresentedindeneddailydosesper1000inhabitantsperday(DDD/1000). FinancialexpendituredataarepresentedinEurosandtheaveragecostperDDD wascalculatedforeachdruggroup. RESULTS: Thetotalusageofdiureticsison average19%higherinCroatiaandcontinuallyincreasinginbothcountries.ConsumptionrateoffurosemidamongdiureticsislowerinSlovenia(43.96%in2008), whileinCroatia67.28%(2008).ThegenericdrugsusageishigherinCroatiaduring thewholeinvestigatedperiod,e.g.in2008,87.4%ofallprescribeddiureticswere genericsinCroatia,andinSlovenia59.4%.Theaveragepriceper1DDDwasmore thantwicelowerinCroatia(0.08EUR/DDDvs.0.17EUR/DDDinSlovenia).Thebeta blockersdrugsusageis13%higherinSlovenia,butgenericdrugsusageishigherin Croatia. CONCLUSIONS: Increaseddiureticsandbetablockersdrugusageiscomparabletoworldwidetrends,butdrugprescribingpatternsaredifferentwhen comparingbothcountries.Althoughthegenericdrugsusageisrelativelyhighin Croatia,itshouldbefurthersupportedandpromoted.Accordingtoourinvestigationthereisapossibilityforfurtherpricereductionofdiureticsandbetablockersin Slovenia.Thisisimportantasatooltodecreasecostsofhealthcaresystems. PCV92 CHANGESINTHEPREVALENCE,TREATMENTANDCONTROLOF HYPERTENSIONINGERMANY?ACLINICAL-EPIDEMIOLOGICALEXAMINATION OF50,000PRIMARYCAREPATIENTSWITHTHEDETECTSTUDYLabeitA1,KlotscheJ2,PieperL2,WehlingM3,WittchenHU2 1UniversityofLeicester,Leicester,UnitedKingdom,2TechnischeUniversitŠtDresden,Dresden, Germany,3UniversitŠtHeidelberg-MedizinischeFakultŠtMannheim,Mannheim,GermanyOBJECTIVES: Toestimate2003pointprevalence,treatmentandcontrolratesof hypertensionfordifferentagegroupsofprimarycareattendeesinGermany,to examinefactorsassociatedwithpharmacotherapyandadequatebloodpressure controlandtoexaminewhetherchangeshaveoccurredinratesoftreatmentand controlsince2001. METHODS: Usingstandardizedclinicaldiagnosticandtreatmentappraisalforms,bloodpressurelevelsandpatientquestionnairesfor55,518 participantsfromtheDETECTstudy(2003)wereanalyzed. RESULTS: Accordingto physiciandiagnosisoverallprevalencewas35.5%,accordingtoNHANEScriteria 56.0%.AmongthosedenedbyNHANEScriteria,treatmentandcontrolrateswere 56.0%and20.3%,respectively.Signicantpredictorsforreceivingantihypertensive medicationwere:increasingage,femalesex,obesity,previousmyocardialinfarctionandtheprevalenceofcomorbidconditionssuchasCHD,dyslipidemiaand diabetesmellitus.SignicantpositivepredictorsofadequatebloodpressurecontrolwereCHDandantihypertensivemedication.Inadequatecontrolwasassociatedwithincreasingage,malesexandobesity.Ascomparedtondingsfrom2001, ratesoftreatmentandcontrolindicatesomeimprovementacrossallagegroups. CONCLUSIONS: RatesoftreatedandcontrolledhypertensionaccordingtoNHANES criteriainDETECTremainconsiderablylowascomparedtotheratesof2001, despiteconsistentindicationsofsomeimprovement.Hypertensivepatientswith riskfactorssuchasobesityandassociatedcomorbiditiessuchasCHD,dyslipidemiaanddiabetesweremorelikelytoreceiveantihypertensivedrugsinaccordance withguidelines.Adequatebloodpressurecontrolwasachievedinmorepatients withobesityandCHD,butnotthosewithdiabetesmellitusordyslipidemia. PCV93 HEALTHEXPENDITURECOMPARISONOFMETOPROLOLSUCCINATEVS METOPROLOLTARTARATEVaidyaV,PatelP UniversityofToledo,Toledo,OH,USAOBJECTIVES: Tocomparehealthcareexpendituresofhypertensivepatientson metoprololtartarate(MT)andmetoprololsuccinate(MS).Metoprolol,aselective-1 blocker,isavailableintwodifferentsaltformsinthemarket-metoprololsuccinate (MS)andmetoprololtartarate(MT).BothofthesesaltsareFDAapprovedforthe treatmentofhypertension.Severalstudieshaveshownsimilarefcaciesbetween thetwosalts;howevertheyarepricedatdifferentlevelsbecauseoftheirpharmacokineticproperties.MTistheshortactingsaltandisgiventwicedaily,whereas MSisthelongactingsaltandisgivenoncedaily. METHODS: TwocohortsofpatientsusingMTandMSwereselectedfrom2008MedicalExpenditurePanelSurvey (MEPS).Propensityscore(PS)matchingtechniquewasusedtobalancethecohorts onvariousparameterssuchasdemographicinformation,insurancestatusand comorbityscore.PatientsusingMTwerematchedtopatientsusingMSonthelogit ofpropensityscoreusingcalipersofwidthequalto0.2ofthestandarddeviationof thelogitofthepropensityscore.Toestimatemeanexpenditurecostsand95% condenceintervals,nonparametricbootstrapsamplingmethodwasused. RESULTS: Atotalof742patientswerefoundtousemetoprolol(MT-388,MS-354). AfterPSmatching,atotalof594patientswereleftinthesamplefornalanalysis (298patientsineachcohort).Theaverageannualexpenditureintwogroupsdidnot differsignicantly.Aftercarryingoutnonparametricbootstrapsamplingusing 1,000samples,MTcohorthad$10,779(95%CI:$10,682-$10,875)averageannual expenditurewhileMScohorthad$9,810(95%CI:$9,810-$10,003). CONCLUSIONS: BoththesaltsofMetoprololwerefoundhavesimilaraverageannualtotalhealthcareexpenditure,howeveronceadailysalt(MS)hashigheroutofpocketcost. PCV94 ANTICOAGULANTUSE,THEPREVALENCEOFBRIDGINGANDRELATIONTO LENGTHOFSTAYAMONGHOSPITALIZEDPATIENTSWITHNON-VALVULAR ATRIALFIBRILLATIONSmoyer-TomicK1,SanderS2,SiuK2,SmithDM3,AminA4,JohnsonBH1 1ThomsonReuters,Washington,DC,USA,2BoehringerIngelheim,Ridgeeld,CT,USA,3Thomson Reuters,Montvale,NJ,USA,4UniversityofCalifornia,Irvine,Irvine,CA,USAOBJECTIVES: ASA/AHAguidelinesforstrokepreventioninpatientswithnon-valvularatrialbrillation(NVAF)recommendanticoagulant(AC)therapywithavitaminKantagonist(e.g.,warfarin)forpatientswithouthemorrhagiccontraindication.Bridgingwithlowmolecularweightheparin(LMWH)wasrecentlyaddedto theseguidelines(ClassIIa),butwithlittlesupportiveevidence(LevelofEvidenceC). LittleisknownabouttheuseofACtherapyintheinpatientsettingandhowit relatestohospitallengthofstay(LOS). METHODS: PatientswithanNVAFdischarge diagnosisbetween7/1/04-9/30/09wereidentiedfromadministrativeclaimsdata representingoverahundredUShospitals.InpatientACusebydrugtypewasreportedtoidentifypatientsreceivingwarfarinaloneandincombinationwith LMWH/pentasaccharide(PS)orunfractionatedheparin(UFH),indicativeofbridging.LOSwasmeasuredandreportedbyACcombinations.Non-parametricstatisticaltestswereusedtocompareLOSamongpatientsreceivingwarfarinonlyto thosewithbridgetherapy. RESULTS: Of6340NVAFpatients,76%receivedinpatient ACtherapy,with3037(48%)receivingwarfarin.Ofpatientsreceivingwarfarin,64% werebridgedwithanotherAC.LMWH/PSwasthemostcommonbridgeagent(45% ofallbridgedpatients),followedbyUFH(36%),andbyLMWH/PSandUFHcombined (18%).ThemeanLOSforpatientsreceivingbridgetherapieswas6.3days(SD 6.1) comparedto4.2days(SD 3.1)forwarfarinalone,p .0001.Bybridgingdrugtype, meanLOSwas5.6days(SD 5.3)forLMWH/PS;6.0days(SD 4.9)forUFH,and8.4 days(SD 8.4)forLMWH/PSandUFH. CONCLUSIONS: LessthanhalfofallNVAF patientsreceivedwarfarinwhilehospitalized.Amongthosetakingwarfarin,bridgingwithotheranticoagulantswascommonandwasassociatedwithlongerLOS thanpatientsreceivingwarfarinalone. PCV95 USINGCOMPUTERIZEDPHYSICIANORDERENTRYOFANTI-HYPERTENSIVE MEDICATIONSTOCONNECTINDICATIONSWITHPRESCRIBINGANDIMPROVE PROBLEMLISTDOCUMENTATIONGalanterW,WaltonS,FalckS,RosencranzH,AdimadhyamS UniversityofIllinoisChicago,Chicago,IL,USAOBJECTIVES: ElectronicMedicalRecordsoftenlackaccurateproblemlists.The objectivesofthisstudyweretoexamineifalertsrequestingindicationinformation duringorderentryofmedicationsusedtotreatacommondisease,hypertension, canimproveproblemlists. METHODS: Inpatientordersgeneratedbypreviously implementedclinicaldecisionsupport(CDS)inanacademichospitalusingCPOE werereviewed.CDSwasdevelopedforclassesofantihypertensiveand3additional diagnoses:CHF,benignprostatichypertrophyandnephropathy.Fivelogicalmedicationgroupsweredeveloped.Arandomsampleof250chartswerecollectedfor eachmedicationgroupandreviewedbyphysicianstodeterminealertyieldand addedproblemaccuracy.Theaccuracyandyield(problemsaddedperorder)were analyzedacrossthemedicationgroups,cliniciantypes,andclinicalvenues. RESULTS: Preliminaryresultsobtainedfromthegroupofmedicationsthatonly treatHTNandnootherdiagnoses(HTNgroup)showedanaccuracyofadditionsto theproblemlistof92%andtheyield,was76%.Forthegroupofmedicationswhich treatedHTNandCHF(CHFgroup),theaccuracyandyieldwerelower,81%and62% respectively.LookingattheHTNgroupalone,theaccuracywas100%forattendingA49VALUEINHEALTH14(2011)A1–A214

PAGE 58

physiciansand55%forresidentsphysicians.Similarlytheaccuracywashigherin theambulatorysettingthanforinpatient,89%vs.62%.TheyieldfortheHTNgroup wasalsohigherwithattendingsthanresidents,100%vs.67%,andtheyieldwas higherinambulatoryversusinpatientsettings,87%vs.64%. CONCLUSIONS: CDS basedonmedicationordersduringCPOEmayimproveproblemlists.Theaccuracy andyieldofdiagnosticinformationissensitivetomedicationtype,clinicalvenue andpossiblycliniciantype.Mandatoryindicationbasedprescribingmaynotproduceveryaccurateindicationdata.Thereisasubstantialneedforfurtherresearch inthisarea. PCV96 TRENDINTHEUTILIZATIONOFPHARMACOLOGICALANDNONPHARMACOLOGICALTREATMENTSTRATEGIESINTHEMANAGEMENTOF NEWLYDIAGNOSEDATRIALFIBRILLATIONPATIENTSINTHEUNITEDSTATESKashyapA,LiC UniversityofArkansasforMedicalSciences,LittleRock,AR,USAOBJECTIVES: Treatmentstrategiesforatrialbrillation(AF)areclassiedintorate controlandrhythmcontrol.Theycanbeemployedthroughtheuseofpharmacologicalandnon-pharmacologicalapproaches.Thisstudydescribedthetrendinthe utilizationoftreatmentstrategiesforthemanagementofAFinnewlydiagnosed patientsbetween1999and2008. METHODS: Thestudywasaretrospectivecohort studyofadministrativeclaimsdatafromalargenationallydispersedgroupof commerciallyinsuredsubjectsfrom1999-2008.NewlydiagnosedAFpatientsaged 18yearsandwithatleastoneclaimforanAFrelatedinterventionwithin12 monthsofdiagnosiswereidentied.Basedoninitialtreatmentreceived,patients wereclassiedinpharmacotherapyornon-pharmacotherapytreatmentgroups. Patientsinitiatedondrug-therapieswereclassiedintorate-control(beta-blocker, calciumchannelblockerordigoxin)orrhythmcontrolgroups.Descriptiveanalysis wasconductedtocomparethepatientdemographiccharacteristicsacrossthe treatmentgroups.OLSregressionwasconductedtoassesslineartrends. RESULTS: Ofthe6284newlydiagnosedAFpatientswhoreceivedtreatmentwithinone-year, 83.7%underwentpharmacotherapyasthe1stlinetherapy.Majority(86.3%)ofpa tientswhoinitiateddrugtherapyreceivedrate-controlmedicationsonly.Relative distributionofdifferenttreatmentsacrosstimedidnotvarysignicantly.However, withinrate-controlgroup,beta-blockersuseincreasedsignicantly(p 0.0001) from47.5%(1999)to70.58%(2008),whereasuseofcalciumchannelblockersand digoxindecreased.Non-pharmacotherapywasemployedmorecommonlyinthe mid-west(44.95%,p 0.0002),inpatientsyoungerthan65years(58.35%,p 0.0001) anditsusedecreasedwithincreasingage.Malesweremorelikelytoundergo non-pharmacotherapy(69.9%,p 0.0001). CONCLUSIONS: Higherutilizationof ratecontrolas1stlinetherapycomparedtoothertreatmentstrategiesisconsistent withthecurrentliterature.Greateruseofnon-pharmacotherapyinyoungerpatientscanbepartiallyexplainedbythehighersuccessrateinthispopulation. PCV97 ISCARDIOVASCULARMEDICATIONUSAGEAMARKERFORCARDIOVASCULAR RISK?POTENTIALIMPLICATIONFORSMART-EDITFORMULARYMANAGEMENT STRATEGIESHunterTD1,GunnarssonC1,BullanoMF2,deVanePJ2 1S2StatisticalSolutions,Inc.,Cincinnati,OH,USA,2AstraZenecaPharmaceuticalsLP, Wilmington,DE,USABACKGROUND: Inthecurrentstatinlandscape,aneedhasemergedforsimpleand lessdata-intensivetoolstoidentifypatientsatriskforcardiovasculareventsandto appropriatelyidentifypatientsforhigher-efcacystatinuse. OBJECTIVES: Toexaminethefeasibilityofapharmacy-basedalgorithmasamarkerofcardiovascular risktoguideappropriatestatinselectionatthepointofservice. METHODS: A retrospectiveanalysisusinganintegrateddatabaseofover70pooledhealthplans wasconductedfortheperiod2003tomid-2010.Patientsonatleastonecardiovascularmedicationofinterest(N 4,276,418)and 18yearsoldwerematchedbyage andgendertocontrolsubjects(notoncardiovascularmedications[N 3,046,021]). Cardiovascularmedicationsincludedwereantihypertensive,antidiabetic,antianginal,antiarrhythmic,anticoagulant,non-statinlipidmodier,andantiplatelet drugs.Subjects'cardiovascularriskstatus(high/moderate/low)wasassessedaccordingtothecurrentNationalCholesterolEducationProgramguidelines,from recordsofdiagnosesandprocedures.Relativeriskswerecalculatedforthelikelihoodofbeingatincreasedrisk(highormoderateriskstatus)astheratioofproportionsofpatientsoncardiovascularmedicationversuscontrolswhowereat increasedrisk. RESULTS: Thestudypopulationhadameanageof53years.Overall, apatientreceivinganycardiovascularmedicationwas2.4timesmorelikelytobeat increasedrisk(95%CI2.39,2.40)andwas2.7timesmorelikelytoexperienceacardiovascularevent(95%CI2.73,2.74)thanasubjectnotreceivinganycardiovascular medication.Additionally,itwasobservedthat27%ofthetotalpatientsinthe healthplanswhowereoncardiovascularmedicationsaccountedfor71.5%ofall cardiovasculareventsinthatgroup. CONCLUSIONS: Apharmacyalgorithmbased onusageofvariouscardiovascularmedicationscanbeusedtoidentifyincreased cardiovascularrisk.Theseinitialndingsdemonstratethepotentialofusing widelyavailablecardiovascularprescriptiondataindesigningstatinformulary managementstrategies. PCV98 USEOFELECTRONICMEDICALRECORDSTOASSESSTHECLINICALAND ECONOMICIMPACTOFNON-MEDICALSWITCHINGBETWEENDIFFERENT ANGIOTENSINRECEPTORBLOCKERS(ARBS)KamalKM1,ZackerC2,CivitareseL3 1DuquesneUniversity,Pittsburgh,PA,USA,2NovartisPharmaceuticals,Emmaus,PA,USA,3PreferredPrimaryCarePhysicians,Carnegie,PA,USAOBJECTIVES: Electronicmedicalrecords(EMRs)increasinglyarebeingusedbyoutcomesresearcherstostudythequalityofmedicalcare.Thestudyassessedthe impactofnon-medicalswitchesinARBtherapyonbloodpressure(BP)andmedical resourceutilizationincludingofcevisitsandotherantihypertensivemedications. METHODS: AretrospectivestudywasconductedusingtheGECentricityEMRofa primarycarephysiciansgroup.Hypertensivepatients 18years,continuouslyenrolled,andreceivingARBtherapywereincluded.Switchmusthaveoccurredafter thepatientwasontherstARBforatleast60days.Thecontrolgroupremainedon thesameARBanddidnotswitch.A1:1matchingwasdoneonthedurationofinitial ARBtherapyandpresenceofdiabetesorrenalimpairment.Demographiccharacteristics,clinicalparameters,ARBuse,andofcevisitswereidentied.Datawas extractedusingMicrosoftSQLandstatisticalanalyseswereconductedusingSPSS version18.0. RESULTS: Atotalof4,851patients(meanage68.23years,female 58.1%)wereprescribedanARBbetween2004and2008outofwhich3,083(63.6%) stayedononeARBand1,768(36.4%)switchedfromoneARBtoanother.Matched pairsofswitchers(n 357)andcontrolgroup(n 357)werethenidentied.There wasnodifferenceinmeanbaselinesystolicBP(SBP)forswitchers(143.79 21.22) andcontrol(144.20 22.16)butswitchershadhigherpost-switchSBP (141.06 18.52)thancontrol(137.97 20.58)(p 0.035).Moreswitcherslostcontrol fromindextorstvisitpost-indexperiod(11.5%)thancontrol(8.7%).Moreantihypertensiveagentswereaddedinswitchers(1.91 1.7)thancontrol(1.02 1.2)inthe post-indexperiod(p 0.001). CONCLUSIONS: EMRshavethepotentialtobridge researchwithclinicalcarebyprovidingreal-worlddata.Thisstudydemonstrates thatnon-medicalswitchesmayresultinlossofBPcontrolandadditionalmedical resourceuse.Thus,carefulconsiderationshouldbegivenbeforeswitchingtherapiesfornon-medicalreasons. PCV99 MEDICALRESOURCEUTILIZATIONANDCOSTSFOLLOWINGHOSPITALIZATION OFPATIENTSWITHCHRONICHEARTFAILUREINTHEUNITEDSTATESKorvesC1,Eldar-LissaiA1,RodermundD1,SwallowE1,CummingsAK1,ArielyR2, DuhMS1 1AnalysisGroup,Inc.,Boston,MA,USA,2NovartisPharmaceuticalsCorporation,EastHanover, NJ,USAOBJECTIVES: Thestudyobjectivewastodeterminemedicalresourceutilization anddirectandindirectcostsfollowinghospitalizationwithchronicheartfailure (HF). METHODS: Patientswith 1hospitalizationwithachronicHFclaim(ICD-9 428.22,428.32or428.42)wereidentiedinaUScommercialinsuranceclaimsdatabasefrom2004-2008.Patientswereobservedfrombeginningofrsthospitalization(indexhospitalization)forchronicHFuntildisenrollmentorendofdataavailability.Inpatient,outpatient,andprescriptiondrugdatawereusedtoestimateper patientpermonth(PPPM)utilizationrates.Costs(2009USD)werecalculatedper hospitalizationandPPPMforpatients 65years,andincludedinsurers'reimbursement,patientout-of-pocketandsickleavecosts. RESULTS: Therewere7,814patients(meanage73.2years,55.7%male)meetinginclusioncriteria.MeanHFhospitalizationlengthofstayincreasedfrom6.7daysatindexhospitalizationto8.2 daysatfourthre-hospitalization.RateofHF-relatedre-hospitalizationremained over0.045PPPMthroughout24monthsoffollow-up,accountingforover78%of all-causehospitalizations.Rateofall-causeandHF-relatedoutpatientvisits peakedat4.0and0.59visitsPPPM,respectively,withinthethreemonthsafter indexhospitalization.Indexhospitalizationwasmostexpensive(directmedical costs $31,998).Patientout-of-pocketcostsaccountedforlessthan10%ofdirect hospitalizationcostsandsickleavecostswerelessthan$1,800atanyhospitalization.Duringthestudyperiod,outpatientcardiovasculardrugsaccountedfora smallproportionoftotalpharmacycosts(approximately20%);averagePPPMcost variedfrom$88to$124,representinglessthan1%oftheaveragecostofaHFrelatedhospitalization. CONCLUSIONS: TreatingchronicHFpatientsisresource intensive.Thegreatestutilizationandcostburdenoccurwithinthethreemonths afterindexhospitalizationandpatientscontinuetobeburdenedafterhospitalizationbyhighinpatientandoutpatientvisitrates. PCV100 HEALTHCAREUTILIZATIONANDCOSTSFORAMEDICATIONTHERAPY MANAGEMENT(MTM)PROGRAMPintoSL,ParthaG UniversityofToledo,Toledo,OH,USAOBJECTIVES: 1)Todeterminecostsandutilizationincurredbyemployeesfollowingenrollmentinanemployer-sponsoredmedicationtherapymanagement(MTM) programand2)toassesstheimpactofattritiononhealthcareexpenditures. METHODS: Alongitudinalstudyusingmedicalclaims.Studyparticipantswere LucasCountyemployeeswithdiabetes,hypertensionand/orhyperlipidemiaparticipatinginanemployer-sponsoredopen-enrollmentMTMprogram.Claimsfor ofce,emergencyroomandinpatientvisitswereanalyzedfortheperiodofJanuary 2005-July2010.Disease-relatedexpendituresweretrackedforaperiodoftwo-year pre-andpostenrollment.Allexpenditureswerecalculatedasaveragecostsperpatientpreandpostjoiningandalsoastotalexpenditurebornebytheemployer. Dropoutswereidentiedfromanappointmentdatabase.Expenditureswerecalculatedusingclaimsforoneyearbeforeandafterthedropoutdate. RESULTS: Claimswerereceivedfor361employeesthatenrolledintheMTMprogram.Ofce visitexpenseswentdownby$71,442.29orby22.36%afterjoiningtheMTMprogram.Anincreaseinspendingforemergencyroomvisitsby$12,597.16wasobserved,therewerenorecurringpost-enrollmentvisitsandthosethatoccurred wereforindicationsthatcouldhaveincreasedlong-termspendingfortheemployer.Totalexpenditureoninpatientvisitswentupbyapproximately$7600but theamountspentoneachvisitwentdownfrom$7,746.36 4,169.13to4,408.07 A50VALUEINHEALTH14(2011)A1–A214

PAGE 59

6,204.77.Numberofemployeeswhohadaninpatientvisitincreasedfrom3to7, withnorecurrenceofpre-visitevents.Adecreaseintotalhealthcareexpenditures byover14%wasobserved.Preliminaryanalysisforthesecondobjectiveshowsthat onaverageemployeesspent$406.97/patient/yearmorewhentheydroppedoutof theprogramthaniftheystayedenrolled. CONCLUSIONS: PharmacistledMTMprogramhelpedreducehealthcareexpenditurefortheemployer.Improvingretentionfortheprogramcouldhelpsubstantiatethesecostsavings. PCV101 EXPLORINGGENDERDISPARITIESINPREVENTIVECAREUTILIZATION AMONGSTTHEUNITEDSTATESPOPULATIONVaidyaV,ParthaG,KarmakarM UniversityofToledo,Toledo,OH,USAOBJECTIVES: Toidentifyexistinggenderdifferencesinutilizationofpreventive careservicesintheUnitedStates(US)populationusinganationaldatabase.Few studieshavepointedtowardsroleofgenderindeterminingutilizationofpreventivecareservicebuttheissuehasn'tbeenexploredusinganationallyrepresentativepopulation. METHODS: Aretrospective,cross-sectionalstudyusing2008MEPS (MedicalExpenditurePanelSurvey)data,anationallyrepresentativesurveyofUS population,whichreportstheirpatternofmedicalcareutilization.Guidelinerecommendationswidelyusedinclinicalpractice,suchasNCEP,AmericanDental Society,JNC-VIIetc.,wereusedtodetermineappropriateutilizationofpreventive careservices.Descriptivestatisticswereusedtodescribethepopulationcharacteristicswhilemultivariatelogisticregressionmodelwasbuilttopredicttheutilizationofthevariouspreventivecareservices(bloodpressurecheckup,lipid screening,dentalcheckup,sigmoidoscopy/colonoscopyanduvaccination),using genderastheprimarypredictorvariable,whilecontrollingforage,income,race/ ethnicityetc. RESULTS: Outof33,066respondents,20,336mettheinclusioncriteria forbloodpressurecheckup,23,058mettheinclusioncriteriafordentalcheckup, 19,543mettheinclusioncriteriaforuvaccinations,2,986mettheinclusioncriteriaforlipidscreening,and4,195mettheinclusioncriteriaforsigmoidoscopy/ colonoscopyscreenings.Genderwasfoundtopredictutilizationofpreventivecare services.Maleswerefoundlesslikelytoutilizebloodpressurecheckup(OR 0.327, CI 0.297 0.359),lipidscreening(OR 0.768,CI 0.694-0.85),dentalcheckup(OR 0.634,CI 0.599-0.671),anduvaccination(OR 0.680,CI 0.637-0.726).Incaseof sigmoidoscopyandcolonoscopyscreenings,thedisparitieswerenotsignicant, thoughutilizationwasstillfoundtobehigherinfemales. CONCLUSIONS: Thestudy helpeddeterminegenderdisparitiesinutilizingpreventivecareservicesinUS population.Utilizationofpreventivecareservicesneedstobeencouragedinmales. PCV102 COMPLICATIONSARISINGDURINGHOSPITALIZATIONFORHEMORRHAGICOR ISCHEMICSTROKE:EVIDENCEFROMALARGEADMINISTRATIVEDATABASECandrilliSD,KarveSJ,WirthPE RTIHealthSolutions,ResearchTrianglePark,NC,USAOBJECTIVES: Limiteddataexistregardingtheeconomicburdenofcomplications arisingduringstroke-relatedhospitalizations.Thisstudysoughttodocumentthe rateofcomplicationsarisingduringhospitalization(i.e.,denedasdiagnosesrecordedupondischargebutnotobservedatadmission)forhemorrhagic(HS)or ischemicstroke(IS),anddescribecharacteristicsofcomplicatedhospitalizations. METHODS: DataforhospitalizationswithaprimarydiagnosisofHS(ICD-9-CM codes430.xx,431.xx,or432.xx)orIS(433.x1,434.xx,or436.xx)inthe2008HCUP MichiganStateInpatientDatabasewereanalyzed.IncidenceofcomplicationsdevelopedduringhospitalizationamongpatientswithaprimarydiagnosisofHSorIS wereassessed,andresource-basedoutcomes(e.g.,totalcost,lengthofstay[LOS]) among,andothercharacteristicsof,stayswithcomplications,comparedtouncomplicatedhospitalizations,analyzed. RESULTS: Ofthe1.3millionhospitalizationsoccurringinMichiganin2008,19,065hadaprimarydiagnosisofHSorIS. Amongthese,20.6%(n 3,922)hadevidenceof 1complicationsarisingduringthe stay.Nodifferencesinpatientage(mean:70.4versus70.7years;p 0.3293)or genderdistribution(53%versus54%female;p 0.3476)betweencomplicatedand uncomplicatedhospitalizationswereobserved.Thetop-5mostfrequentlyobservedcomplicationswereurinarytractinfection,sitenotspecied;hypopotassemia;acuterespiratoryfailure;pneumonitisduetoinhalationoffoodorvomitus; andacutekidneyfailure,unspecied.Comparedtouncomplicatedhospitalizations,meanLOSandtotalcostsforcomplicatedstaysweresignicantlygreater: 10.5versus4.5days(p 0.0001)and$28,608versus$10,747(p 0.0001),respectively. Patientswith 1complicationsspent2.8(SD 6.4)daysinanICU,andagreater proportionwithcomplicationsthanwithoutdiedduringhospitalization(11.4% versus6.6%;p 0.0001). CONCLUSIONS: Thecostofstroke-relatedhospitalizations withcomplicationsissignicant, 3timesgreaterthanstroke-relatedhospitalizationswithoutcomplications.Effortstoimproveinpatientstrokemanagement strategiesmayhelplowertheincidenceofcomplications,reduceassociatedcosts, andimprovepatientoutcomes. PCV103 PHARMACISTINTERVENTIONSWITHINACOMMUNITYPHYSICIANBASED MEDICALHOMEPRACTICE:DIABETESCLINICALOUTCOMESWilleyVJ1,ReinholdJA1,KozuchJM1,AminSM1,KellyBL2,KimEA2,WilleyKH2 1UniversityoftheSciences,Philadelphia,PA,USA,2QualityFamilyPhysicians,Hockessin,DE, USAOBJECTIVES: Thepatient-centeredmedicalhomehasbeentoutedasawayto improvepatientcareandreduceoverallhealthcarecosts.Pharmacistsarequalied toprovidemanyoftheservicesthatarecoretothemedicalhomeconceptaspart ofthephysician-directedteam;however,thepharmacist'sroleinthemedical homehasreceivedlittleattention. METHODS: Medicalrecordreviewwasperformedonallpatientsreferredtothepharmacistfrom7/1/2009to12/1/2010within acommunity-based,medicalhome,primarycarepractice.Patientsreferredincludedthosenon-complianttopriorphysicianrecommendationsforlifestyle modicationsand/orthosenotachievingtherapeuticgoals.Pharmacistinterventionsincludeddiseasestateeducation,therapeuticlifestylemodicationandmedicationcounseling,andrecommendationsfortherapyoptimization.Primaryanalysesexaminedpre/postchangesinthesubsetofpatientswithdiabetes.Outcomes assessedwerechangesinhemoglobinA1c(HbA1c),lipidfractions,bodymassindex(BMI),weight,andgoalattainmentforHbA1candlow-densitylipoproteincholesterol(LDL-C),utilizingpairedt-tests,Wilcoxonsigned-rankandMcNemar's testsasappropriate. RESULTS: Onehundred-sevenpatientswerereferredtothe pharmacist,49withdiabetes.Diabetespatientshadameanageof57 9years;53% weremale.HbA1cvaluesdecreasedfrom8.7%to7.4%( 1.3%;95%CI 0.5%to 2.1%;p 0.003).ThepercentageofpatientsachievingHbA1clevelsbelow8%rose from50%to75%(p 0.021)andbelow7%rosefrom28%to47%,althoughthiswas notstatisticallysignicant(p 0.109).Statisticallysignicantdecreaseswereobservedindiastolicbloodpressure,LDL-C,totalcholesterol,triglycerides,BMIand weight.ThepercentageofpatientsachievingLDL-Clevels 100mg/dLincreased from30%to74%(p 0.002). CONCLUSIONS: Pharmacistinvolvementinthiscommunitybasedmedicalhomewasassociatedwithpositiveimprovementsinclinical markersforthesediabeticpatients.Thesepilotstudyresultssupporttheinclusion ofpharmacistsashealthcareteammembersinfuturemedicalhomedemonstrationprojects. PCV104 REAL-WORLDSIDEEFFECTDATAONCHOLESTEROLMEDICATIONS–OUTPUTS FROMANONLINEPATIENTCOMMUNITYCascadeE1,PlummerRC2,ManginiNS2 1MediGuard.org,Rockville,MD,USA,2MediGuard.org,Princeton,NJ,USAOBJECTIVES: Tocomparesideeffectdatareportedfrommembersofapatientregistrytoinformationcontainedintheproductlabelingoffourdifferentcholesterolmedicationseachrepresentingadifferenttherapeuticclass. METHODS: A randomsampleofU.S.MediGuard.orgmemberswhoreportedtakingniacin(Niaspan),fenobrate(Tricor),simvastatin(Zocor),orcolesevelam(Welchol),wereinvitedtocompleteavalidatedonlinetreatmentsatisfactionsurveythatincludes questionsrelatedtosideeffects.MediGuard.orgisafreemedicationmonitoring servicethatprovidesinformationtoover2.5millionmembersintheUS,UK, France,Germany,Spain,andAustralia.Acomparisonlistofadverseeffectsand frequenciesreportedduringclinicaltrialswasextractedfromthebrandedpackage insertforeachmedication. RESULTS: Feedbackwasobtainedfrom56colesvelam, 108niacin,216fenobrate,and660simvastatinpatients.Niacinpatientshadthe highestprevalenceofsideeffects(62%),primarilyushing(55%)andpruritis(14%). 23%ofcolesevelammembersreportedsideeffects:themostcommonwereconstipation(14%)andbloating/gas(2%).Forpatientstreatedwithfenobrate,16% reportedsideeffectswithmyalgia(4.2%)andarthralgia(1.4%)beingthemostcommon.Similartofenobrate,16%ofsimvastatinpatientsreportedsideeffectsand again,myalgia(6.2%)andfatigue(2.1%)hadthemostmentions.Finally,2%of niacin,22%offenobrate,29%ofsimvastatin,and36%ofcolesevelampatient reportsincludedsideeffectsnotincludedintheproductlabeling. CONCLUSIONS: On-linepatientcommunitiesareanemergingresourceforconrmingadverse eventsreportedduringclinicaltrialsandforcapturingpreviouslyundocumented signals.Intandemtothecurrentpost-marketingspontaneousadverseeventreportingsystem,longitudinalpatientregistriescanprovideinsightnotonlyonthe numberofadverseevents,butalsoaprevalencerateofthosewhoexperience versusthosewhodonotexperiencesideeffects. PCV105 THEINITIATIONANDEXTENTOFDOSETITRATIONOFACEINHIBITORSAND -BLOCKERSPOSTACUTEMYOCARDIALINFARCTION:APROSPECTIVEAUDITAlowayeshMS1,WrightP2 1VirginiaCommonwealthUniversity,Richmond,VA,USA,2TheLondonChestHospital,London,UKOBJECTIVES: Theobjectivesofthisauditare:(1)todeterminethepercentageof patientswhoaredischargedonsecondarypreventionmedicationfollowingacute myocardialinfarction(AMI)including:aspirin,clopidogrel,ACEinhibitors(ACEi),-blockers(BB),andstatins;(2)toidentifywhatdoseeachpatientisdischargedon regardingACEiandBBandwhentheyareinitiated;(3)toexploretherelationship betweenbloodpressureandACEidosetitration;(4)toexploretherelationship betweenheartrateandBBdosetitration. METHODS: AprospectiveauditwascarriedoutattheLondonChestHospital(LCH)fromJune15-June28,2009.Allpatients whowereadmittedtothecoronarycareunit(CCU)withanaldiagnosisofAMI wereincluded.Patientswereexcludediftheydiedpriortohospitaldischarge. Patients'demographics,vitalsigns,drughistory,pastmedicalhistory,drugsduringhospitalstayandatdischargewerecollected. RESULTS: 33patientswereincludedinthisaudit(meanage59.7 12.7years,79%males,21%females)withan averagelengthofstayof2.2days.88%ofthepatientswerestartedACEiandBBon day2ofhospitalization.ForpatientsreceivingACEiandBBonly41%weretitrated towardstheoptimaldose.78%oftheopportunitiestotitrateACEiaccordingto bloodpressureand55%oftheopportunitiestotitrateBBaccordingtoheartrate werenottaken.Atdischarge,100%wereprescribedaspirin,statins,andBB;while 97%wereprescribedACEiandclopidogrel. CONCLUSIONS: Thisauditrevealshigh useofsecondarypreventionmedicationattheLCHfollowingAMI.Althoughthere areopportunitiesforfurtherdosetitrationpriortodischarge,furtherworkisrequiredtoestablishreasonsformissingtheseopportunities.A51VALUEINHEALTH14(2011)A1–A214

PAGE 60

PCV106 RE-HOSPITALIZATIONRATESOFACUTECORONARYSYNDROMEPATIENTSIN REALWORLDCLINICALPRACTICE:OBSERVATIONSFROMANATIONAL ADMINISTRATIVECLAIMSDATATunceliO1,GandhiSK2,BhandaryD2,StephensonJJ1,GoldA2,FuAC1,KernDM1, SingerJ1 1HealthCore,Inc.,Wilmington,DE,USA,2AstraZenecaPharmaceuticalsLP,Wilmington,DE,USAOBJECTIVES: Re-hospitalizationandmortalityratesareincreasinglybeingusedas qualityofcaremeasuresthathavesignicantreimbursementimplications.We examinetheratesofre-hospitalizationandmortalityofacutecoronarysyndrome (ACS)patientsinreal-worldclinicalpractice. METHODS: Commercially-insured patients(age 18years)withaninpatienthospitalizationforACS[ICD-9-CMcodes foracutemyocardialinfarctionorunstableangina(UA)]between1/1/2007-5/31/ 2010wereidentiedfrommedicalclaimsintheHealthCoreIntegratedResearch Database(HIRD).PatientswithACSeventswithinoneyearpriortoindexhospitalizationwereexcluded.All-causeandACS-relatedre-hospitalizationsandmortality rateswithin30daysand12monthsafterindexeventwereevaluated. RESULTS: Of 66,772ACSpatients(60%male;meanage66.6years),21%haddiagnosticcodingfor STelevationmyocardialinfarction(STEMI),31%hadcodingfornon-STelevation myocardialinfarction(NSTEMI),37%hadUA,and11%hadnototherwisespecied (NOS)ACS.Approximately,90%and52%ofpatientshad 30-daysand 12months ofcontinuouseligibilityrespectively.The30-dayall-causere-hospitalizationrate was16.3%(STEMI:16.4%,NSTEMI:19.0%,UA:13.3%,NOS:20.6%;)and6.3%(STEMI: 8.8%,NSTEMI:6.6%,UA:5.2%,NOS:4.5%)foranACS-relatedre-hospitalization.The 12-monthallcausere-hospitalizationratewas41.3%(STEMI:39.0%,NSTEMI: 46.4%,UA:38.2%NOS:46.6%,)and16.6%foranACS-relatedre-hospitalization.The 30-daypost-indexmortalityratewas2.4%(STEMI:1.8%,NSTEMI:4.3%,UA:0.5%, NOS:5.2%;)andthe12-monthratewas7.0%.Forpatientswithages 65years,the 30-dayall-causeandACS-relatedre-hospitalizationrateswere21.2%and7.0%, respectively. CONCLUSIONS: There-hospitalizationandmortalityrateforACS patientswithin30daysand12-monthspost-indexhospitalizationdischargeas observedinreal-worldclinicalpracticeremainhigh.UseofmoreeffectivetherapiesmayprovideanopportunitytoimproveimportantclinicalandeconomicoutcomesinACSpatients. PCV107 IDENTIFYINGEFFICIENTACUTECLINICALPATHWAYSFORCHESTPAIN:USING RISKADJUSTEDCOST-EFFECTIVENESS(RAC-E)TOCOMPAREHOSPITALSUSING LINKED,ROUTINELYCOLLECTEDDATACaffreyO1,PhamC2,KarnonJ2,Ben-TovimDI3,HakendorfP3,CrottyM4 1UniversityofBirmingham,Birmingham,Warwickshire,UK,2UniversityofAdelaide,Adelaide, SouthAustralia,Australia,3FlindersMedicalCentre,Adelaide,SouthAustralia,Australia,4FlindersUniversity,Adelaide,SouthAustralia,AustraliaOBJECTIVES: Cost-effectivenessanalysisisnowwellestablishedinrelationtothe evaluationofhealthtechnologies,butnotinrelationtobroaderinstitutionallevel variationsinclinicalpathways.Linked,routinelycollecteddataprovidesanopportunitytoevaluaterealworldactivity.Anovelapproachtoanalysingtherisk-adjustedcost-effectiveness(RAC-E)ofacuteservicesforpatientspresentingwith chestpainatthefourmainpublichospitalsinSouthAustraliaispresented. METHODS: Routinelycollecteddataonhospitalseparations(includingcosts)and mortalityrecordswerelinkeddeterministically.Relevantintermediateendpoints overatwoyearfollow-upperiodinacohortofpatientspresentingwithchestpain intheyeartoJuly2006wereidentied.Lifetimecostsandsurvivalwereextrapolatedfromtheseendpointsusingdatafromchestpainpatientspresentingbetween July2002andJune2008.Theresultingestimatesofcostsandsurvivalwerestandardizedusingseparateregressionmodelsthatestimatedexpectedcostandsurvivalvaluesforeachpatient. RESULTS: Inthebasecase,twoofthefourhospitals weredominatedbyhospital1.Hospital2hadlowerstandardizedlifetimecosts thanhospital1,andtheincrementalcostperlifeyeargainedbetweenthesetwo hospitalswasAus$2,909.Abootstrappedprobabilisticsensitivityanalysisshowed hospital1tohaveveryhighprobabilitiesofcost-effectivenessatrelevantdollar valuesforalifeyeargained.Analysisofdifferencesincostcomponentsbetween thehospitalsshowedthathospital1spentrelativelylessonpathologyandimaging,whilstspendingmoreonnursingtimeandpharmaceuticals. CONCLUSIONS: RAC-Eprovidesausefulframeworkforidentifyingimportantdifferencesinthe costsandbenetsassociatedwithvariationsinclinicalpractice.Potentialdeterminantscanbepartiallyinvestigatedwiththedata,butfurtherprimaryanalysisof clinicalpathwaysatkeyhospitalsisrequiredtofullyinformthedisseminationof bestpractice. PCV108 CARDIOVERSIONTREATMENTPATTERNSANDOUTCOMESAMONGACUTE ATRIALFIBRILLATIONPATIENTSIN5EUROPEANCOUNTRIESBuonoJL1,RussellT2,WorsfoldA2,BashLD3 1St.John'sUniversity,Queens,NY,USA,2AdelphiRealWorld,Maccleseld,UK,3Merck&Co., Inc.,WhitehouseStation,NJ,USAOBJECTIVES: Weexaminecurrenttreatmentpatternsassociatedwiththeuseof antiarrhythmicsforpharmacologiccardioversionandevaluatetimetoconversion amongpatientswithacuteatrialbrillation(AF)inFrance,Italy,Germany,Spain, andUK. METHODS: Datawerecollectedcross-sectionallyasobservedby303physiciansfromApriltoJune2010. RESULTS: Among2,997patients,1,352(45%)receivedpharmacologiccardioversiononly,with1,082(80%)successfullycardioverting.Ofthese,931(69%ofthosetreatedpharmacologically,86%ofthose successfully)hadcompleteinformationontimetoachievesinusrhythmandwere furtherassessedhere.Amiodaronewasadministeredtomost(49%),withsimilar proportionsreceivingintravenous(IV)(53%)andoral(46%)formulations.AmiodaroneIVwasassociatedwithasignicantlyshortermedianconversiontime(8 hours)comparedtooral(36hours).Patientstreatedwithecainide(26%)hadthe shortestmediantimetocardioversion(4hours),whilepatientsonpropafenone (15%)werenext(6hours).Mediancardioversiontimevariedbycountry.Specically,patientsinFrancetooklongertoconvertonamiodaroneIVandoralata mediantimeof12and48hourscomparedtoanaverageof7and24hours,respectively,inothercountries.InGermany,littledifferencewasobservedwithmedian conversiontimesof9and10hoursforamiodaroneIVandoral,respectively.MedianconversiontimeswerealsosimilarbetweenamiodaroneIVandecainidein SpainandItaly(adifferenceof1and2hours,respectively)comparedtoanaverage differenceinmedianconversiontimeof4hoursbetweenthesetreatmentsoverall. CONCLUSIONS: Whileamiodaronehadalongermediantimetocardioversion, fasteractingagents,suchasecainideandpropafenone,hadconversiontimes substantiallylongerthantheirknowntimeofaction,reectingotherinuenceson treatmentadministration.Differencesinconversiontimesuggestcountry-and physician-specicpracticesintheuseofpharmacologiccardioversiontherapies. CardiovascularDisorders–ResearchonMethods PCV109 BEYONDCASEFATALITY:ANEWMETHODTOESTIMATETHEEFFECTOF INCREASINGTREATMENTUPTAKEONMORTALITYMitsakakisN1,WijeysunderaHC2,KrahnM1 1TorontoHealthEconomicsandTechnologyAssessment(THETA)Collaborative,Toronto,ON, Canada,2SunnybrookHealthSciencesCentre,Toronto,ON,CanadaOBJECTIVES: Epidemiologicalmodelshavebeenwidelyusedtoestimatehowincreaseduptakesofmedicalandsurgicaltreatmentsaffectmortalityandrelated outcomes.Standardmethodsrelyontheestimateofthecasefatality,denedas theriskofdeathintheabsenceofthetreatment.Becausemostpatientsreceive sometreatment,mortalityrateswheresometreatmentispresentareoftenused insteadofcasefatalityrates,leadingtobiasedresults.Amethodthatdoesnotrely oncasefatalityestimatesisneeded. METHODS: Weborrowthemechanismused forthecalculationofthePotentImpactFraction(PIF),anepidemiologicalmeasure thatisequaltotheproportionalreductionintheincidenceofadiseaseormortality, resultingfromaspecicchangeinthedistributionofariskfactorinthepopulation, andapplyittotheestimationofthe relative reductionofmortalitycausedbythe increaseoftreatmentuptakeinthepopulationatrisk.Weapplythismethodto estimatethereductionofcardiovasculardiseasedeathsinOntario,iftreatment ratesforCHDinterventionsweretobeincreasedfrom2005levelstotherecommendedbenchmarkutilizationof90%.TheMant-Hicksmodelforpolypharmacyis adopted,whiletheuptakesofmultipletreatmentsareassumedtobeindependent fromeachother. RESULTS: UsingtheproposedPIF-basedmethod,weestimated thatincreasingtreatmenttobenchmarklevelsuptakeresultsinareductionof cardiovascularmortalityof22.5%.Thestandardmethodgivesareductionof17%, probablyduetounderestimationofthecasefatality. CONCLUSIONS: Herewepresentanalternativemethodfortheestimationoftheeffectoftreatmentuptake increasetothereductionofmortality.Ourexamplesuggeststhatthemagnitudeof biasassociatedwiththestandardmethodmaybesubstantial.Thisapproachmay beausefultoolforepidemiologicalandhealthcareresearch. PCV110 DEVELOPMENTANDVALIDATIONOFASHORTPROMEASUREOFHEALTH STATUSFORINDIVIDUALSWITHACUTEMYOCARDIALINFARCTION:THE MYOCARDIALINFARCTIONDIMENSIONALASSESSMENTSCALE(MIDAS)JenkinsonC1,ThompsonD2,RoebuckA3,ChurchmanD4 1UniversityofOxford,HealthServicesResearchUnit,Oxford,Oxfordshire,UK,2Australian CatholicUniversity,Melbourne,Victoria,Australia,3UnitedLincolnshireNHSTrust,Lincoln, Lincolnshire,UK,4IsisInnovationLtd.,Oxford,Oxfordshire,UKOBJECTIVES: Todevelopandvalidateadisease-specichealthstatusmeasurefor individualswithmyocardialinfarction(MI). METHODS: Thedevelopmentofthe MyocardialInfarctionDimensionalAssessmentScale(MIDAS)followedthreemain stages.Stage1consistedofin-depth,semi-structured,exploratoryinterviewsconductedonasampleof31patientstoidentifyareasofsalienceandconcernto patientswithMI.Theseinterviewsgenerated48candidatequestions.Instage2the 48-itemquestionnairewasusedinapostalsurveytoidentifyappropriaterephrasing/shortening,todetermineacceptabilityandtohelpidentifysub-scalesofthe instrumentaddressingdifferentdimensionsofMI.Finally,instage3theconstruct validityofMIDASsubscaleswasexaminedinrelationtoclinicalandotherhealth outcomes.Setting-Asinglecentre(districtgeneralhospital)inEnglandwasused forstages1and3andanational postalsurveywasconductedforstage2.Patients-A totalof410patientswererecruitedforthenationalsurvey(stage2).Fulldatawas availableon348(85%)patients.155patientswererecruitedtotestconstructvalidity(stage3). RESULTS: TheMIDAScontains35questionsmeasuringsevenareasof healthstatus:physicalactivity,insecurity,emotionalreaction,dependency,diet, concernsovermedicationandsideeffects.Themeasurehashighface,internaland constructvalidityandislikelytoproveusefulintheevaluationoftreatmentregimesforMI. CONCLUSIONS: TheMIDAShasacceptablevalidityandreliability.It issuitableforuseinavarietyofsettingsforpatientswithmyocardialinfarction. PCV111 IDENTIFICATIONOFRESPONSESHIFTAMONGHYPERTENSIVEPATIENTSWITH CORONARYARTERYDISEASEUSINGTWOSTRUCTURALEQUATIONMODELING TECHNIQUESGandhiPK1,RiedLD2,HuangIC1,KimberlinC1,KaufT1,SuhDC3 1UniversityofFlorida,Gainesville,FL,USA,2SouthwesternOklahomaStateUniversity, Weatherford,OK,USA,3SchoolofPharmacy,RutgersUniversity,Piscataway,NJ,USAA52VALUEINHEALTH14(2011)A1–A214

PAGE 61

OBJECTIVES: Toidentifyresponseshiftusingtwostructuralequationmodeling (SEM)techniqueswiththeSF-36HealthSurvey. METHODS: Hypertensivepatients withcoronaryarterydisease(CAD)whocompletedbothbaselineandoneyear follow-upoftheSF-36wereincluded(n 909).Anoccurrenceofresponseshift usingOortandSchmittSEMtechniqueswasspecicallyidentied.LISRELsoftware wasusedtoconductSEMprocedures.Avarietyoftindiceswereusedtodeterminemodelt.ForbothSEMapproaches,responseshiftisdenedbasedon changesinvariousparametersinthemeasurementmodel.Effectsizeindiceswere calculatedforthecontributionofresponseshiftonthechangeofSF-36domains scores.Wehypothesizedthedivergenceindeningtypeofresponseshiftlinkedto changesinvariousparameterswillleadtodifferentndings. RESULTS: Onlythe SF-36physicalfunctioning(PF)scalewasidentiedwithrecalibrationresponse shiftbyusingbothOortandSchmittSEMapproaches.WithOortapproach,recalibrationwasidentiedbythechangeinintercepts,whereasSchmittapproachdenesrecalibrationasthechangeinfactorvariancesorfactorloadingsovertime. EffectsizeoftherecalibrationresponseshiftonthechangeofPFdomainscores wasmarginal: 0.118. CONCLUSIONS: Thisistherststudytoidentifyresponse shiftinhypertensiveCADpatientsusingSEMapproach.Recalibrationresponse shiftwasidentiedusingbothOortandSchmittSEMapproaches.DifferentinterpretationofspecicPFitemsbyhypertensiveCADpatientsbeforeandaftertreatmentsmaycontributetotherecalibrationresponseshift.Bylookingmoreclosely attheSF-36PFdomainscoresamonghypertensiveCADpatientswillenableusto providenuancedattentionanddirecttreatmentforthemostimpairedaspectsof qualityoflife. PCV112 EXTENSIONOFMETA-ANALYSISINCOMPARINGOFFIMASARTANWITH LOSARTANINBLOODPRESSURELOWERINGEFFECTNaY,LeeEK SookmyungWomen'sUniversity,Seoul,SouthKoreaOBJECTIVES: Anewdrugmasartanhadbeendevelopedandcurrentlyapproved inKoreafortreatingessentialhypertension.Thisstudyaimedtoidentifywhether mainresultsofdirectcomparativestudymaintainsconsistencywiththoseofextensionofmeta-analysisinthebloodpressureloweringeffectofmasartanwith losartan. METHODS: Systematicreviewsofliteraturesofclinicaltrialsincluding masartanorlosartanwereconducted.Thebloodpressurechangefrombaseline wereusedasaneffectivenessmeasureandwaspooledinRevMan4.0.Fordirect comparison,thehead-to-headrandomizedcontrolledtrial(RCT)ofmasartanand losartanwasused.Forindirectcomparison,itfollowedtomethodofadjusted indirectcomparison(Bucher1997)usingcommoncomparatorandusedITC(IndirectTreatmentComparison)program(CADTH).Inaddition,Bayesianmixedtreatmentcomparison(MTC)wasperformedonwhichcombineswholepairwisecomparisonstudiestogetherbyWinBugsprogram.Afterthat,theresultswere comparedwiththatofadirectcomparison. RESULTS: Inregardtodirectcomparison,thereistheonlyhead-to-headtrial(PhaseIII)reportofcomparingmasartan withlosartan,whichhadconductedlocally.Forindirectcomparison,thesearch identiedonereportofthetrialofcomparingmasartanwithplacebo(PhaseIIb), andthe6ofarticlesofcomparinglosartanwithplacebowereselectedandmeasuredestimateswerepooled.Thechangeinreductionindiastolicbloodpressureof losartanandmasartanversusplacebowas 2.7mmHgasdirectestimateretrievedfromRCT,and 3.315mmHgascombinedestimatethroughMTCanalysis, respectively.InSBPreduction,itwas 4.3mmHgand 3.995mmHgrespectively. CONCLUSIONS: Thisstudyfoundthatestimatesobtainedfromindirectcomparisonweresimilarwiththoseofdirect.Theresultsfromthetwocomparisonmethodsbothindicatedconsistentlythattheantihypertensiveeffectofmasartanis betterthanlosartan. SensorySystemsDisorders–ClinicalOutcomesStudies PSS1 STEVENS-JOHNSONANDREDMANSYNDROME:ACASEREPORTONADVERSE DRUGREACTIONSOFSIMULTANEOUSUSEOFPHENYTOINANDVANCOMYCINKlerSK DepartmentofGeneralMedicine,GovernmentMedicalCollege&Hospital,Chandigarh,IndiaOBJECTIVES: Adversedrugreactions(ADRs)areacommoncauseofinjurytohospitalizedpatientsandarelikelypreventable.Monitoringspontaneousadversedrug reactionsisoneoftheepidemiologicalmethodsforassessingthesafetyofdrugsin ahospitalsetting.ManystudiesreportedthatcutaneousreactionsandgastrointestinaldisturbancesastopmostreportedADRs. METHODS: Theauthorreportsa caseof22-year-oldgirlwhodevelopedStevens-JohnsonsyndromeandRedMan syndromeafterreceivingphenytoinandvancomycinsimultaneously. RESULTS: Thiswas22yearoldgirlpresentedwithchiefcomplaintsofheadache,nausea, vomitingandsweating.Diagnosedasacaseofpostfossatumorandoperatedon3rddayafteradmission.Shewastreatedwithdifferentmedicationsalongwithvancomycin1gm12hourly,phenytoin100mg8hourlyanddexamethasone2gm8 hourly.Ontheday15ofvancomycinandphenytointreatment,thepatientdevelopeddiffusepolymorphicerythemawithmouthlesionsandtypicalpalmo-plantar lesionscharacteristicofStevens-Johnsonsyndrome.Alsoaskinrashandlesions predominantlyontheface,neck,upperbackandfrontregionscharacterizedasRed Mansyndrome.Useofvancomycinandphenytoinwasdiscontinuedandreplaced bycepimepluslinizolidandphenobarbitonerespectively.Thereactionpersisted forapproximately55dayswithprogressiveimprovementandthepatientwas treatedwithhydroxyzine30mg,6hourly,hydrocortisone100mg,6hourlywith gradualtaperingofthedoseandtopicaltreatmentofthelesions. CONCLUSIONS: Althoughtherearereportsontheofadversedrugreactionwithmonotherapyof phenytoinandvancomycinorininvolvementwithothermedications.However, therearenoreportsonthedevelopmentofStevens-JohnsonandRedmansyndromewhenphenytoinandvancomycinwereusedsimultaneouslywithother drugtreatment.ItalsoconrmsthepossibleincreasedriskofdevelopingStevensJohnsonsyndromewhenphenytoinisassociatedtocorticosteroids. PSS2 DIAGNOSEDANDUNDIAGNOSEDDRYEYE,SYMPTOMSEVERITY,AND ASSOCIATEDFACTORSAMONGMENANDWOMENINTHEUNITEDSTATESSchaumbergDA1,LiJZ2 1BrighamandWomen'sHospital,Boston,MA,USA,2Pzer,Inc.,SanDiego,CA,USAOBJECTIVES: Toexaminefactorsassociatedwithdryeyedisease(DED)intheUS. METHODS: Weconductedacross-sectionalsurveyof4000participantsinthe Women'sHealthStudyandPhysicians'HealthStudieswithdiagnosedDEDor severesymptoms.WeassessedthecurrentlevelofsymptomsbytheOcularSurfaceDiseaseIndex(OSDI)andSymptomAssessmentinDryEye(SANDE)questionnaires,diagnosis,co-morbidities,treatments,andpatientsatisfaction. RESULTS: 3390(84.8%)subjectsreturnedquestionnaires.2099participantsreportedadiagnosisofDED,and1291deniedDEDdiagnosis(73.9%ofthesehadreportedDED diagnosispreviously).Among451subjectsselectedbasedonseveresymptoms alone,114(25.3%)reportedanewdiagnosisofDED,whichwasmorestrongly associatedwithseveresymptomsbySANDE 40(OR 2.24,p 0.001),thanbyOSDI 33-100(OR 1.38,p 0.25).Blepharitis(OR 2.03,p 0.05)wasalsoassociatedwith newDEDdiagnoses.AmongthosewhocurrentlydeniedDEDdiagnosis,15.9%had severe(SANDE 40)and40.4%hadmild-moderatesymptoms(SANDE15-40).Adjustingforageandsex,participantswithsymptomsonlywerelesslikelythan diagnosedpatientstohaveaneyeexam 1x/year(OR 0.71,p 0.002),useantidepressants(OR 0.76,p 0.04),articialtears(OR 0.67,p 0.0002),orotherDED treatments(OR 0.65,p 0.0001);andmorelikelytoreportrefractivesurgery (OR 1.67,p 0.02),andcontactlenswear(OR 2.51,p 0.0001).Inage-andsexadjustedmodelsincludingallrespondents,thosewhocurrentlyreportedDEDdiagnosesweremorelikelytohaveaneyeexam 1x/year(OR 1.42,p 0.0003), severesymptoms(SANDE 40,OR 2.00,p 0.0001),blepharitis(OR 1.41, p 0.007),useantidepressants(OR 1.43,p 0.003),articialtears(OR 2.01, p 0.0001),orotherDEDtreatments(OR 1.70,p 0.0001);andlesslikelytouse glaucomamedications(OR 0.82,p 0.02). CONCLUSIONS: Theseobservationssuggestthepossibilityofunder-diagnosisofDED,andarealsoconsistentwithamilder and/ormoreintermittenttypeofDED.IndividualswithdiagnosedDEDaremore likelytohaveseveresymptoms,despitetherapy. PSS3 PERSISTENCEWITHSTATINSANDTHERISKOFAGERELATEDMACULAR DEGENERATIONINALARGEHEALTHORGANIZATIONINISRAELChodickG,ShalevV,GoldsteinI,SrorM MaccabiHealthcareServices,TelAviv,IsraelOBJECTIVES: Toinvestigatetheassociationbetweenpersistentuseofstatinsand theriskofage-relatedmaculardegeneration(AMD). METHODS: Apopulationbasedretrospectivecohortamongadultswhobeganstatintherapybetween1998 and2006inalargehealthorganizationinIsrael.Theorganization'scentralcomputerizeddatabaseswereusedtocollectsataonincidentAMDcasesdiagnosedby ophthalmologists. RESULTS: Atotalof108,973individualsaged55orolderwere identied.Duringthestudyfollow-upperiod409,113person-years,therewere 2,732incidentAMDcases(6.68per1,000person-years).ThecrudeincidencedensityrateofAMDamongpatientsatthelowestquintileofpersistencewithstatins (7.18per1,000)wascomparabletothatofhighestpersistencequintile(7.13per 1,000).Afteradjustmentforpotentialconfounders,patientsinthehighestquintile ofpersistencewithstatinshadahazardratioof0.99(95%CI:0.78-1.26)forAMD comparedwithpatientsinthelowestPDCquintile.Inadditiontoage,AMDwas foundtoassociatewithpastsmoking,asthma,diabetesandfrequentvisitsto ophthalmologistsorprimaryphysicianspriortoindexdate. CONCLUSIONS: Our studyagreeswithpreviousstudiesthatshowednoassociationbetweenpersistent useofstatinsandreducedriskofAMD.Theseresultssuggestthattheearlyreports onastrongprotectiveeffectofstatinsagainstAMDdevelopment,wereprobablya resultofasmallstudyeffect. PSS4 SYSTEMATICREVIEWOFTHEEPIDEMIOLOGICLITERATUREONATOPIC DERMATITISINCHILDRENBerbariJ1,ReynoldsS1,Heil-RuessM2,IskedjianM3 1PharmIdeasResearchandConsulting,Ottawa,ON,Canada,2MHR-NutritionMarketing Services,Lausanne,Switzerland,3PharmIdeasResearchandConsulting,Oakville,ON,CanadaOBJECTIVES: Asystematicreviewoftheliteraturewasperformedtogatherthe epidemiologicalevidencerelatedtoatopicdermatitis(AD)inapediatric population. METHODS: OVIDMEDLINE¨wassearchedusingtermsrelatedtoAD, epidemiology,incidence/prevalenceinapopulation 18yearsold.Tworesearchersundertooktheinclusion/exclusionprocessonthe466citationsthathadbeen identied.Athirdpersonactedasanoverallreviewerandadjudicator. RESULTS: ThemajorndingwastheInternationalStudyofAsthmaandAllergiesinChildhood(ISAAC)whichreportedtheprevalenceofADinchildren6-7yearsoldfrom35 countriesattwodistincttimeperiods(around1995and2002).Afurther32independentstudieswereidentiedforinclusioninthisreview.Thesestudiesreported onincidence(n 2),prevalence(n 25)orboth(n 5)inEurope(n 23),Southeast Asia(n 8)andAfrica(n 1).Thenumberofstudyparticipantsdifferedgreatly (n 137ton 317,926).AccordingtoISAAC,theworldwideprevalencerateofAD increasedbyarateof36%from12.9%in1995to17.5%in2002.Overthatsametime period,a46%increasewasreportedforNorthAmericaandWesternEuropewithA53VALUEINHEALTH14(2011)A1–A214

PAGE 62

thelargestincrease(95%)observedinLatinAmerica.By2002,many(n 9)countries hadreachedabsoluteprevalencerates 30%.ThehighprevalenceofADasreportedbyISAACwasconrmedintheremainingliterature,especiallyinapopulationofyoungchildrenandinfants,withratesofupto60%.Thereportedincidenceratesforasimilarpopulationwereashighasapproximately10%. CONCLUSIONS: ADisoneofthemostprevalentdiseaseconditionsinchildhood, withanincreasingtrendinmostpartsoftheworld. PSS5 WHATINFLUENCESDENTALCAREAMONGCHILDREN?HufstaderM1,VaidyaV2,White-MeansS3,DishmonB4,SudharshanL5,SharmaG5 1Covance,gaithersburg,MD,USA,2UniversityofToledo,Toledo,OH,USA,3Universityof Tennessee,Memphis,TN,USA,4ParadigmDental,Memphis,TN,USA,5CovanceMarketAccess Services,Inc.,Gaithersburg,MD,USAOBJECTIVES: Theobjectivewastoexaminethefactorsthatinuencedentalcare, dentalexpenditures,anddentalinsuranceamongchildren(2-17)intheUS.Dental cariesarethemostprevalentchildhoodchronicdisease.Lackofdentalcarein childhoodcanleadtopoorhealthoutcomesanddecreasedqualityoflife. METHODS: TheMedicalExpenditurePanelSurveydata(2008),descriptivestatistics,linearregression,andlogisticregressionwereusedtoperformacrosssectionalretrospectiveanalysis.Covariatesincludedrace,residence,insurancestatus, andfamilyincome. RESULTS: Meandentalexpendituresperchildin2008was $1,246.Half(52%)ofchildrenhaddentalinsurance,andhalf(54%)usedpreventativedentalservices.Childrenthatutilizedpreventativeserviceshadannualmean dentalexpenditures($559)thatweresignicantlylowerthanchildrenthatdidnot utilizepreventativeservices($2,092).Logisticregressionanalysisindicatedthat childrenwithoutcariesweremorelikely(OR:2.43[2.42-2.44])tohavepreventative carethanchildrenwithcaries.Higherincomeanddentalinsurancewerepositive predictorsofpreventativedentalcare.Thosewithdentalinsurancehadhigher meanannualexpenditures($1,579)comparedwiththoseuninsured($975),indicatingthatdentalinsuranceincreasesdentalresourceuse.Predictingfactorsfor dentalinsurancewerelackofcaries,preventativeserviceutilization,higherincome,residenceintheNortheastorMidwest,Asiandescent,andwhiterace. Linear regressionshowedthatdentalexpenditureswereloweramongthosewithout dentalinsurance,withpreventativeservices,residinginanon-metropolitan statisticalarea,blackrace,Hispanicethnicity,andresidingoutsidethe northeast. CONCLUSIONS: Thedataindicatethatdentalinsuranceisnottheonly barriertochildren'sdentalcare.Certainracialandethnicgroupsandlowerfamilial incomearenegativepredictorsfordentalcare.Preventativeserviceslowerdental expendituresandpreventcaries.Thereforethereturnoninvestmentforpreventativeservicesispositive.Healthreformshouldenactpoliciesencouragingpreventativedentalservices. PSS6 SAFETYPROFILEOFLATANOPROSTVERSUSTIMOLOLINPATIENTSWITH OPENANGLEGLAUCOMAOROCULARHYPERTENSION:POST2000SETTINGGuptaJ,SiddiquiMK,JindalR,SehgalM HeronHealthPrivateLtd,Chandigarh,Chandigarh,IndiaOBJECTIVES: Tosystematicallycomparethesafetyoflatanoprostversustimolol throughmeta-analysisofrandomizedcontrolledtrials(RCTs)inopenangleglaucoma(OAG)orocularhypertension(OH). METHODS: MedlineandCochraneCentralRegisterofControlledTrials(CCTR)weresearchedforRCTspublishedpost 2000,assessingheadtoheadcomparisonoflatanoprostvs.timololinOAGorOH. Studieswereassessedforinclusion/exclusionbasedonaprespeciedprotocol. Tworeviewersundertookdataextractionindependently.Anydisagreementwas resolvedbyathirdreviewer.Acustomizedspreadsheetwasusedtoextractthe relevantoutcomes,includingthestudyduration,samplesize,numberofpatients withlocalandsystemicsideeffectsandtheproportionofwithdrawalsdueto adverseevents. RESULTS: FourRCTsenrolling1688patientsmettheinclusion criteriaoutofthe289studiesidentied.Latanoprostwasassociatedwithanincreasedincidenceofhyperemia(riskratio(RR) 2.7,95%CI 1.37to5.28),iris pigmentation(RR 10.81,95%CI 1.37to85.47)andpruritus(RR 3.05,95%CI 1.06to8.73)comparedtotimolol(LevelIevidence).Also,oneRCT(LevelIIevidence) reportedthatlatanoprostwasassociatedwithincreasedincidencesofeyelash changes(riskdifference(RD) 0.23,95%CI 0.16to0.29;numberneededtoharm (NNH) 4,95%CI 3to6).OtherAEscommonlyobservedwithboththetreatments wereblepharitis,cataract,dryeye,eyeirritation,eyepainandkeratitis.Therewas nostatisticallysignicantdifferenceforwithdrawalsduetoAEsacrossboththe treatments(RD 0.02,95%CI 0.02to0.05). CONCLUSIONS: Timololdisplayeda bettersafetyprolethanlatanoprostforadverseeventsincludinghyperemia,iris pigmentationandpruritus.However,longtermsafetycouldnotbeassessedasfew includedstudiesevaluatedpatientsbeyondsixmonths. SensorySystemsDisorders–CostStudies PSS7 MEDICALCARECOSTSASSOCIATEDWITHVISIONLOSSINTAIWANWangMT1,NgK2,LoYW1,LeeWJ1,WuBJ1,YehWS3 1NationalDefenseMedicalCenter,Taipei,Taiwan,2AllerganSingaporePteLtd,Singapore, Singapore,3Allergan,Irvine,CA,USAOBJECTIVES: Toassesstheeconomicburdenfromvisionlossonthehealthcare systeminTaiwan. METHODS: Thisisaretrospectivematchedcohortstudy,with studypopulationretrievedfromtheNationalHealthInsuranceResearchDatabase between01/01/2000and12/31/2008.BasedontheICD-9-CMcodes,casesinitially diagnosedwithmoderatevisionloss(369.6-369.9),severevisionloss(369.1-369.4), orblindness(369.0-369.08)wereidentied.Eachcasewasmatchedtoarandomly selectedcontrolwithnormalvisionbyage( 1year)andgender.Eachcontrolwas assignedthesameindexdate,theinitialdiagnosisdateofvisionloss,asthecorrespondingcase.Foreachpair,theannualmedicalcostswerecalculatedduring theoneyearafterindexdate.Allmedicalcostswereadjustedto2010NewTaiwan dollars(NT$,US$1 NT$30).Ageneralizedlinearmodelwithgammadistribution andlog-linkwasusedtoestimatethedifferencesinpost-indexannualincreased costswhileadjustingforbaselinecharacteristicsandpre-indexloggedcosts.Bootstrappingtechniquewasusedtogenerate95%condenceinterval(CI). RESULTS: A totalof2,233,283,and350patientswithmoderatevisionloss,severevisionloss, andblindnesswereidentied,respectively.Patientswithmoderatevisionloss, severevisionloss,andblindnesshadNT$13,888(CI8,564-19,236),NT$23,768(CI 7,764-42,379)andNT$75,461(CI56,123-94,685)higherannualmedicalcostsascomparedwiththeirmatchedcontrols.Inpatientservices(62.3%)constitutedthelargestcomponentoftotalmedicalcostsforblindness,followedbyoutpatientservices (26.8%)andemergencecare(4.3%).Similarpatternswereobservedformoderate andseverevisionloss. CONCLUSIONS: ConsistentwiththeWesternworld,significantlyhighermedicalcostsareassociatedwithvisionlossinTaiwan.Theadditionalcostsseemtoberelatedtotheextentofthevisionloss,withblindness incurringthehighestcosts,followedbyseverevisionlossandthenmoderatevisionloss. PSS8 SYSTEMATICREVIEWOFTHEECONOMICLITERATUREONATOPIC DERMATITISINCHILDRENIskedjianM1,NavarroV2,KhondokerF1,FarahB3 1PharmIdeasResearchandConsulting,Oakville,ON,Canada,2PharmIdeasEuropeSAS,Lyon, France,3PharmIdeasResearchandConsulting,Ottawa,ON,CanadaOBJECTIVES: Asystematicreviewoftheliteraturewasperformedtogatherthe economicevidencerelatedtoatopicdermatitis(AD)inchildren. METHODS: OVID MEDLINE¨andEMBASEwereexploredbytworeviewersforacombinedsearch withtermsrelatedtoeconomicsandqualityoflifeinapaediatricpopulation,for theperiod1996-2010.Thisabstractreportstheresultsoftheeconomicreview.Two reviewersbrowsedabstracts,retrievedthesuitablearticles,separatedtheminto costofillness(COI)andeconomicevaluationsincludingcostcomparisons(CC), cost-effectiveness(CEA)andcost-utilityanalyses(CUA),andsummarizedkey points.Athirdpersonactedasoverallreviewerandadjudicatorincaseof disagreement. RESULTS: Fromaninitialsearchyielding704references,16articles wereincludedinthereview,7COIand9economicevaluations,including2CC,3 CUAand1CEAoftreatmentwithimmunomodulators,1CCofcorticosteroidsand 1CEAofhomeopathyaswellas1CEAonapartiallyhydrolysedinfantformulaused inprevention.Theinvestigationperiodspanfrom1995to2010.Fourstudieswere performedintheUK,3intheUS,2inAustralia,2inCanada,2inGermany,1in France,1inItalyand1covered10Europeancountries.Therangeofyearlymean costfortreatingADisverywideandincreasesaccordingtoseverity.Themaincost driversseemtobeconsultations,hospitalizations,emollientsandprescription medications,aswellastimelossforparents. CONCLUSIONS: Givenitshighprevalence,ADcarriesasubstantialeconomicburdenonsuchstakeholdersasthe nationalhealthcaresystemsandthefamilyoftheaffectedchildren.Inlightof thesendings,anapproachbasedonpreventionratherthantreatmentmayhave agreatereconomicimpact.Inturn,furtherresearchiswarrantedonthepotential economicbenetfrominterventionsaimedatprevention. PSS9 COST-EFFECTIVENESSOFCOLLAGENASEVERSUSHYDROGELDRESSINGSFOR CHRONIC-WOUNDDEBRIDEMENTINALONG-TERMCARESETTINGWaycasterC HealthpointLtd.,FortWorth,TX,USAOBJECTIVES: Toassessthecost-effectivenessofcollagenaseenzymaticdebridementrelativetohydrogelautolyticdebridementforpressureulcersinalong-term caresetting. METHODS: A2-stageMarkovmodelwasusedtodeterminetheexpectedcostsofwounddebridementbetweencollagenaseandhydrogeldressings forpressureulcersinalong-termcaresetting.ThetwoMarkovstatesusedinthe analysiswereanecroticnon-viablewoundbedandadebridedviablewoundbed. Outcomedatausedintheanalysisweretakenfromarandomizedclinicaltrialthat directlycomparedcollagenaseandhydrogeldressingsinalong-termcarenursing facility.Theprimaryoutcomeintheclinicaltrialwastheproportionofpatients achievingaviable,granulatedwoundbedat42days.Transitionprobabilitiesfor theMarkovstateswereestimatedfromtheclinicaltrial.Resourceutilizationwas basedonthewoundcaretreatmentregimenusedintheclinicaltrial.Resource utilizationcostswerederivedfromstandardcostreferencesandmedicalsupply wholesalers.Allcostsareexpressedin2010U.S.dollars.Thepayerperspectivewas takenintheeconomicanalysisandnocostdiscountingwasperformedduetothe shorttimehorizonofthestudy.One-wayandprobabilisticsensitivityanalyses wereperformedtoestablishtherobustnessoftheeconomicresults. RESULTS: The clinicaltrialestablishedthat85%ofthecollagenasesubjectsversus29%ofthe hydrogelsubjectsachievedadebridedwoundbedatthe42-dayassessment. Theeconomicanalysisindicatedthat42daysafterinitiatingtreatmentthecosteffectivenessratioforcollagenasewas$85perdebrideddaycomparedto$184per debrideddayusingahydrogeldressing.One-wayandprobabilisticsensitivityanalysesindicatedthattheresultswererobustwithinthemodeledparameters. CONCLUSIONS: Resultsindicatethatcollagenaseointmentiscosteffectiveforthe debridementofpressureulcerscomparedtoautolysiswithahydrogeldressing.A54VALUEINHEALTH14(2011)A1–A214

PAGE 63

PSS10 ACOST-EFFECTIVENESSANALYSISOFETANERCEPTFORTHETREATMENTOF MODERATEANDSEVEREPSORIASISINMEXICOVargas-ValenciaJ1,Sotelo-Guzm‡nM1,Mould-QuevedoJF2,Muci–o-OrtegaE3, Galindo-SuarezRM3 1EconopharmaConsultingS.A.deC.V.,MexicoCity,Mexico,2Pzer,NewYork,NY,USA,3Pzer,MexicoCity,MexicoOBJECTIVES: Biologictreatmentshaddramaticallychangedthetherapeutics,outcomesandcostofmanagementofpsoriasis,acommonchronicdiseasethat stronglyaffectsqualityoflifeofpatients.Theaimofthisstudywastoassessthe cost-effectivenessofbiologicalternativescurrentlyavailableinMexicofortreatmentofmoderatetoseverepsoriasisfromaninstitutionalperspective. METHODS: Adecision-treemodelwasdevelopedtosimulatetheclinicalcourseofpatients treatedwithetanercept,adalimumab,iniximaborustekinumabasrst-linetherapies,aswellastreatmentassociatedcosts(2-yeartimeframewitha5%annual discountrate).Effectivenessmeasuresweretheproportionofpatientsreaching 75%improvementinthePsoriasisAreaandSeverityIndex(PASI-75)andquality adjustedlifeyearsgained(QALY = s).Costsconsideredincluded:biologicsdrugs, concomitantmedication,medicalfollow-upandsideeffectsmanagement.Clinical responseofalternativeswasextractedfrompublishedliterature,whileunitcosts werecollectedfromInstitutoMexicanodelSeguroSocial(IMSS)ofcialdatabases. Probabilisticsensitivityanalyseswerecompleted. RESULTS: Aftertwoyears,the proportionsofpatientsreachingPASI-75were59%,62.1%,62.7%,and64.5%for adalimumab,etanercept,iniximabandustekinumab,respectively(p 0.077, Friedmantest);QALY = sassociatedtoeachalternativewere:1.5554,1.5633,1.5650 and1.5695,respectively(p 0.392,Friedmantest).Giventhatinthetimeframe consideredtherearenodifferencesbetweeneffectivenessoftherapies,acostminimizationratherthanacost-effectivenessanalysiswasperformed.Theexpectedmeancostsperpatientwere:US$27,281.95,US$25,929.85,US$31,542.76and US$38,050.59,respectively.EtanerceptisUS$1,352.1( 5.1%)lesscostlycomparedto adalimumab;US$5,612.91( 17.8%)comparedtoiniximabandUS$12,120.74 ( 46.74%)comparedtoustekinumab(themostcostlyalternative). CONCLUSIONS: Giventhateffectivenessofthebiologictreatmentsanalyzedissimilaroverthe time-horizonused,etanercepttreatmentrepresentsthelessexpensivealternative forthemanagementofmoderateandseverepsoriasisatIMSS. PSS11 COST-EFFECTIVENESSANALYSISOFFORMULARYMANAGEMENTSTRATEGIES FORFIRST-LINEAGENTSUSEDINOPEN-ANGLEGLAUCOMAANDOCULAR HYPERTENSIONBlaserDA1,OusterhoutMM1,GagnonJM1,CalabreseD2 1UMassMedicalSchool,Shrewsbury,MA,USA,2MedMetricsHealthPartners,Worcester,MA, USAOBJECTIVES: Todeterminewhichformularymanagementstrategyofophthalmic prostaglandins(PGs)isthemostcost-effectivefromtheperspectiveofamanaged careorganization(MCO)overoneyear. METHODS: Thiscost-effectivenessmodel containseightuniquearms,whichreectdifferentformularymanagementstrategiesforthePGs(bimatoprost,latanoprost,travoprost).Threearmsexaminethe cost-effectivenessofeachPGasrst-linetherapy,withtimololaddedassecondlinetherapyforpatientsnotreachinggoalintraocularpressure(IOP).Thenext threearmsexaminerst-linetherapywithtimolol,analternativeandlesscostly primarytherapy,withapreferredPGassecond-linetherapy.Anadditionalarm examinesnotselectingapreferredPGasrst-linetherapywithtimololassecondlinetherapy,andthenalarmexaminestimololasrst-linetherapy,butdoesnot selectapreferredPGassecond-line.Aneffectivelytreatedpatientwasdenedasa patientachievinganIOPlessthan18mmHgafterthreemonthsoftherapy.If patientswereunabletoreachgoalIOPwithsecondary-linetherapy,dorzolamide wasadded.PatientsunabletotoleratetherapyorachievegoalIOPrequiredadditionalphysicianvisits.Costsincludedprescriptionmedicationsandphysicianvisitsin2010USdollars. RESULTS: Thecost-effectivenessratios(CER)forrst-line therapywithapreferredPGwere$815.13,$961.71,and$889.13pereffectively treatedpatientforbimatoprost,latanoprost,andtravoprost,respectively.The CERsforrst-linetherapywithtimolol,followedbyapreferredPG,were$436.77, $499.77,and$462.21forbimatoprost,latanoprost,andtravoprost,respectively.Ifa preferredPGwasnotselected,theCERswere$910.95forrst-linetherapyand $477.77forsecond-linetherapy.Sensitivityanalysesshowedthatreducingthe priceoflatanoprostby9percentandtravoprostby3percentyieldsequivalentCERs asbimatoprost. CONCLUSIONS: Timololfollowedbyabimatoprostisthemost cost-effectiveoftheeighttreatmentstrategiesexaminedinthismodel. PSS12 ADISCRETEEVENTSIMULATIONTOOPTIMISETHEALLOCATIONOF CONSTRAINEDHOSPITALRESOURCESFORGLAUCOMACraneGJ1,KarnonJ1,KymesS2,CassonR1,MetcalfeA1,HillerJE1 1UniversityofAdelaide,Adelaide,SouthAustralia,Australia,2WashingtonUniversity,St.Louis, MO,USAOBJECTIVES: Aliteraturereviewidentiedthatconstrainedresourcemodelsof servicedeliveryandorganisationdonotincludepatient-basedoutcomes.Theaim ofthisstudyistodevelopanevidence-basedframeworkrepresentingcostsand healthbenets,toinformbestuseofconstrainedhealthcareservicesinaroutine clinicalsetting. METHODS: Adiscreteeventsimulation(DES)modelwasdeveloped torepresenttheuseofglaucomaservicesattheRoyalAdelaideHospital.TheDES describesdiseaseprogressionandpathwaysofcare,wherediseaseprogressionis inuencedbythefrequencyandcontentofhospitalvisits,andviceversa.Individuals,whoarecompetingforaccesstohospitalservices,areassignedcharacteristicsthatrepresentrelevantpatientanddiseasecharacteristics.Acrossallindividuals,thesedatainformthedemandforservicesovertime,whichiscombinedwith informationonthesupplyofavailableresourceswithinthesystemtoanalyze alternativeapproachestotheuseofavailableresources. RESULTS: Thebasecase modelhasbeenvalidatedagainsthospitaldatarepresentingbothprocessand clinicalmeasuresovertime.Analysisofthemodelshowsvariationinthetotal QALYsgainedbycohortsofglaucomapatientsovertheirremaininglifetime,accordingtoalternativetreatmentdecisionalgorithms(e.g.medicationversussurgicalintervention);surveillanceandimagingalgorithms(e.g.variationinrelative follow-upschedulesforhighandlowprioritypatients). CONCLUSIONS: TheappliedframeworkillustratesthepotentialvalueofDESinmodellingthecostsand healthbenetsofalternativeapproachestoorganisingscarcephysicalresources, providingestimatesofhealthgainsthatcanbeachievedintheabsenceofthe introductionofnewtechnologies. PSS13 INCREMENTALCOST-UTILITYANALYSISOFDEXAMETHASONEINTRAVITREAL IMPLANTFORTHETREATMENTOFMACULAREDEMAFOLLOWINGRETINAL VEINOCCLUSIONKowalskiJ1,YehWS1,O'LearyB2,AckermanS3,SharmaS4,AdenaM5 1Allergan,Irvine,CA,USA,2CovancePtyLtd,NorthRyde,NSW,Australia,3CovanceMarket AccessServices,Inc.,SanDiego,CA,USA,4Queen'sUniversity,Kingston,ON,Canada,5DatalyticsPtyLtd,Bruce,ACT,AustraliaOBJECTIVES: Macularedema(ME)followingretinalveinocclusion(RVO)isacommoncauseofvisionloss.Theobjectiveofthisresearchwastoassessthelifetime incrementalcost-effectivenessofdexamethasoneintravitrealimplant0.7mg(DEX) treatmentversusobservationforMEfollowingcentralorbranchRVO(CRVOor BRVO)fromaUSpayerperspective. METHODS: AnExcel-basedMarkovmodelwith 6heathstatesbasedonvisualacuity(VA)plusoneabsorbing(Death)statewas developed.Transitionmatriceswerederivedusingindividualpatient-leveldata pooledfromtwoidenticalphase3studies.Enrolledstudypatientsatbaselinehad meanageof65yearsandstudy-eyeVAof20/80.DEXpatientswereassumedto receiveupto6-treatmentsover3years.Directmedicalcostsincludeddrug,procedure,maintenancecare,andadverseevents.Additionaldirectscostsandmortality wasassignedtopatientswhobecamelegallyblind( 20/200inBSE).Utilityscores weredirectlycalculatedfromstudydataviatheVisualFunctionQuestionnaire UtilityIndexandhealthstatesdifferentiallyvaluedbywhetherthetreatedeyewas thebetter-(BSE)orworse-seeing-eye(WSE).90%ofpatientswereassumedtohave RVOintheirWSEatmodelentryandatime-dependentriskoffelloweyeoccurrence(FEO)wasincorporated.Costsandoutcomeswerediscountedat3%.Deterministicandprobabilisticsensitivityanalyseswereconducted. RESULTS: ReferencecaseICERswere$23,416and$20,597perQALYforBRVOandCRVO(respectively); andsensitivetothepercentofpatientsincurringtheRVOintheBSE,riskofFEO,and costofvisionloss. ProbabilisticsensitivityanalysesdemonstratedthattheICERsfall belowathresholdof$50,000perQALYin87%and92%ofsimulationsforBRVOand CRVO,respectively. CONCLUSIONS: Usingathresholdof$50,000,DEXtreatment comparedtoobservationissupportedasacost-effectivetreatmentoptionforME followingBRVOorCRVO. PSS14 COST-UTILITYANALYSISOFRANIBIZUMAB(LUCENTIS)INWET-AMDBASED ONREAL-LIFEDATACOLLECTEDINTHEHELIOSSTUDYAFTERRANIBIZUMAB REIMBURSEMENTINBELGIUMMoeremansK1,GerlierL1,DrieskensS2,AnnemansL3,BriŽH4,DeruyckM4,FaesP4, VancayzeeleS4 1IMSHealthConsulting,Vilvoorde,Belgium,2PanaceaOfcinalisBVBA,Brussels,Belgium,3Ghent University-BrusselsUniversity,Ghent,Belgium,4NovartisPharmaAG,Vilvoorde,BelgiumOBJECTIVES: Assesscost-utilityofranibizumab(Lucentis¨)inwet-AMDbasedon real-lifeevidenceregardingqualityoflife,treatmentpatternsandeffectiveness collectedintheBelgianobservationalHELIOSstudyincluding253patientstreated withranibizumabafterreimbursementapproval. METHODS: Anexistingmodel, developedforclinicaltrial-basedanalyses,wasextendedtoincludeareal-life ranibizumabarm.This10-yearMS-ExcelMarkovmodelwith5visualacuity(VA) levelsand1deathstatepredictsVAinpatientstreatedwithranibizumab,Visudyne¨photodynamictherapy(PDT)orbestsupportivecare(BSC).Transitionprobabilities andtreatmentfrequencyforranibizumab(onaverage5injectionsduringyear1) wereprovidedbytheHELIOStrial(1-yearinterimdata).Forcomparatorsthesedata wereobtainedfromseveralclinicaltrials(ANCHOR,MARINA,PIERandTAP).Baselinecharacteristicsandutilitiescorrelatingwithtreated-eyeVAwereobtained fromtheHELIOSstudy.Two-yeartreatmentdurationwasmodelled,followedby BSC.Costs(2010)fromtheperspectiveofthehealthcarepayer socialservices patientwereobtainedfromliteratureandexpertopinion.Costs(3%)andoutcomes (1.5%)werediscounted.Sensitivityanalysescoveredvariabilityinefcacy,costs, treatmentfrequencyandutilities. RESULTS: Visiongaininreal-lifewasinlinewith PRNclinicaldatabutlowerthanobservedinmonthly-dosingclinicaltrials.Injectionfrequencyandcostswerelowerinreal-lifethananticipatedfromranibizumab clinicaltrials.Base-caseanalysesversusBSC(MARINA,PIERandTAP)consistently revealedcost-utilityofranibizumab,withresultsrangingfromdominanceto 8,367 /QALY.Base-caseanalysesversusPDT(PIERandTAP)suggesteddominance ofranibizumab.Resultsweremostsensitivetothecostofblindnessandtime horizon,butremainedgenerallywithinacceptablelimits.Atanacceptability thresholdof35,000 /QALY,theprobabilitythatreal-liferanibizumabiscost-effectiverangedfrom84%to100%. CONCLUSIONS: Real-lifeuseofranibizumabin wet-AMDappearedhighlycost-effectivecomparedtoBSCandPDT.A55VALUEINHEALTH14(2011)A1–A214

PAGE 64

SensorySystemsDisorders–Patient-ReportedOutcomes&Preference-Based Studies PSS15 DEVELOPMENTOFAMULTIATTRIBUTEINSTRUMENTFORESTIMATING UTILITIES(PREFERENCEWEIGHTS)INPEOPLEWITHGLAUCOMAFROMTHE NATIONALEYEINSTITUTEVISUALFUNCTIONINGQUESTIONNAIRE(NEI-VFQ)KymesS1,PetersCM1,BeusterienK2,KotakS3,GrinspanJ2,StwalleyD4,PleilA5 1WashingtonUniversitySchoolofMedicine,St.Louis,MO,USA,2OxfordOutcomesLtd., Bethesda,MD,USA,3Pzer,Inc.,NewYork,NY,USA,4WashingtonUniversity,St.Louis,MO, USA,5Pzer,Inc.,SanDiego,CA,USAOBJECTIVES: Weusedmulti-attributetheorytodevelopautilityelicitationinstrumentforpeoplewithglaucomabasedontheNationalEyeInstitute-VisualFunctioningQuestionnaire(NEI-VFQ),avisionspecicqualityoflifeinstrumentfrequentlyusedinclinicaltrials. METHODS: NEI-VFQresponsesfor1677people enrolledinglaucomapreventiontrialswereanalyzedtoidentifyitemsresponsive todifferencesinvisionstatus.Wethenconstructedaconjointinterviewadministeredto48peoplewithglaucomaandidentiedthe6itemsofgreatestimportance topeoplewithglaucoma.Usingtheseresults,weconstructedaweb-basedinterviewusingstandardgamble(SG)andvisualanalogscales(VAS)administeredtoa communitysampleof404people. RESULTS: Theveattributesmostimportantto peoplewithglaucoma(andresponsivetochangeindiseasestatus)were:abilityto read,drivingatnight,abilitytoleavehome,needinghelpwithactivitiesandability toaccomplishtasks.Weaddedasixthattribute,peripheralvision,eventhoughit wasnotratedhighlybytheparticipantsasthisisanimportantfunctionlostearly inthediseaseprocess.Finalutilityestimatesweremadeusingamethodsimilarto thatemployedintheHealthUtilityIndex.Theresultshadexcellentfacevalidity andwenotetherangeofutilitylossfrommildlossoffunctiontoseveredifculty: 1)Abilitytoread(0.011-0.039);2)Drivingatnight(0.011-0.037);3)Leavinghome (0.014-0.056);4)Needinghelpwithactivities(0.027-0.063);5)Abilitytoaccomplish tasks(0.016-0.046);and6)Peripheralvision(0.008-0.032). CONCLUSIONS: Wehave developedaninstrumentthatcanbeusedtoestimatetheutilitylossinpeoplewith glaucomabasedupontheNEI-VFQ.Whenusedinclinicaltrials,theinstrumentwill provideanestimateoftheutilitylossassociatedwiththeprogressionofglaucoma. Furtherworkwillberequiredtoreneandvalidatetheseestimates. PSS16 DEVELOPMENTOFADISEASESPECIFICVERSIONOFTHEEQ-5DFORUSEIN PSORIASISLloydA1,SwinburnP1,BoyeKS2,EdsonE2,BowmanL2 1OxfordOutcomesLtd.,Oxford,Oxfordshire,UK,2EliLillyandCompany,Indianapolis,IN,USAOBJECTIVES: TheEQ-5Disawidelyusedgenericmeasureofhealth.Previousresearchhashighlightedpossibleshortcomingsincontentvalidityincertainconditions(e.g.vision,hearingloss).Theobjectiveofthisstudywastodetermineifthis wastrueforpeoplewithpsoriasisandtoaddressthisthroughthedevelopmentof adisease-specicversionoftheEQ-5Dforuseinpsoriasis. METHODS: Aliterature review,consultationwithmedicalexpertsandin-depthqualitativeinterviewswith psoriasispatientswithvaryingdegreesofdiseaseseverity(n 8)wereundertaken tounderstandthequality-of-lifeburdenassociatedwithpsoriasis.Aniterative processidentiedfournewcandidatedimensions(skinirritation,skinappearance, self-condenceandsocial/relationshipdifculties).Thesewereincludedalongside theexistingvedimensionsofEQ-5D.Inavalidationstudy100psoriasispatients completedthenewversionofEQ-5D;theDermatologyLifeQualityIndex(DLQI) andSelf-AdministeredPsoriasisAreaSeverityIndex(SAPASI). RESULTS: Psychometricanalysessuggestedthatallcandidatedimensionsweresensitivetodisease severityandhadexcellentconcurrentvalidity.Thenewdimensionssignicantly improvedthepredictivepowerofEQ-5Dinmeasuringhealthstatusasassessedby theDLQIwithanR2increaseof0.213whenthreecandidatedimensionswere added.Almost20%ofpatientsindicatednoproblemsoneachdimensionofconventionalEQ-5D(state11111),butreportedsubstantialimpairmentsonthepsoriasisdimensions,indicatingthelimitationsofexistingEQ-5Dinpsoriasis. CONCLUSIONS: ThisstudyispartoflargerresearchattheEuroQolGrouptoinvestigatethepotentialdevelopmentofdiseasespecicversionsofEQ-5D.Thestudy demonstratesthatincertaindiseaseareasthereisneedformoreaspecicmeasure.Theadditionofappropriatedimensionscansignicantlyincreasethesensitivityoftheinstrument.Furtherworkisneededtocapturepreferenceweightsfor thisnewinstrumenttosupportitsinclusionineconomicevaluation. PSS17 THEEFFECTOFORALCP-690,550ONPRURITUSINPATIENTSWITH MODERATE-TO-SEVEREPLAQUEPSORIASISMamoloCM1,BushmakinAG2,CappelleriJC2,StewartM2 1Pzer,Inc.,NewLondon,CT,USA,2PzerGlobalResearchandDevelopment,Groton,CT,USAOBJECTIVES: ToassesstheeffectoftheoralJanuskinaseinhibitorCP-690,550on pruritusinpatientswithpsoriasis. METHODS: Ina12-week,double-blind,placebocontrolledPhase2bstudy,patients(n 197)withmoderate-to-severeplaquepsoriasiswererandomizedtoCP-690,550(2,5or15mgBID)orplacebo.Prurituswas patient-assessedusingtheItchSeverityScore(ISS) ,a0(noitching')to10(worst possibleitching')numericratingscale,recordeddailyduringtherst2weeksof treatmentandatallclinicvisits.Theclinicallyimportantdifference(CID)and clinicallyimportantresponder(CIR)ontheISSweredenedusingthePatient GlobalAssessmentasananchor. RESULTS: AtBaseline,mean(SD)ISSforpatients inthe2,5,and15mgBIDandplacebogroupswere7.04(2.65),6.98(2.27),6.96(2.23), and6.78(2.77),respectively.TheestimatedCID(95%CI)was1.64(1.50;1.78)andCIR was29.8%(23.30%;36.40%).OveralldifferencesinISSvaluesbetweenactivedrug armsandplacebowere 2.82(95%CI: 3.71; 1.92), 3.02( 3.93; 2.12),and 3.87 ( 4.76; 2.98)for2,5,and15mgBIDvsplacebo,respectively;alldifferenceswere larger(inabsolutevalue)thantheCIDof1.64,indicatingthatthetreatmenteffects werenotonlystatisticallysignicantbutalsoclinicallyrelevant.AtWeek2,the numberofISSresponderswas77.78%,68.75%,and76.60%for2,5,and15mgBID, respectively,vs34.00%forplacebo(p 0.0001);atWeek12thenumberofISSrespondersincreasedto91.89%,87.18%,and100%inthe2,5,and15mgBIDgroups, respectively,vs29.41%forplacebo(p 0.0001). CONCLUSIONS: Inpatientswith psoriasis,CP-690,550producessignicantandclinicallymeaningfulreductionsin pruritus,asmeasuredbythepatient-reportedISS. PSS18 DISEASESEVERITYEVALUATIONAMONGDERMATOLOGICALOUT-PATIENTS: ACOMPARISONBETWEENTHEASSESSMENTSOFPATIENTSANDPHYSICIANSTabolliS,SampognaF,PagliarelloC,ParadisiA,SpagnoliA,AbeniD IDIIRCCS,Rome,ItalyOBJECTIVES: Theassessmentofapatient'sdiseaseseverityisanessentialcomponentintheformulationoftreatmentstrategies.Thisstudywantstocomparethe dermatologicaldiseaseseverityassessmentbypatientsandbyphysicians,andto describethepossiblediscrepanciesbetweenthem. METHODS: Foreachpatient,we obtainedthePhysicianGlobalAssessment(PhGA)andthePatientGlobalAssessment(PtGA).Datawerecompletedfor2.578patients.Sixty-onephysiciansparticipatedinthestudy.WecalculatedtheagreementbetweenPtGAandPhGAscores usingtheweightedkappastatistics;amultinomiallogisticregressionwasperformedtoassesstheriskofdisagreementconsideringbothpatientandphysician variables. RESULTS: Differencesinthepercentagesofseveritylevel,identiedby patientsandbyphysicians,werealwaysstatisticallysignicant(p 0.05).Overall, theweightedCohen'skappawasintherangeof0.09-0.34,dependingonthe diseases.Genderdifferencesbetweenpatientsandphysiciansdidnotinuencethe agreement.Inthemultinomialmodelfemalepatients(OR 1.38;95%CI,1.07-1.77), patientswithhighereducationallevels(OR 2.71;95%CI,2.12-3.46),andpatients withimpairedqualityoflife(OR 1.56;95%CI1.23-1.97)hadahigherrisktobe underestimatedfortheirdiseaseseveritybyphysicians,independentlybyphysiciangenderandexperience. CONCLUSIONS: Combiningthesubjectivereportwith theobjectiveseverityassessmentofthelesions,dermatologistsmayreachabetter determinationofhowseverityofdiseaseisperceivedbytheirpatientsandhow theyfeelabouttheeffectivenessoftreatment.PtGAandPhGAmightbeconsidered inroutineclinicalassessmentsandnotonlyforresearchactivities. PSS19 PSYCHOMETRICEVALUATIONOFTHENATIONALEYEINSTITUTEVISUAL FUNCTIONQUESTIONNAIRE25ANDVISUALFUNCTIONQUESTIONNAIRE UTILITYINDEXINPATIENTSWITHNON-INFECTIOUSINTERMEDIATEAND POSTERIORUVEITISNaikRK1,GriesKS2,RentzA3,KowalskiJW1,RevickiDA3 1Allergan,Inc.,Irvine,CA,USA,2UnitedBioSourceCorporation,Seattle,WA,USA,3United BioSourceCorporation,Bethesda,MD,USAOBJECTIVES: ToevaluatethepsychometricpropertiesoftheNationalEyeInstitute VisualFunctionQuestionnaire25-item(NEIVFQ-25)andVisualFunctionQuestionnaireUtilitiesIndex(VFQ-UI)inpatientswithnon-infectiousintermediateand posterioruveitis. METHODS: Secondaryanalysisofpooleddatafroma26-week, multicenter,masked,randomized,sham-controlledPhase3trialofdexamethasoneintravitrealimplant.Visionspecichealthrelatedqualityoflifewasassessed usingtheNEIVFQ-25atbaselineandweeks8,16,and26/earlyexit.TheEQ-5Dand SF-36wereadministeredatbaseline.InternalconsistencyreliabilityandreproducibilitywasmeasuredusingCronbach'salphaandintraclasscorrelationcoefcient (ICC).ValiditywasassessedwithSpearman'sproduct-momentrankcorrelations andknown-groupsofBCVAandvitreoushazeseverity.Clinicallysignicantdifferencewasassessedusingdistributionandanchor-basedmethods. RESULTS: The studyincluded224subjectswithnon-infectiousintermediate(80.4%)orposterior uveitis(19.6%);meanage44.6years;63.4%female;60.3%Caucasian;medianvitreoushazescoreatbaseline2.0andmedianvisualacuity(ETDRS)62.5letters(84.4% treatedintheirworseseeingeye).TheNEIVFQ-25andtheVFQ-UIdemonstrated goodinternalconsistency(Cronbach'salpha0.87-0.94)andtest-retestreliability (ICC0.72-0.88).Spearman'sproduct-momentrankcorrelationsbetweentheNEI VFQ-25andVFQ-UIscoresandtheSF-6D,EQ-5D,andBCVAatbaselinerangedfrom lowtomoderate.Therewasasignicantassociationbetweenvisualfunctioning andknowngroupsofvisualacuity,withgroupcomparisonsdemonstratingsignificantdecreaseinvisualfunctioningbetween 20/40and 20/40to 20/200 (p 0.05).Atweek8,clinicallysignicantdifference,basedonthestandarderrorof measurement,was0.04fortheVFQ-UIandrangedfrom4.63to8.92fortheVFQ-25 domainsand3.86forthecompositescore. CONCLUSIONS: TheNEIVFQ-25andthe VFQ-UIarereliableandvalidmeasuresofvision-relatedfunctioningandpreference-basedstatusinpatientswithnon-infectiousintermediateandposterioruveitis. PSS20 MEDIATIONMODELINGANDMEASUREMENTCHARACTERISTICSOFTHEITCH SEVERITYSCOREFROMAPHASE2BTRIALOFORALCP-690-550INPATIENTS WITHMODERATETO-SEVEREPLAQUEPSORIASISMamoloC1,BushmakinAG2,CappelleriJC2,StewartM2 1Pzer,NewLondon,CT,USA,2PzerGlobalResearchandDevelopment,Groton,CT,USAOBJECTIVES: ToassessdirectandindirecteffectsoftheoralJanuskinaseinhibitor CP-690,550onpruritusinpatientswithpsoriasis. METHODS: Ina12-week,doubleblind,placebo-controlledPhase2btrial,197patientswithmoderate-to-severepsoriasiswererandomizedtoCP-690,550(2,5or15mgBID)orplacebo.Prurituswas patient-assessedusingtheItchSeverityScore(ISS) ,a0(noitching')to10(worstA56VALUEINHEALTH14(2011)A1–A214

PAGE 65

possibleitching')numericratingscale,recordeddailyduringtherst2weeksof treatmentandatclinicvisits.MediationmodelingwasusedtodeterminetherelationshipbetweenISS,thePhysician'sGlobalAssessment(PGA,aclinicalmeasure ofpsoriasisseverityassessingerythema,induration,andscaling),andtreatments (2,5,15mgBIDvsplacebo).MediationmodelingincludedISSasthedependent variable(averagedpost-treatmentISSduringweeks2-12foreverypatient),PGAas themediatorvariable(averagedpost-treatmentPGAscoreduringweeks2-12for everypatient),andtreatments(witheachdrugdosecomparedwithplacebo)asthe independentvariables.Psychometric(correlational)analyseswereperformedon ISSusingpost-baselineassessments. RESULTS: ThedirecteffectsofallCP-690,550 dosesonISSwere70%-81%(vsplacebo;p 0.001),indicatingthatdrugeffectson pruritusweremostlyindependentofimprovementsinerythema,induration,and scaling.DailyISSmeasurementshadacceptabletest-retestreliability(intraclass correlation:0.83)inpatientswhodidnotchangeonPGAduringtherst2weeksof thetrial.PearsoncorrelationsbetweenISSandothermeasureswereconsistent withexpectations;forexample,weeklyISScorrelatedwithweeklyPGA(range: 0.3 0.5)andPatientGlobalAssessment(range:0.6 0.7). CONCLUSIONS: Inpatientswithpsoriasis,CP-690,550hasadirect,benecialeffectonpatient-reported pruritusthatisindependentfromimprovementsinclinician-reportedpsoriasis severity.Post-baselineassessmentsofISSshowedfavorablepsychometriccharacteristics,indicatingitspotentialuseasasimpletoolforassessmentofpruritusin clinicaltrials. PSS21 GAPANALYSISFORPATIENT-REPORTEDOUTCOMESMEASURESFORALOPECIADennee-SommersB1,GalipeauN1,FitzgeraldK1,EvansC1,DanielsS2,BurgessSM2 1MAPIValues,Boston,MA,USA,2Allergan,Inc.,Irvine,CA,USAOBJECTIVES: Toidentifyanexistingpatient-reportedoutcome(PRO)measureto substantiatelabelingorpromotionalclaimsfortreatmentinalopecia. METHODS: MEDLINE¨,PROandQualityofLifeInstruments(PROQOLID)andPROandDrug MarketingAuthorizations(PROLabels)databasesweresearchedandrelevantinstrumentswereidentiedbyreviewingabstracts,articles,andlabelsforconcepts ofinterest(i.e.,color/darkness,length,fullness/thickness,generalappearance/ attractiveness,self-esteem).Theinitialinstrumentswerefurtherreviewedand subsequentlyexcludediftheywere:notPROs(e.g.,haircoverage),notspecicto scalphairgrowth,ordevelopedwithoutpatientinput.Thenallistofinstruments wasevaluatedbasedonrequirementsoutlinedintheUSFoodandDrugAdministration'sFinalPROGuidanceforIndustry(e.g.,developmentandconrmationof conceptualframework,patientinvolvementinitemgeneration,contentvalidity, recallperiod,responseoptions,considerationofpatientpopulationusedfordevelopmentinrelationtofutureclinicaltrials). RESULTS: Forty-veinstrumentswere identied;23werepatient-reported,17wereinvestigator-rated,threeweredevices andtwowereobjectivehairmeasures.Forty-onewereexcludedbasedonthecriteriaoutlined.Thefourinstrumentsconsideredfurtherincludedalopecia-specic itemsandinstrumentsusedtosupporttheapprovalofdrugstostimulatehair growth(e.g.,Propecia¨),specically,theKingsleyAlopeciaQuestionnaire(KAP), theHairGrowthQuestionnaire,theHairProblemList,andtheWomen'sAndrogenicAlopeciaQualityofLifeQuestionnaire(WAA-QOL). CONCLUSIONS: Noneof theinstrumentsmetthePROGuidancerequirements;however,somecouldbe adapted.TheHairGrowthQuestionnairecouldberevisedandtestedtoconrmif thecontentoftheinstrumentisrelevantforwomenwithalopeciaasitwasdevelopedbasedonmaleinputonly.Alternatively,theHairProblemListortheWAAQOLcouldbesupplementedtoincludemoreconceptsofinterestandtestedto conrmifthecontentoftheinstrumentisrelevantforbothmenandwomenwith alopecia. PSS22 SYSTEMATICREVIEWOFTHEQUALITYOFLIFELITERATUREINCHILDREN WITHATOPICDERMATITISIskedjianM1,NavarroV2,KhondokerF1,FarahB3 1PharmIdeasResearchandConsulting,Oakville,ON,Canada,2PharmIdeasEuropeSAS,Lyon, France,3PharmIdeasResearchandConsulting,Ottawa,ON,CanadaOBJECTIVES: Asystematicreviewoftheliteraturewasperformedtoelicitthe publishedevidencerelatingtoqualityoflife(QOL)inchildrenwithatopicdermatitis(AD). METHODS: OVIDMEDLINE¨andEMBASEwereexploredbytworeview ersforacombinedsearchwithtermsrelatedtoeconomicsandQOLinapaediatric population,fortheperiod1996-2010.ThisabstractreportstheresultsoftheQOL review.Tworeviewersbrowsedabstracts,retrievedsuitablearticlesandsummarizedkeyndings.Athirdpersonactedasoverallreviewerandadjudicatorincase ofdisagreement. RESULTS: Fromaninitialsearchyielding704references,51primaryresearcharticleswereincludedinthereview,14reportingonQOLasprimary outcomeand37assecondaryoutcome.QOLasaprimaryoutcomewasreportedfor Europe(n 7),NorthAmerica(n 4),Asia(n 3),Australia(n 1)andSouthAmerica (n 1).Onestudyreportedpreference-basedoutcomes.Moststudieswerebasedon AD-specictoolssuchastheChildren'sDermatologyLifeQualityIndexorthe DermatologyLifeQualityIndex(n 6),theInfants'DermatitisQualityofLifeIndex (n 4),theQualityofLifeIndexforAtopicDermatitisortheParent'sIndexofQuality ofLifeAtopicDermatitis(n 2)andtheDermatitisFamilyImpactQuestionnaire (n 5).OnestudywasbasedongeneralQOLmeasuresfromthe12-itemShort-Form HealthSurvey(SF-12).Thestudiestargetingveryyoungchildrenmostoftenused parentsasproxies.MoststudiespointedtoaninversecorrelationbetweenQOLand severityaswellascorrelationbetweenvariousinstruments.Studiesreportingon QOLasasecondaryoutcomeconrmedthosendings. CONCLUSIONS: Thereisa clearlackofstudieselicitinghealthstateutilities.Furthermore,mostAD-specic toolsdonotprovideastandard,quantitativemeasurementinrelationtoperfect healthaswoulddopreferencebasedstudiesrequiredforcost-utilityanalyses. Furtherresearchshouldfocusonutilitymeasurement. SensorySystemsDisorders–HealthCareUse&PolicyStudies PSS23 COMPARINGHEALTH-RISKBURDENANDTOTALHEALTHCARECOSTSOF PSORIASISWITHTOPFIVECHRONICCONDITIONSNaimA1,PittsJ2,ChenCY3,WrightD3,ChalkMB2,EdingtonD3 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2Chalk&Associates,Austin,TX, USA,3UniversityofMichigan,AnnArbor,MI,USAOBJECTIVES: Psoriasis(PsO)isachronic,episodic,immunologicalskindiseasethat affectsapproximately2-2.5%ofUSpopulation.Itmayresultinworkproductivity lossandhighoverallhealthcarecostsforemployers.However,theimpacton health-risksisnotwellstudied.Wecomparedhealth-risks,lostproductivity,and totaldirecthealthcarecostsamongindividualswithPsOandvemostprevalent chronicconditions. METHODS: Health-riskinformationandself-reportedillness daysusinghealth-riskassessment(HRA)datawereexaminedanddirectHealthcarecosts(medicalandpharmacy)usingclaimsdataforemployees,retirees,and theiradultdependentsofalargeself-insuredemployerwerecomparedfrom20022006amongindividualswithRAandvemostprevalentchronicconditions; asthma,coronaryarterydisease/congestiveheartfailure(CAD/CHF),diabetes,hypertension,chronicobstructivepulmonarydisease(COPD). RESULTS: 54individualswithPsOwereidentied.ThePsOcohorthadmoderatehealth-riskscore(2.8/5) whichwascomparabletoindividualswithasthma(2.8/5),CHF(2.9),hypertension (2.6)andlowerthandiabetes(3.2/5)andCOPD(3.4/5).Ahigherproportion(35.2%)of individualswithPsOhad 5illnessdaysperyearascomparedwithindividuals withasthma(31.8%),CAD/CHF(25.9%),hypertension(20.4%),diabetes(28.4%),and COPD(33%).Annualdirecthealthcarecostswerealsohigherforindividualswith RA($28,933)ascomparedtoindividualswithasthma($25,814),CAD/CHF($22,916), hypertension($18,632),anddiabetes($28,224),andlowerascomparedtoindividualswithCOPD($38,839). CONCLUSIONS: IndividualswithPsOhavesimilar health-risksbuthigherillnessdays,anddirecthealthcarecostsascomparedtothe individualswithvemostprevalentchronicconditions.Psoriasisisahighcosts andlostproductivitydriverforemployers.Populationhealthbasedprogramsthat engageemployeesinappropriatelymanagingtheirchronicconditionscanhelp employersreducehealth-risks,improveproductivity,andmayhelpreducehealthcarecostsaswell. PSS24 MEDICATIONCHOICEANDASSOCIATEDHEALTHCAREOUTCOMESAND COSTSFORPATIENTSWITHACNEANDACNERELATEDCONDITIONSINTHE UNITEDSTATESPatelP1,LinHC2,FeldmanSR3,FleischerAB3,NahataM4,BalkrishnanR2 1UniversityofGeorgia,Athens,GA,USA,2UniversityofMichigan,AnnArbor,MI,USA,3Wake ForestUniversity,Winston-Salem,NC,USA,4OhioStateUniversity,Columbus,OH,USAOBJECTIVES: Acneisacommonconditionforwhichmultipletreatmentoptions areavailable.Thepatternsofpharmacotherapyforacneandsimilarconditions, andtheeffectofthosepatternsoncost,arenotwellcharacterized.Thisstudy examinedtheimpactsofpatientdemographicsandmedicationchoicesonpatient'shealthstatusandassociatedmedicationcosts. METHODS: Aretrospective cross-sectionalstudywasconductedusingthe2007MedicalExpenditurePanel Survey(MEPS)database.Informationonpatientdemographics,healthstatus,medicationutilization,andmedicationcostswereobtainedfromthedatabaserepresenting3,784,816patientswithacneandsimilarconditions. RESULTS: Weighted multiplelinearregressionanalysesindicatedthattheuseoftopicalretinoidswas preferredincombinationwithothertreatmentsratherthanitsmonotherapy.Oral antibioticswerewidelyprescribedanditsusewasassociatedwithasignicant decreaseintotalannualprescriptionspending.Useoforalretinoidsandoralcontraceptivesincreasedtheannualprescriptioncostssignicantly.Increaseinannualdrugrellswasnotassociatedwiththeimprovementinhealthstatus. CONCLUSIONS: Weobservedanassociationwithmedicationchoiceforacneand acnerelatedconditionsonmedicationspending.Pharmacologictreatmentofacne signicantlyaddstoacnerelatedannualhealthcarecostscomparedtononpharmacologictreatment. PSS25 MEDICATIONADHERENCETOTOPICALMEDICATIONSANDHEALTHCARE EXPENDITURESINMEDICAID-ENROLLEDCHILDWITHATOPICDERMATITISOuHT1,FeldmanSR2,BalkrishnanR3 1UniversityofMichiganCollegeofPharmacy,AnnArbor,MI,USA,2WakeForestUniversity, Winston-Salem,NC,USA,3UniversityofMichigan,AnnArbor,MI,USAOBJECTIVES: Toidentifyfactorsthatpredictadherencetotopicalmedicationin pediatricpopulationwithatopicdermatitis(AD),andassesstheirimpacton healthcareexpenditures. METHODS: ADpatient'sageunder12yearsoldusing topiccorticosteroidortopicalcalcineurininhibitor(TCI)wasidentiedusingMarketScanTMMedicaiddatabasefrom2005to2007.AdherencetoADmedicationand costsforallhealthcareclaimsandcostsforADrelatedclaimswereoutcomesof interestandmeasuredover12monthswhenADmedicationstarted.Medication adherencewasmeasuredusingmedicationpossessionratio.Multipleregression analysiswascarriedtoexaminepredictorsformedicationadherenceandtheir predictionsonhealthcarecosts. RESULTS: 4,182patientswereincluded,witha meanageof4years.AdherencetoADmedicationaveragewaslow(41%),withthe lowestratesinpatientswithlowpotencycorticosteroidtherapyaloneincluding alclometasone,desonideorhydrocortisone(39%)andthehighestratesinthoseA57VALUEINHEALTH14(2011)A1–A214

PAGE 66

whousedcombinationtherapy(topicalcorticosteroidandTCI)(68%).Typeof healthplan,Medicaideligibilitystatus,numberoftherapeuticclass,comorbidity, hospitalizationornotandADrelatedcostsduring12monthsbeforeADmedication startedweresignicantlyassociatedwithADmedicationadherence.Adherenceto ADmedicationwassignicantlynegativelyassociatedwithtotalannualhealthcarecosts(p 0.001)andwithADrelatedcosts(p 0.001),adjustedfor patientdemographic,comorbidity,andhealthcareutilizationcharacteristicsbeforeADmedicationstarted. CONCLUSIONS: Pooradherencetotopicalmedicationwasobservedin pediatricAD,andadherenceratesdifferedbythetypeandcombinationofAD medicationtherapy.Thedetrimentaleffectofpooradherenceonhealthcareeconomicoutcomewassignicant,whichimpliesaneedtoimproveadherencein ordertoreducethenancialimpactofnon-adherence.Factorswhichcouldcontributenon-adherenceandnancialburdenneedtoberenedandtargetedby interventiontoimprovehumanisticandeconomicoutcomesoftreatment. PSS26 PATIENT’SEVALUATIONOFTHEQUICKNESSOFACTIONOFGINGIVAL INFLAMMATIONTREATMENTSMattoutP1,RahhaliN2,WattM3,AugesM2,TaiebC2 1GEPI,Marseille,France,2PFSA,BoulogneBillancourt,France,3PFOC,Castres,FranceOBJECTIVES: Gingivitisisdenedaslesionsonthegingivalmargin,expressed throughgumredness,bleeding,localizededema,andgingivalsensitivity.Itismost oftencausedbysubstancesproducedbybacterialplaque,ordentalbiolm,which developsalongthegingivalcrevice.Toevaluate,usingpatient'sinterviewresults, thequicknessofactionofseveraltreatmentsforgingivalinammation. METHODS: Observational,prospective,longitudinal,multicentricstudycarriedout inFrance,usingdatacollectedbyparticipatingdentistsanddentalsurgeons. RESULTS: Atotalof316patientswithgingivitisreturnedtheirquestionnaire.women:65.25%,currentsmokers:22.93%and28.51%wereex-smokers.Lightandheavy bleedingduringbrushingwasreportedby45.74%and33.33%ofpatientsrespectively.45.70%reportedvisibleredness,56.34%reportedswollengums,13.19%had lesions,butaboveall50.52%reportedpain.Finally,62.26%hadprevioushistoryof gingivitissymptoms.Asfordentalsurgeons:96.36%hadperformedscaling,15.15% gingivalcurettage,and13.64%radicularresurfacing.Atotalof78.4%judgedgingivalinammationtobemoderatetosevere,63.10%saidithadspread( 30%).In termsoftreatment:98.62%gavepatientsoralhygieneadvice,87.98%advisedon brushingmethods,69.91%recommendedspecictoothpaste,and78.85%amouthwash.Atotalof30.61%hadgeneralizedinammationafter1month,reducingto just11.24%at2monthsand15.63%at3months.A totalof88.08%reportedimprovementininammationaftertherstmonth,91.59%at2monthsand93.59%at3 months.Atotalof83.93%feltlesspainafter1monthoftreatment,87.90%after2 monthsand92.08%after3months.(p 0.0418).89.13%felttheirtreatmentwaseffectiveafter1month,97.79%after2monthsand96.15%after3months.(p 0.0036). CONCLUSIONS: Intermsofsatisfaction,86.52%weresatisedafter1month,94.85% after2monthsand95.92%after3months.(p 0.0076).87.57%felttheirtreatmentwas easytofollowafter1month,86.76%after2monthsand92.08%after3months.Above all,aftertherstmonthoftreatment,88.83%saidtheywouldcontinueusingthe treatmentinpreventionevenaftercompletedisappearanceofgingivitis. SensorySystemsDisorders–ResearchonMethods PSS27 EFFECTOFTREATMENTSWITCHONTHECOST-EFFECTIVENESSOFBIOLOGICS INPSORIASISINPERUANDCOLOMBIAAlandeteJC Janssen-Cilag,Bogota,ColombiaOBJECTIVES: Toevaluatetheeffectoftreatmentswitchonthecost-effectiveness ofbiologicsusedinpatientswithmoderateorseverepsoriasisinColombiaand Perœ. METHODS: Inapreviousstudy(Alandete,JCacceptedintheISPOR13thAnualEuropeanCongress)costeffectivenessofetanercept,adalimumab, ustekinumabandiniximabwasestimatedbasedonlabelinformationforrst(induction)yearandsecond(maintenance)yearassuminga100%treatment continuation($1USD COL$1.832 SOL$2.75).For etanercepttwoinduction schemeswereconsidered:50mgweekly52weeks-D1-and100mg12weeksfollowedby50mg40weeks-D2-.Effectivenesswasevaluatedas75%reductionin PsoriasisAreaandSeverityIndex-PASI75-iniximab 80%;ustekinumab 69%; adalimumab 59%;etanerceptD252%;etanerceptD1 39%.Iniximaband ustekinumabeffectivenesswerenotsignicantlydifferent.Bothweresignicantly superiortoetanercept(Hawkinsetal.meta-analysispresentedinthe14thInterna tionalISPOR).Inthisabstractwedevelopedanewmodelestimatingswitching probabilitiesduetotreatmentfailureatweek12andadverseevents.Biologicscosts wereadjustedconsideringtimeonthepreandpost-switchingperiods.Treatment effectivenesswasadjustedwhenbiologicswereusedafterswitchingduetotreatmentfailure. RESULTS: IntroductionofswitchingeffectratiedustekinumabdominanceinColombia($US44,675in2years)generatingcostsavingsof-$US4.049 versusetanerceptD1;-$US4.049versusadalimumab;-$US7.844versusetanerceptD2and-$US27.517versusiniximab;withhigherorsameeffectivenessthan theotherbiologicsinthatcountry.InPeru,ustekinumabchangedfrombeingthe mostcost-effectiveoptionandbecamethedominantoption($US41.827in2years) generatingcostsavingsof-$US283versusetanerceptD1;-$US489versusadalimumab;-$US3581versusetanerceptD2and-$US13.499versusiniximab. CONCLUSIONS: Inthestudiedcountriesinclusionoftheswitchingeffectdueto treatmentfailureandadverseeventsratiescost-savingsobservedinColombia andmakesustekinumabthecost-savingoptioninPeru.Theseresultscorroborate thoseobservedintheUSAandEurope. PSS28 EFFECTOFDIFFERENTRECALLPERIODSONDRYEYESYMPTOMRATINGSRuizWM1,LiJZ2,JohnsonME3 1G&SResearch,Inc.,Indianapolis,IN,USA,2Pzer,Inc.,SanDiego,CA,USA,3BristolEye Hospital,Bristol,UKOBJECTIVES: Clinicalstudiesofdryeyedisease(DED),ahighlysymptomaticdisease,oftenaskpatientstoevaluatetheirDEDsymptomsusingpatient-reported outcomesinstruments.Mostoftheseinstrumentsuseaone-weekrecallperiod. TheeffectofthisrecallperiodontheaccuracyofDEDsymptomassessmentshas notbeendocumented.Thepurposeofourresearchwastocompareself-reported DEDsymptomsbetweenone-weekanddailyrecallperiods. METHODS: Weenrolled156DEDpatientstoaweb-basedobservationalstudytoassesstheirDED symptomsonceadayfor9days.Foreachofthe14symptoms,weaskedthe patientstoratethefrequencyandintensityona0-6ratingscale,withahigher scoreindicatingworsesymptom.TheassessmentsonDays1and9hadaone-week recallperiod,whiletheassessmentsonDays2-8hadaone-dayrecallperiod.We thencalculatedthemeanweeklyscoresforDay1andDay9andthemeandaily scoresforDays2-8,andtestedthedifferencesbetweenthemeanweeklyanddaily scoresusingmatched-pairttestswithoutmultiplicityadjustment. RESULTS: The Day1meanweeklyscoresweresignicantlyhigherthanthemeandailyscoresfor all14symptomsinbothfrequencyandintensity.TheDay1meanweeklyscores werealsosignicantlyhigherthantheDay9meanweeklyscoresin10frequency and11intensityscores.TheDay9meanweeklyscoreswereslightlyhigherthan thedailyscores;however,mostofthedifferenceswerenotstatisticallysignicant. CONCLUSIONS: Patients'self-ratingsoftheirDEDsymptomsusingaone-week recallperiodareconsistentlyinatedwhencomparedtotheirratingsusingaonedayrecallperiod.Suchinationshouldbeconsideredwhendesigningclinical studiesforDED. PSS29 DEVELOPMENTOFTHEMODIFIEDOCULARCOMFORTINDEX(MOCI)JohnsonME1,RuizWM2,LiJZ3 1BristolEyeHospital,Bristol,UK,2G&SResearch,Inc.,Indianapolis,IN,USA,3Pzer,Inc.,San Diego,CA,USAOBJECTIVES: Dryeyedisease(DED)ischaracterizedbysymptomsofoculardiscomfort,visualdisturbanceandreducedtolerancetoenvironmentalstressors. DEDhasasignicantnegativeimpactonthequalityoflife(QOL)ofpersonsaffected,andimposesamassiveburdenonmedicalresourcesowingtoitshigh prevalenceandchronicnature.Itisnotknownifavailablepatient-reportedoutcome(PRO)instrumentsfullycapturethescopeofDEDsymptomsandtheirimpact onQOL.ThepurposeofourongoingresearchistodevelopaPROinstrumentthat meetstheneedsofclinicalstudiesinvestigatingpotentialtreatmentsforDED. METHODS: PatientswithDEDinvecountries(UnitedStates,UnitedKingdom, Spain,JapanandKorea)wereinterviewedtoidentifytheirsymptomsandtheimpactofthediseaseonQOL(n 120).Basedontheseresults,itemsweredraftedthat weretestedintwoweb-basedstudieswithmild-moderateDEDsubjects(n 106 and156)andface-to-faceinterviewswithsevereDEDsubjects(n 22). RESULTS: Itemsenquiringabout8additionalsymptomexperiences(16itemsgroupedin doubletsaskingaboutfrequency/intensity)wereaddedtotheoriginalOcularComfortIndex(OCI)usingthesamequestionformatandresponsestructure(uctuatingvision,lightsensitivity,redness,foreignbodysensation,excessivetearing,excessiveblinking,ocularirritationandstickiness).Additionally,2itemsthat enquiredaboutthemostbothersomesymptomandtheextentofbother,and12 itemsthatappraisedhowsymptomsinterferedwiththeabilitytoperformdaily activitieswereincluded. CONCLUSIONS: Patientinterviewssuggestthatavailable PROinstrumentsdonotfullycapturethescopeofDEDsymptomsandtheirimpact onQOL.ThemodiedOCI(mOCI)willbeusedinclinicalstudiestofacilitateits renementandvalidation. PSS30 BURDENOFINFANTILEHEMANGIOMA:DEVELOPMENTOFAQUESTIONNAIRETaiebC1,VoisardJJ2,RuizF3 1PFSA,BoulogneBillancourt,France,2PFD,Lavaur,France,3Clinsearch,Bagneux,FranceOBJECTIVES: Infantilehemangioma(IH)developsduringtherstweeksoflife;it normallyformswithin3to6months,thenregressesveryslowlyoveradurationof 3to7years.Incomplicatedforms,itispossibletoencounterhaemorrhaging, necrosesandulcerations,infectionsand,moreexceptionally,respiratorydistress, cardiovascularshunt.Toexplorethehandicap,initslargestsense,generatedbyIH usingaquestionnairetoexpresstheburdenonthedailylifeoftheparents. METHODS: Thequestionnairewasdevelopedfollowingastrictmethodological process,involvingamultidisciplinaryteamincorporatingvariousplayers(doctors, nurses,socialworkers)whoareinvolvedinthetreatmentofpatientsorwhoare specialisedintheconstructionofquestionnaires.Areviewoftheliteratureand discussionswiththefamilieswereconductedinordertoidentifytheconcepts relatedtothepathology. RESULTS: Exploratoryassessmentsshowedthattheconceptofburdencouldbestructuredaroundtwomainmodules:assesstheimpact directlyfortherst-module.TheconsequencesofHIondailylife,familyand personalrelationships,work,nancialsituationandpsychologicalimpactforthe second-module.Athirdmodulefocusesonthebehaviourofthechild;thismodule willevolveovertimeanddependingontheanalyses.Fifty-sixpreliminaryitems wereidentiedfollowingarstdiscussion.Arstanalysismanagedtoreduce theseitemsto36whilstconservingthe3modulesbutmakingiteasiertousethe analysis. CONCLUSIONS: TheHemangioma-Burden-Questionnairewillallowclinicianstobetterunderstandtheimpactandconsequencesofthepathologyonthe family.Itwillalsoallowthedevelopmentoftheburdentobemonitoredaccording totherateofdevelopmentoftheillnessanditstreatment.ItwillalsoallowtheA58VALUEINHEALTH14(2011)A1–A214

PAGE 67

familiesofthechildrentobetterdefendtheirinterestbeforethehealthauthorities intermsofexpenditure(medicalorother)forwhichthepartremainingtheirresponsibilityisincreasingsignicantly. SystemicDisorders/Conditions–ClinicalOutcomesStudies PSY1 SEVERERENAL,HEPATICANDGASTROINTESTINALEVENTSASSOCIATED WITHDEFERASIROXINPATIENTSWITHTRANSFUSION-DEPENDENTANEMIAHuangWF1,HsiaoFY2,ChouHC3,TsaiYW4,YenHC5,KeWM5 1NationalYang-MingUniversity,Taipei,Taiwan,2NationalTaiwanUniversity,Taipei,Taiwan,3InstituteofHealthandWelfarePolicy,NationalYang-MingUniversity,Taipei,Taiwan,4InstituteofHealth&WelfarePolicy,NationalYang-MingUniversity,Taipei,Taiwan,5Taiwan DrugReliefFoundation,Taipei,TaiwanOBJECTIVES: Ironchelators(deferasiroxordesferrioxamine)areessentialtopatientswhoneedlife-longbloodtransfusion(e.g.-Thalassemia).However,in2010, theUSFoodandDrugAdministration(FDA)hadissuedawarningonpotential adverseeventsassociatedwithironchelators,especiallydeferasirox.Theobjective ofthisretrospectivecohortstudywastocomparetheriskofrenalimpairment, hepaticimpairment,andgastrointestinalbleedinginpatientswithtransfusiondependentanemiausingdeferasiroxordesferrioxamine. METHODS: Patientswith transfusion-dependentanemia(sicklecelldisease,-thalassaemia,myelodysplasticsyndromeandaplasticanemia)andwereprescribedironchelators(deferasirox ordesferrioxamine)wereidentiedfromthe2005 2009Taiwan'sNationalHealth Insurancedatabase.Coxproportionalhazardsmodelswereusedtoassessthe associationbetweenironchelatorsandoccurrencesofadverseevents(renalimpairment,hepaticimpairment,andgastrointestinalbleeding).Allmodelsadjusted forage,sex,drugexposure(days),typeoftransfusion-dependentanemiaandmedicalhistory. RESULTS: Patientswerecategorizedintodeferasirox(n 180),desferrioxamine(n 586),andmixedusers(n 202),basedonthedrugtheyreceivedduringthefollow-up.Thecruderatesofadverseeventswere4.14,3.16and0.65per 10,000person-yearindeferasirox,desferrioxamineandmixusers,respectively. Afteradjustingcovariates,therewasnoassociationbetweendeferasiroxandadverseevents(hazardratio[HR]0.84;95%CI,0.592.00)comparedtodesferrioxamineusers. CONCLUSIONS: Inthispopulation-basedanalysis,transfusion-dependantanemiapatientsusingdeferasiroxanddesferrioxaminewereatsimilarriskof adverseevents. PSY2 THEASSOCIATIONBETWEENTHERAPYWITHANGIOTENSIN-CONVERTING ENZYMEINHIBITORSANDHEMOGLOBINLEVELChodickG1,RazR1,LeshemE2,SteinvilA2,BerlinerS2,ZeltserD2,RogowskiO2, ShalevV1 1MaccabiHealthcareServices,TelAviv,Israel,2TelAvivSouraskyMedicalCenter,TelAviv, IsraelOBJECTIVES: ToassessthehematologicaleffectsAngiotensin-convertingenzyme (ACE-I)inhibitorsandAngiotensinIIreceptorblockers(ARB)inpatientswithout concomitantrenalimpairment. METHODS: Inthepresentretrospectivecohort studyweusedtheMaccabiHealthcareServices'databasetoidentifynewusersof ACE-I(N 14754)ARB(N 751),orcalciumchannelblockers(CCB,N 3087)with availablehemoglobin(Hb)testsbetween2004and2009.Excludedwerepatients purchasingdrugsfrommorethanonemedicationclass,diagnosedwithrenalimpairmentorcancer.MedianHblevelsoneyearbeforeandafterrstmedication purchase(indexdate)werecalculatedandcomparedaccordingtotheproportionof dayscoveredwithmedicationclass. RESULTS: PersistentuseofACE-IandARBwas associatedwithasignicantdecrementinhemoglobinlevel.PatientsatthehighestPDClevelwereatasignicantlyhigherriskofdevelopinganemiaamongACEI (OR 1.59,p 0.001),andARB(OR 2.21,p 0.05).TherelationshipbetweenCCB therapyandHbdecrementwassubstantiallyweaker. CONCLUSIONS: Hblevelsare reducedduringtherstyearofACE-IorARBtherapy.Thisassociationisdosedependentandisnotlikelytobecausedbyartifactsrelatedtopatientadherence. PSY3 OPIOIDSINNON-MALIGNANTPAIN:ARETHEYEQUIVALENTINSAFETY PROFILE?ANETWORKMETA-ANALYSISSiddiquiMK,GuptaJ,BhutaniM,SehgalM HeronHealthPrivateLtd,Chandigarh,IndiaOBJECTIVES: Severenon-malignantpainaffectsalargenumberofpatients.Opioidsareanimportantoptionforanalgesia.However,thereisrelativelylittleinformationaboutthecomparativesafetyofopioids.Wesoughttocomparethesafety andtolerabilityofcommonlyusedopioidsinnon-malignantpainthroughnetwork meta-analysesofrandomizedcontrolledtrials(RCTs). METHODS: Medlineand Embaseweresearchedfrom2000to2011forRCTscomparingcommonlyused opioids(tramadol,oxycodone,hydrocodone,propoxyphene,codeine)innon-malignantpain.Studieswereassessedforinclusion/exclusionbasedonaprespecied protocol.Tworeviewersundertookdataextractionindependently.Anydisagreementwasresolvedbyathirdreviewer.Anetworkmeta-analysiswasusedtocombinedirectandindirectevidenceforsafetyoutcomesreportedinthetrials.Based ontheincidenceofadverseevents(AEs)foreachintervention,aprobability-based ranking(probability(P)ofbeingworst)wasgeneratedusingWinBUGS. RESULTS: Of the1156studies,5RCTsenrolling1399patientswereeligibleforinclusion.The mostcommonlyreportedAEswerenausea,vomiting,somnolence,dizziness, headache,constipationanddrymouth.WithdrawalsduetoAEsweremostcommonlyobservedwithcodeine(P 42%)followedbyhydrocodone(P 28%),tramadol (P 19%),andoxycodone(P 10%).Theprobabilityofoccurrenceofnauseaand somnolencewasthehighestwithcodeine.Dizzinesswasmostfrequentlyassociatedwithoxycodone(P 53%).However,theincidenceofdizzinessandheadache wasthelowestwithcodeine.Tramadolwasobservedtobeassociatedwiththe highest(P 40%)incidenceofvomiting,whilehydrocodonehadthelowestincidence(P 15%). CONCLUSIONS: Codeinewasobservedtohavethehighestincidence ofwithdrawalsduetoAEs.Itwasobservedthattheprobabilityofoccurrenceofany particularAEvariedacrossincludedopioidanalgesics.Codeinewasobservedtohave beenmorefrequentlyassociatedwithnausea/somnolencewhiletramadolandoxycodonehadthehighestincidenceofvomitinganddizzinessrespectively. PSY4 TRENDSINHYPONATREMIAMANAGEMENTANDASSOCIATEDOUTCOMESIN HOSPITALSETTINGS:INTERIMRESULTSFROMANOBSERVATIONAL, PROSPECTIVE,MULTI-CENTER,GLOBALREGISTRYINHOSPITALIZEDPATIENTSDastaJF1,AminA2,ChiongJR3,GreenbergA4,HauptmanPJ5,VerbalisJG6 1OhioStateUniversity,Columbus,OH,USA,2UniversityofCalifornia,Irvine,Irvine,CA,USA,3LomaLindaUniversity,LomaLinda,CA,USA,4DukeUniversity,Durham,NC,USA,5Saint LouisUniversitySchoolofMedicine,SaintLouis,MO,USA,6GeorgetownUniversity,Washington, DC,USAOBJECTIVES: Althoughhyponatremia(HN)isthemostcommonelectrolyteabnormalityinhospitalizedpatients,littleisknownregardingtheinuenceofHNandits managementonpatientoutcomesandhealthcareresourceusage.TheHNRegistry isanovelprospectiveefforttodocumenttheclinicalandhealthcareoutcomesof HNanditsmanagement.Resultsfortherst25HNpatientsenrolledaredescribed here. METHODS: Afterinformedconsentorwaiver,datawereextractedfrommedicalchartsofenrolledpatients.HNwasdenedasaserumsodium 130mmol/L. Thepilotdataweresummarizedappropriatelybysamplesizeandforcategorical databypercentage.SubjectswhohadHNonadmissionwerecategorizedaspreexistingHNpatientsandthosewhowereadmittedforanotherreasonanddevelopedHNwhileinthehospitalwerecategorizedashospital-acquiredHNpatients. RESULTS: Overall,only20%oftheenrolledpatientsreceivedanypharmacologic managementforHNandapproximately44%weredischargedwithpersistentHN (21%oftreatedvs.79%ofuntreated).AmongthepatientsdischargedwithHN,55% hadapreviousepisodeofHN.Inaddition,amongthepatientswithpreviousHN, 46%weredischargedwithpersistentHN.ThelengthofstayforpatientswithpreexistingHNwas1.3dayslongercomparedtopatientswithhospital-acquiredHN. Thesendingswillbefurtherevaluatedandreportedasmoredatainthislarge registrystudyareaccumulated. CONCLUSIONS: Amonghospitalizedpatients,HNis frequentlyuntreated,andnearlyhalfofpatientsaredischargedwithoutnormalizationofserumsodium.HNcommonlypersiststhroughseveralhospitaladmissions. PSY5 ANINVESTIGATIONINTOTHERELATIONSHIPBETWEENOBESITYANDSKIN ANDSOFTTISSUEINFECTIONSREQUIRINGHOSPITALIZATIONSwineyJ UniversityofKentucky,Lexington,KY,USAOBJECTIVES: TheUnitedStatesisexperiencinganobesityepidemicwith67%of adultsbeingeitheroverweightorobese.Whileitisknownthatexcessiveweight increasestheopportunityforskininfections,thisrelationshiphasnotbeenwell studied.Thisstudycontributestotheknowledgeabouttherelationshipbetween thesetwoconditions. METHODS: UsingtheH-CUPnationaldatabaseforinpatient hospitalizations,thisstudyanalyzedthedatafromhospitalsintheSouthernstates forthenumberofskinandsofttissueinfectionsin2003,2005and2007inadults. Theproportionofpatientswhowerealsocodedasobeseinthispopulationwas quantied.Twot-testswereperformedcomparingtheaveragelengthofstayfor patientswhowereobeseandnotobeseandtheaveragetotalhospitalchargesfor patientswhowereobeseandnotobese.Twolinearregressionsanalyzedtheimpactofobesityonthecostofhealthcarebyusinglengthofstayandtotalhospital chargesasdependentvariables. RESULTS: Theproportionofpatientshospitalized forskinandsofttissueinfectionswhowerealsoobeseincreasedfrom48.09%in 2003to51.7%in2007.Theaveragelengthofstaywas7.97daysfornon-obese patientsverses4.6daysforpatientscodedasobesewhichwasstatisticallysignificant.Theaveragetotalhospitalchargeswere$26,653fornon-obesepatientscomparedto$20,876forobesepatients.Thiswasalsostatisticallysignicant.Surprisingly,theco-morbidityofobesityhasanegativepredictivevalueforbothhospital lengthofstayandtotalhospitalcharges. CONCLUSIONS: Itispossiblethatpatients whoareobesearebeingdischargedsoonerbecauseofdifferencesinseverityof infections.Moreresearchisneededtodeterminewhetherobesityisacausalfactor inskinandsofttissueinfectionsandhowthisisaffectingthecostanddeliveryof healthcare. PSY6 TRANSFUSIONALIRONOVERLOAD(TIO)MONITORINGANDTREATMENT: FINDINGSFROMANELECTRONICMEDICALRECORDSREVIEWSTUDYATTHE MOFFITTCANCERCENTERANDRESEARCHINSTITUTEDuhMS1,WetzsteinG2,GuoA3,SasaneM3,SardaSP1,KorvesC1,WangST1,WeiR1, ClintonB1 ,RayL2 1AnalysisGroup,Inc.,Boston,MA,USA,2H.LeeMofttCancerCenterandResearchInstitute, Tampa,FL,USA,3NovartisPharmaceuticalsCorporation,EastHanover,NJ,USAOBJECTIVES: Weexaminedproportionsofpatients:a)monitoredforTIOafter receiving 10, 20,and 30unitsofredbloodcells(RBC)andb)receivingiron chelationtherapy(ICT).Wealsoexaminedoverallsurvival(OS)among:a)monitoredvs.unmonitored;b)ICT-treatedvs.ICT-untreatedgroups. METHODS: Medicalrecordsofpatients 18yearsreceiving 10RBCunits 6monthsbeforedata abstractionwereidentiedattheMofttCancerCenterandResearchInstitute (December2009-June2010).Observationperiodspannedfrom10thRBCunittoend offollow-up(i.e.,death,clinicdeparture,orendofobservationperiod).TIOmoni-A59VALUEINHEALTH14(2011)A1–A214

PAGE 68

toringwasdenedasreceiptof 1serumferritintestafter10thunit.ICT-eligibility wasdenedas 2serumferritintests 1,000g/Lor 20unitstransfused.ICT treatmentwasdenedas 1prescription/administrationfordeferasiroxordeferoxamine.Kaplan-MeierandCoxproportionalhazardsregressionmethodswere usedtocompareoverallsurvival. RESULTS: Atotalof163patientswereincluded; 58.3%and30.1%received 20and 30RBCunits,respectively.Atotalof50.3%had leukemiaand20.2%hadmyelodysplasticsyndromeastheunderlyingdisease.A totalof44.8%,53.7%and67.3%ofpatientsreceiving 10, 20and 30RBCunits, respectively,weremonitoredforTIO.Comparedtounmonitoredpatients,TIOmonitoredpatientshadsignicantlylongermedianOS(16.13vs.2.30months; log-rankp-value 0.001)evenafterconfounderadjustment(hazardsratio0.271; p-value 0.001).Among99ICT-eligiblepatients,9%receivedICT;8(88.9%)received deferasirox.MedianOSforICT-treatedpatientsanduntreatedpatientswas9.48 and7.16months,respectively(log-rankp-value 0.593). CONCLUSIONS: Toconclude,TIOmonitoringafter10RBCunitsmaybeassociatedwithsurvivalbenets, but 50%ofpatientsreceiving 10transfusionsweremonitoredforTIO.Clinical benetsofchelatingcancerpatientshavenotbeenestablishedandadditional studiesarenecessarytoascertaintheimpactofTIOandchelationutilityamong cancerpatients. PSY7 ANOVELMECHANISMOFCAPTURINGPOST-MARKETINGSAFETY INFORMATIONONRECOMBINANTFACTORVIIA(RFVIIA)INTHERARE DISORDERACQUIREDHEMOPHILIA:THEACQUIREDHEMOPHILIA SURVEILLANCE(AHS)PROJECTLentzS1,TandraA2,DoucetteK3,CooperDL3 1UniversityofIowa,IowaCity,IA,USA,2IndianaHemophilia&Thrombosis,Indianapolis,IN, USA,3NovoNordisk,Inc.,Princeton,NJ,USAOBJECTIVES: Acquiredhemophilia(AH)isararedisorder(1per1.3million)characterizedbyauto-antibodiestofactorVIII.AHischaracterizedbylife-threatening bleedingandhighmortality.RecombinantfactorVIIa(rFVIIa)receivedFDAapprovalfortreatmentofAHin2006.Aspartofthepost-approvalcommitment,Novo Nordiskagreedtouseasociety-ownedresearchregistrytomonitortreatmentof AHbleedingepisodes.ThisrequiredIRBapprovalandinformedconsent.Forsites withfewpatientsand/orlimitedresearchstaff,theregistrywasnotafeasible option METHODS: Analternativeweb-based,IRB-exemptreportingmethod,AcquiredHemophiliaSurveillance(AHS,www.novosevensurveillance.com),wasestablishedtocollectdataonrFVIIauseinAHandassociatedadverseeventsunder theHIPAAsafetysurveillancewaiver. RESULTS: FromApril2008-January2011,32 reporterssubmitted80casereports(32male/48female).Themeanagewas65 years(range16-97).Commonassociatedconditionswereautoimmunedisorders (32patients),malignancy(6patients)andpost-partumstate(5patients).Onaverage,5discretebleedingeventsperpatient(range0-100)werereported;75case reportsdescribedbleedingasspontaneous(88%),surgical(16%),and/orrelatedtoa procedure(15%).Nobypassingagentwasreportedtohavebeenusedin17(21%). rFVIIawasusedin50(63%),themajorityofwhichwererst-line(39,78%).AHwas reportedtohavebeen"resolved"in44(55%),"notresolved"in19(24%),and"unsure"in16(20%);mean(median)timetoresolutionwas7.2(2)months(range1-52). Therewerenoreporteddeaths.Noneofthe50rFVIIa-treatedcasessufferedan adverseorthromboticevent(AE). CONCLUSIONS: AHSprovidesaninnovativeapproachforhemophiliatreatmentcentersandhematology/oncologypracticesto capturebasicsafetysurveillancedataforpatientswithAcquiredHemophilia.The AHSprojectprovidesadditionalinformationaboutAHtreatment,andreafrmsthe safetyofrFVIIaandthelowrateofthromboticcomplications. SystemicDisorders/Conditions–CostStudies PSY8 DIRECTANDINDIRECTCOSTSOFPRIVATELY-INSUREDPATIENTSTREATED WITHOXYMORPHONEEXTENDED-RELEASEOROXYCODONECONTROLLEDRELEASETABLETSKirsonNY1,WhiteA1,BirnbaumHG1,SchillerM2,WaldmanT1,Ben-JosephR3, SummersK3 1AnalysisGroup,Inc.,Boston,MA,USA,2AnalysisGroup,Inc.,NewYork,NY,USA,3Endo Pharmaceuticals,ChaddsFord,PA,USAOBJECTIVES: Comparecostsofprivately-insuredoxymorphoneextended-release tablet("oxymorphone")userswiththoseofoxycodonecontrolled-releasetablet ("oxycodone")users,fromanemployer'sperspective. METHODS: Patients,ages 18 ,with 1claimforoxymorphone/oxycodoneduringQ2:2006-Q4:2009were identiedinde-identiedprivatepayerclaimsdataandobservedfromtherst suchclaim("indexdate")untiltheearliestof:useofcomparatordrug;endof continuouseligibility;or12months("studyperiod").Continuouseligibilitywas requiredthroughouta6-monthbaselineperiodand 1monthaftertheindexdate. Patientswithclaimsforanyformulation(extended-releaseorotherwise)ofthe comparatordrugduringtherst30daysofthestudyperiodwereexcluded,aswere patientspregnantduringthebaseline/studyperiods.Studyperiodrisk-adjusted costswereestimatedperpatient-monthusinggeneralizedestimatingequations controllingforbaselinedemographics,resourceuse,andcosts.Medical/drug(direct)costspaidbyprivateinsurerswerereportedforpatientsages18-64(N 8,354) and65 (N 3,515)aswellassubsetsofeachwithoutcancerduringthebaseline/ studyperiods(N 7,090andN 2,444,respectively).Medically-relatedabsenteeismanddisability(indirect)costswerereportedforallemployees,ages18-64,(N 1,313)andemployeeswithoutcancer(N 1,146).Bias-correctedbootstrappingwas usedtocomparethird-partypayercosts. RESULTS: Oxymorphoneusers,ages1864,hadlowerdrugcosts($693vs.$763,p 0.0035)andsimilarmedicalcosts($1,875 vs.$1,976,p 0.3570)perpatient-monthcomparedwithoxycodoneusers.Indirect costswerenotdifferent($662vs.$670,p 0.9370).Oxymorphoneusers,ages65 hadsimilarMedicaresupplementaldrugcosts($533vs.$588,p 0.0840)andsignificantlylowermedicalcosts($459vs.$747,p 0.0001).Resultswerecomparablefor thesubsetswithoutcancer. CONCLUSIONS: Aftercontrollingforbaselinecharacteristics,real-worldevidencesuggeststhat,fromanemployer'sperspective,oxymorphoneusersmayincurlowermonthlydirectcoststhanoxycodoneusers. PSY9 HEALTHCONSEQUENCESANDCOSTSOFTACHOSILASHAEMOSTATIC TREATMENTINLIVERSURGERYVERSUSTISSUCOL:ARETROSPECTIVE OBSERVATIONALCOHORTSTUDYOrtegaA,PardoF Cl’nicaUniversitariadeNavarra,Pamplona,SpainOBJECTIVES: Theobjectiveofthisstudywastocomparetheeffectivenessand costsofTachoSil¨carrier-boundcollagenspongesversusTissucol¨bringluein patientsundergoingliversurgeryincurrentmedicalpractice. METHODS: Aretrospectiveobservationalcohortstudywascarriedoutincluding244patients(128in TachoSil¨groupand116inTissucol¨group)thatunderwentaliversurgerya universityhospital.Dataondemographics,medicalbackground,effectiveness,resourceconsumptionandcomplicationswereobtainedbyreviewingpatients = medicalrecords.Unitcostswereobtainedfromnancialhospitalrecordsanddrug costsfromtheSpanishhealthauthorities.Collectionofpatientdatatookplace fromdenitivesurgeryuntilhospitaldischargeusingahospital = sperspective. Variablescollectedtoestimatecostsandeffectivenesswerehaemostaticdrugs, surgicaltime,drainage,bloodtransfusionsandhospitalisationdays.Researchoutcomeswerepresentedasthedifferencesineffectivenessmeasuresandcostsfor thecollectedvariablesusingstatisticalanalysis(p 0.05). RESULTS: WithTachoSil¨signicantlylesspatientsrequiredadrainage(12.5%vs.52.6%,p 0.001)andpatientswerehospitalizedforaminornumberofdays(median6vs.8days, p 0.00013).Nostatisticaldifferencesbetweenbothgroupswereobservedinthe restoftheeffectivenessdatapre-andpostoperative.TheTachoSil¨groupshowed signicantlowermeancostswithrespecttohaemostaticdrugs(mean[SD]euro, 974 [787 ]vs.987 [384 ],p 0.03)anddrainage(3 [7 ]vs.12 [11 ],p 0.00).No signicantdifferencesinothercostsbetweenbothtreatmentswerereported. CONCLUSIONS: TachoSil¨showedtoreducedrainageandhospitalisationdays. HowevernosignicantlowertotalcostsfortheTachoSil¨group(10,212 [8,197 ]) vs.theTissucol¨group(10,224 [7,394 ])couldbedemonstrated. PSY10 COSTANALYSISOFANEMIATREATMENTWITHERYTHROPOIESESSTIMULATINGAGENTS(ESAS)INCANCERPATIENTSRECEIVING CHEMOTHERAPY:AMULTICOUNTRYAPPROACHDuranA1,SpaepenE2,LamotteM3,WalterE4,LucioniC5,PinheiroB6,BrosaM7, KutikovaL8 ,PujolB8,AnnemansL9 1IMSHealth,London,UK,2SBDAnalyticsBVBA/SPRL,Bekkevoort,Belgium,3IMSHealth, Vilvoorde,Belgium,4InstituteforPharmacoeconomicResearch,Vienna,Austria,5WoltersKluwer HealthItaly,Milan,Italy,6CentrodeInvestiga‹oSobreEconomiaPortuguesa,Lisbon,Portugal,7OblikueConsulting,Barcelona,Spain,8Amgen(Europe)GmbH,Zug,Switzerland,9I-CHER InteruniversityCentreforHealthEconomicsResearch,Gent,BelgiumOBJECTIVES: Spaepenetal.(theOncologist2008;13:596607)publishedacostanalysiscomparingdarbepoetinalfa(DARB),epoetinalfa(EPO-A)andepoetinbeta (EPO-B)inthetreatmentofchemotherapy-inducedanemia,usingpropensityscore matching.ThestudywasperformedusingtheIMSHospitalDiseaseDatabase,a longitudinaldatabaseuniquetoBelgiumcontainingindividualpatient/admissionleveldataondiagnoses,procedures,andpharmaceuticals.GiventhelimitedavailabilityofotherdatabasesreportingthesamelevelofinformationinotherEuropeancountries,theobjectivesofthisstudyweretoassesstheapplicabilityofthe Belgiananalysis,andtoestimatedifferencesincostsbetweenESAsinAustria, Italy,Portugal,andSpain. METHODS: Toadapttheanalysis,costswerereplaced withcountry-speciccostsanddiscrepanciesinepidemiologyandtreatmentpatternswereexamined.Adjustingforcountrydiscrepancies,costswereanalyzed usingamixed-effectsmodelstratifyingforpropensityscorequintilesasinSpaepen2008.SourcesincludedEurostat,nationalcancerregistries,IMSsalesdata, andreimbursementandtreatmentguidelinesforproceduresanddrugs. RESULTS: AllpopulationswerecomparabletoBelgiumintermsofage,gender,ESAuse,and bloodtransfusions.Adjustingforchemotherapyuseandtumor-speciccancer incidence,totalcosts(Euro,2010)withDARBwere22-26%lowercomparedtoEPO-A and20-35%lowercomparedtoEPO-B.Anemia-relatedcostswerelowestforDARB 2,585 154,EPO-A 3,102 99,EPO-B 2,969 166inAustria;DARB 3,144 211, EPO-A 5,049 119,EPO-B 3,656 230inItaly;DARB 2,153 117,EPO-A 2,446 63, EPO-B 2,654 124inPortugal;andDARB 2,378 143,EPO-A 3,349 75,EPO-B 2,857 153inSpain. CONCLUSIONS: Totalandanemia-relatedcostswerelowest inpatientsreceivingDARBcomparedtoEPO-AorEPO-Binallcountries.Although itwasnotpossibletoaccountforalldifferencesamongcountries,thendingsare inlinewiththosefromtheBelgiananalysis,anddemonstratethefeasibilityof adaptingsuchdatatoothersettingsaccountingforpatientcharacteristicsand treatmentcosts. PSY11 BURDENOFILLNESSINFIBROMYALGIASYNDROME:THEPATIENTS’ PERSPECTIVESLeeSS1,KimSH2,KimSK3,LeeCK4,LeeHS5,LeeSH6,ParkYB7,KoSK8,ParkHJ8 1ChonnamNationalUniversityHospital,Gwangju,SouthKorea,2InjeUniversityHaeundaePark Hospital,Busan,SouthKorea,3DaeguCatholicUniversityMedicalCenter,Daegu,SouthKorea,4AsianMedicalCenter,Seoul,SouthKorea,5HanyangUniversityGuriHospital,Gyeonggi-do, SouthKorea,6KonkukUniversityHospital,Seoul,SouthKorea,7SeveranceHospital,Seoul,South Korea,8PzerPharmaceuticalsKoreaLtd.,Seoul,SouthKoreaA60VALUEINHEALTH14(2011)A1–A214

PAGE 69

OBJECTIVES: Incommonwithotherchronicconditions,bromyalgiasyndrome (FMS)adverselyaffectsHealth-RelatedQualityofLife(HRQoL)andfunctionalstatus.FMSoftenresultsindisabilityandprofoundlyaffectstheperformanceofdaily activities,bothatworkandathome.Thisstudywasconductedtoassessthe patient-levelburdenamongsubjectswithFMS. METHODS: Aprospective,observationalstudywascarriedoutonFMSpatientsfrom44medicalcentersthroughout Korea.HRQoLscoreswereobtainedusingtheShortForm(SF)-36HealthSurvey.To estimaterecent3monthhealthcareandnon-healthcarecost,andproductivityloss ofFMSpatients,participantscompletedastandardizedquestionnaire.Costitems mainlyincludedhealthcarecostsuchasoutpatient,pharmacy,inpatient,andorientalmedicine;non-healthcarecostsuchastrafcexpenses,nursingcost,complementaryandalternativemedicine.Costincludedinsurance-coveredcostas wellaspatient'sout-of-pocketexpensesduring3months.Toestimateproductivity lossduetomorbidity,daysawayfromworkduetoFMSwerealsoinvestigated. RESULTS: Among1,040FMSpatientswhocompletedthequestionnaire,female was92%(n 957).Approximately45.7%ofpatientshadrheumaticdiseases.HRQoL physicalcomponentmeanscoresweredeterminedas35.9 7.5,andmentalcomponentmeanscoresweredeterminedas34.6 12.6.Numberofout-patientvisits within3monthswas4.0 6.5.Totalcostsover3monthswere1,609 2,055USD. Within3months,23.4%and91.2%ofpatientswithFMShadbeenawayfromwork andreportedthedecreasedworkproductivity,respectively.Inthepreceding3 months,FMSpatientsmissed20(SD 28.1)daysofwork,leftworkearlyanaverage of28(SD 31.0)days,andaccomplishedlessatworkanaverageof65(SD 30.0) days. CONCLUSIONS: Ourresultssuggestedthatthepatient-levelburdenamong FMSsufferersinKoreaissignicant,evidencedbynegativeeffectonhealth-related qualityoflife,healthresourceuseandwork/activitylimitations. PSY12 BASELINECHARACTERISTICSANDPRE–INDEXTOTALHEALTHCAREAND RHEUMATOIDARTHRITIS-RELATEDCOSTSINPATIENTSRECEIVING GOLIMUMABTHERAPYCarterC1,TandonN1,SmithD2 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2IMSHealthIncorporated, Watertown,MA,USAOBJECTIVES: GolimumabwasrecentlyapprovedbytheFoodandDrugAdministrationonApril24,2009foruseinpatientswithrheumatoidarthritis(RA),psoriatic arthritis(PsA),orankylosingspondylitis(AS).Theobjectiveofthisstudywasto comparebaselinedemographicandclinicalcharacteristicsandpre-indextotal/RArelatedhealthcarecostsinpatientsreceivinggolimumabwith/withoutbiologic experience. METHODS: TheIMSLifeLinkHealthPlandatabasewasutilizedto identifypatientswithan:age 18yearsatindex;indexgolimumabpharmacy claimstarted4/24/2009-10/31/2009; 1RA,PsA,orASICD-9diagnosiscode;and24 monthspre-and 2monthspost-indexcontinuousenrollment. RESULTS: Atotal of282patients(n 206withpre-indexbiologicexperience)receivinggolimumab wereidentied;mean SDagewas52 12;72%female.Nosignicantdifferences indemographiccharacteristicsexistedbetweenpatientswith/withoutbiologicexperience.ThetopthreerheumatologicconditionswereRA(79%),PsA(26%),andAS (10%).Hypertension(45%),dyslipidemia(40%),anddiabetes(17%)werethemost prevalentcomorbiditiespriortostartinggolimumab.Pre-indextotalhealthcare costs(mean,median)weresignicantlyhigherforbiologicexperiencedpatients ($59,979,$46,075)comparedtonon-biologicexperiencedpatients($21,002,$12,662; p 0.001).Amongthe178patientswithanRAdiagnosisonly,pre-indextotalRArelatedcosts(mean,median)weresignicantlyhigherforpatientswithbiologic experience($29,701,$27,464)thanthosewithout($6,045,$1,937;p 0.001).Overall mean SDRA-relatedmedicationcosts($17,193 $16,619)weregreaterthan RA-relatedoutpatientorinpatientcosts($5,239 $14,912and$225 $3,001, respectively). CONCLUSIONS: Whenexaminingbaselineclinicalcharacteristicsof patientsreceivinggolimumab,resultsshowedthatthemajorityofpatientswere diagnosedwithRA,andhadconcomitanthypertension,dyslipidemia,ordiabetes. Mostpatientsreceivedanotherbiologicpriortogolimumab.Biologicexperienced patientsweremorecostlyatbaseline.Thesecharacteristicsshouldbeadjustedfor incomparativeanalysesincludinggolimumab. PSY13 THECOSTOFOBESITYANDITSMAJORCOMORBIDITIES:ANALYSISFROMA COMMERCIALCLAIMSDATABASEPadulaWV,NairKV UniversityofColoradoHealthSciencesCenter,Aurora,CO,USAOBJECTIVES: Todeterminepaymentsmadebycommercialhealthcareproviders foradultsdiagnosedwithobesity,andthosecomorbidwithanycombinationof certainchronicconditions:diabetesmellitus(DM),hypertension,depression, and/orcongestiveheartfailure(CHF). METHODS: Weutilizeda10%randomsamplingfromMarketScan,acommercialclaimsandencountersdatabase (n 12,416,190).Thestudypopulation(n 50,717)waslimitedtothosewho:(1)were adultsage18-64;(2)ledclaimsbetween2006-2007;(3)hadatleastoneinpatient visit,oneoutpatientvisit,oroneemergencydepartmentvisit;and,(4)weregivena primaryorsecondarydiagnosisofobesity.Personswereidentiedandcategorized iftheyhadoneormorecomorbiddiagnoses(DM,hypertension,depression,and/or CHF)inadditiontoobesity.Afteradjustingforageandgender,wecalculatedthe meantotalnetexpenditures(in$US2007)foreachcombinationofcomorbidconditions.AllcalculationswereperformedinStata. RESULTS: Amongthosediagnosedwithobesity,themeannetexpendituresforserviceswere$1799perpatient. Personsdiagnosedwithobesityandothercomorbiditiesobservedanincreasein totalnetexpenditures.Obesityandhypertensionobservedthehighestincrease amongsinglecomorbiditiesat$4298.Forpersonswithobesityandtwoothercomorbidities,DManddepressionwasthehighestat$14364.Themostexpensive conditioninthestudysamplewasobesity,DM,hypertensionanddepressionat $14843.Allresultswerestatisticallysignicantatthe95%condencelevel. CONCLUSIONS: Comparedtotheaveragemedicalclaim,personsdiagnosedwith obesityalongwithothercommonchronicconditionsexperiencesignicantincreasesinhealthcarecosts.Thesecostsareoftendrivenhigherbythetimespentas inpatients.Inmanycases,obesitymaybethecauseofotherchronicconditions thatresultinthesehighcosts.Thus,bycontrollingandreducingtheprevalenceof obesity,wemayseesignicantdecreasesinhealthcareexpenditures. PSY14 ECONOMICIMPACTOFPOTENTIALDRUG-DRUGINTERACTIONSAMONG PATIENTSTAKINGOPIOIDANALGESICSSummersKH1,OhsfeldtR2,PuenpatomRA1,RajanN1,Ben-JosephR1 1EndoPharmaceuticals,Inc.,ChaddsFord,PA,USA,2TexasA&MHealthScienceCenter,College Station,TX,USAOBJECTIVES: Patientsmanagingchronicnon-cancerpainwithcytochromeP450 (CYP450)-metabolizedopioid(codeine,fentanyl,hydrocodone,methadone,oxycodone,tramadol)analgesicswhoconcurrentlytakeanotherCYP450-metabolized medicationexperienceadrug-drugexposure(DDE),whichputsthematriskfora pharmacokineticdrug-druginteraction(DDI).Thisstudyexaminedwhetherpatients withanincidentDDEwiththepotentialtocauseaDDIhadgreaterhealthcarecosts comparedtosimilarpatientswithoutsuchexposure. METHODS: Propensityscore matchingwasusedtocontrolforbaselinedifferencesinaninsuredpopulationchronicallyusingtheseopioidanalgesicsduringtheperiod,January1,2004throughDecember31,2008.EstimatesofthepredictedlikelihoodofaDDEoccurrenceforthematchingmodelwerederivedfrompatientage,gender,geographicallocation,numberof uniqueconcurrentmedications,co-morbidities,servicesutilization,andtotalcostsin thesix-monthspiortotheindexdate.Comparisonsweremadebetween122,586DDE and122,586no-DDEpatients. RESULTS: Comparisonsyieldedmeantotalcostssix monthsafteranincidentDDEthatweresignicantlyhigherforyounger( 65years old)patientswithDDEversusmatchedno-DDEpatients($8,165vs.$7,498,respectively,resultinginadifferenceof$667,p 0.01).Similarlyamongolder( 65yearsold) patients,meantotalcostsatsixmonthsweresignicantlyhigherforpatientswith DDEcomparedtomatchedno-DDEpatients($9,598vs.$9,030,respectively,resulting inadifferenceof$568,p 0.01).Thedirection,magnitudeandsignicanceofthese differencespersistedaftersensitivityanalyses. CONCLUSIONS: Althoughitisimpossibletoestablishcausalrelationshipsinclaimsdatabasestudies,thisstudydemonstratesastrongassociationoftheeconomicconsequencesofDDEsthatareavoidable. SinceconcurrentexposuretomultipledrugsmetabolizedthroughtheCYP450enzyme systemmayberelativelycommon,policydecisionsshouldincludeconsiderationof theuseoflong-actingopioidsthatarenotmetabolizedthroughtheCYP450pathway. PSY15 COSTSOFCAREFORPRIVATELYINSUREDMALESWITHHEMOPHILIAINTHE UNITEDSTATES,2008GuhS1,GrosseS1,McAlisterS1,KesslerCM2,SoucieM1 1CentersforDiseaseControlandPrevention,Atlanta,GA,USA,2GeorgetownUniversity, Washington,DC,USAOBJECTIVES: Althoughhemophiliamayhavealargeeconomicimpactonpatients andhealthcaresystems,publishedestimatesofcostsforcareofU.S.hemophilia patientsaresparse.Theobjectiveofthisstudyistoestimateaverageannualcosts ofhealthcareforhemophiliapatientsintheUnitedStates,stratiedaccordingto theinuenceofage,typeofhemophilia[A(factorVIIIdeciency)versusB(factor IX)],presenceofneutralizingalloantibodyinhibitors,andexposuretoblood-borne viralinfections. METHODS: DatafromtheMarketScan¨CommercialandMedicare ResearchDatabaseswereusedfortheperiod2002-2008toidentifycasesofhemophiliaandtoestimatemedicalexpendituresduring2008. RESULTS: Atotalof1,164 malehemophiliapatientswereidentied,933withhemophiliaAand231with hemophiliaB.Theiraverageannualcostsforhealthcarein2008were$155,136. Averageannualcostsfor30(3%)hemophiliaApatientswithaninhibitorwere5 timeshigherthanforotherhemophiliapatients,approximately$697,000and $144,000,respectively.Clottingfactorconcentrateaccountedfor70%82%oftotal costs.Averagecostsfor207adultpatientswithHCVorHIVinfectionwere1.5times higherthanthoseforadultswithoutinfection.Asubsetofadultswithviralinfectionhadparticularlyhighuseoffactorconcentrateandcosts. CONCLUSIONS: Hemophiliatreatmentisassociatedwithasignicanteconomicburden,particularlyforpatientswithneutralizingalloantibodyinhibitorswhorequirebypassing agentsfortreatment.Theexcesscostsofcareassociatedwithblood-borneviral infectionareconcentratedamongasubsetofhemophiliapatients. PSY16 PROJECTINGTHELIFETIMEECONOMICCOSTSOFOBESITY:ABMI TRAJECTORY-BASEDMODELWangBCM1,GarrisonL1,Alfonso-CristanchoR1,WongE2,FlumD1,ArterburnD3, SullivanSD1 1UniversityofWashington,Seattle,WA,USA,2DepartmentofVeteransAffairs,Seattle,WA, USA,3GroupHealthResearchInstitute,Seattle,WA,USAOBJECTIVES: Obesity,denedasbodymassindex(BMI) 30,isamajorcontributor toincreasedmorbidity,mortality,andhealthcareexpenditures.Thereisatendencyforobeseindividualstocontinuallygainweightovertheirlifetimes.However,neitherthestrengthofthistendencynoritseconomicconsequencesarewell understood.Weconstructedahealthoutcomesmodelfromasocietalperspective toassessthis. METHODS: OurmodelprojectedlifetimetrajectoriesforBMI,death, cost,andquality-adjustedlifeyears(QALYs).Usingtheseprojections,wecomputedtheneteconomicvalueofpotentialweightlossforanindividualgivenbase-A61VALUEINHEALTH14(2011)A1–A214

PAGE 70

linecharacteristics:Age,Sex,andBMI.Weusedexistingliteratureandpublicdatasetstoestimatetheparametersofthemodel.FortheBMItrajectory,weutilized existingliteratureonpooleddatabasesoftheindividuals'BMIovertime.Forthe probabilityofdeath,weconstructedlifetablesbyestimatingthe5-yearprobability ofdeathusingalogisticregressionmodelusingdatafromtheNationalHealth InstituteSurveylinkedwithNationalDeathIndex(NHIS-NDI)between1997and 2005.ForcostsandQALYs,weuseMedicalExpenditurePanelSurveyPanels6-10to estimatehedoniclinearregressionsineachperiod.CostsandQALYswerediscountedat3%. RESULTS: Inthebasecasefora45year-oldfemaletheexpected ageofdeathforbaselineBMIof25,35,or45was83,80,and77,respectively.The projecteddifferenceindiscountedlifetimehealthcarecostsbetweenthisnonobese(BMI 25)personandsomeoneBMI 45isabout$26,000.IfthelossinQALYs werevaluedat$100,000/QALY,theneteconomicvaluelossisprojectedat$271,000. CONCLUSIONS: Obesityisassociatedwithhighermedicalcosts,lowerqualityof life,andreducedlifeexpectancy.Thesocietalcostofdeliveringeffectiveweight lossinterventionstoobeseAmericansshouldbeconsideredinthecontextofthese lifetimeoutcomes. PSY17 APHARMACOECONOMICEVALUATIONOFROMIPLOSTIM(NPLATE)FORTHE TREATMENTOFCHRONICIMMUNETHROMBOCYTOPENIA(ITP)INMEXICOArochoR1,NorthridgeK2,RiveraHurtadoR3,Garc’aCh‡vezJ4 1Amgen,Inc.,Barcelona,Spain,2OutcomesInsights,Inc.,WestlakeVillage,CA,USA,3Amgen Mexico,MexicoCity,Mexico,4MexicanInstituteforSocialSecurity(IMSS),MexicoCity,MexicoOBJECTIVES: ITPischaracterizedbyreducedplateletcountsandincreasedriskof bleeding.Romiplostim,arst-in-classthrombopoietinmimetic,safelyincreases andsustainsplateletcountsinmostadultpatientswithchronicITPforaslongas needed,whilereducingtheneedforconcurrentandemergencymedications.We evaluatedtreatmentcostsperoverallplateletresponsewithromiplostim concurrenttreatmentvs.placebo concurrenttreatmentinchronicadultITP,froma PublicMexicanHealthcareperspective. METHODS: Overallresponse,denedas 4weeklyplateletresponses( 50x109/L)fromweeks2to25,wasderivedfromtwo randomizedparalleltrialswithsplenectomisedandnon-splenectomisedpatients overperiodof24weeks.AllpatientswereallowedtoenteronconcurrentITP medication(danazol,corticosteroids,azathioprine)andreceiverescuemedication (eg,intravenousimmunoglobulin).Treatmentcostsincludedintervention,rescue medicationandmanagementofbleeding-relatedeventsduringoneyearperiod. Unitarycostswereobtainedfromthe2010OfcialPriceListofthePublicHealthcareSysteminMexico($MXP).Meantreatmentcostperresponsewascalculated forsplenectomizedandnon-splenectomizedpatients. RESULTS: Costperresponse waslowerforromiplostimcomparedtoplacebo.Overallresponserateswere79% forromiplostimand0%forplaceboinsplenectomizedpatientsand88%forromiplostimvs.14%forplaceboinnon-splenectomizedpatients.Meantreatmentcosts wereMXP$574,580forromiplostimandMXP$301,218forplaceboinsplenectomizedpatientsandMXP$402,083forromiplostimandMXP$180,692forplaceboin non-splenectomizedpatients.CostperresponsewereMXP$727,317forromiplostimandinnitecostperresponseforplaceboinsplenectomizedpatientsand MXP$456,912forromiplostimandMXP$1,290,655forplaceboinnon-splenectomizedpatients.Themaincost-offsetswereduetoreducedimmunoglobulinrescue use. CONCLUSIONS: GiventhelimitednumberofeffectivetherapiesinITP,romiplostimdemonstratesanimportantandcost-efcientoptionforbothnon-splenectomisedandsplenectomisedpatientsfortheMexicanPublicHealthcareSystem. PSY18 ADECISIONANALYSISMODELEXPLORINGTHERESULTSOFAPHASEIITRIAL OFELTROMBOPAGFORPATIENTSWITHCHRONICHEPATITISC,CIRRHOSIS ANDTHROMBOCYTOPENIASchelfhoutJ,KaufT UniversityofFlorida,Gainesville,FL,USAOBJECTIVES: ThrombocytopeniamaypreventchronichepatitisC(CHC)patients frominitiatingormaintaininganti-viraltherapy.Eltrombopagisathrombopoietin receptoragonistcurrentlyundergoingphaseIIItrialsfortreatmentofCHC-related thrombocytopenia.Thisstudyexaminesthecost-effectivenessofeltrombopagin CHCpatientswithcirrhosisandthrombocytopenia. METHODS: Atwo-partmodel wasdevelopedtoestimatequalityadjustedlifeyears(QALYs)gainedwithand withouteltrombopagforahypotheticalpopulationof50-year-old,white,maleCHC patientswiththrombocytopeniaandcirrhosiswhoareotherwiseeligibletoinitiate anti-viraltherapy.Adecisiontreefollowedtherst48weeksofanti-viraltherapy forpatientsreceivingeltrombopag(30,50,or75mgdose)versusaplacebocontrolled,wait-and-see(W&S)strategy.Lifetimeoutcomesofanti-viraltherapysuccessorfailurewereprojectedusingaMarkovmodel.Eltrombopagprobabilities weretakenfromaPhaseIItrial;otherprobabilitiesandcostswerederivedfrom publishedliterature.Modeloutcomesarepresentedasincrementalcosteffectivenessratios(ICERs).Aprobabilisticsensitivityanalysiswasperformedtoanalyze parameteruncertainty. RESULTS: Successiveeltrombopagdosingstrategieswere associatedwithincrementalincreasesincostandQALYs.Allstrategieswerecosteffectiveatathresholdof$50,000/QALYgained.The75mgstrategyweaklydominatedthe50mgand30mgstrategiesandhadanICERof$27,496comparedtothe W&Sstrategy.Resultswererobusttovariationamonginputparameters. CONCLUSIONS: Theresultsofthisstudysupporttheuseofeltrombopagasa cost-effectiveoptioninCHCpatientswiththrombocytopenia.Patientsoneltrombopagaremorelikelytocompleteantiviraltherapy,leadingtoincreasedqualityof lifeanddecreasedfuturecostsoveralifetime.Morestudiesareneededtoanalyze thesafetyofeltrombopagandtoprovidemoreaccurateestimatesofitseffecton plateletcountsandpatientwithdrawalfromantiviraltherapy. PSY19 THECOST-EFFECTIVENESSOFBORTEZOMIBPLUSMELPHALANAND PREDNISONEVERSUSLENALIDOMIDEPLUSMELPHALANANDPREDNISONE WITHCONTINUOUSLENALIDOMIDEMAINTENANCETREATMENTFORTHE INITIALTREATMENTOFMULTIPLEMYELOMAINTHEUNITEDSTATESWangST1,HuangH2,Ba-manciniA2,ShiH2,ChenK1,KorvesC1,DhawanR3,CakanaA4, vandeVeldeH5 ,DuhMS1 1AnalysisGroup,Inc.,Boston,MA,USA,2MillenniumPharmaceuticals,Inc.,Cambridge,MA, USA,3Johnson&JohnsonPharmaceuticalServices,LLC,Raritan,NJ,USA,4Janssen-CilagPtyLtd, Saunderton,HighWycombe,UK,5Johnson&JohnsonORD,Beerse,BelgiumOBJECTIVES: Thisstudyaimedtoassessincrementalcost-effectivenessofVMPvs MPR-Rinpatientsnewlydiagnosedwithmultiplemyeloma(MM),ineligiblefor ASCT,withaverageageof70years. METHODS: AMarkovmodelfromUSpayer's perspectivewasdevelopedtoassesscost-effectivenessofVMPvsMPR-RfortreatmentofMM.Themodelincludedsevenhealthstates:responses(multiple),treatment-free/maintenance,progression,second-linetreatment,anddeath.Monthly transitionprobabilitieswereestimatedfrompatient-leveldata(VISTA)forVMP andMPandfrompublisheddata(MM-015)forMPR-R.Costsincludeddrug,medical, adverseevent,second-linetreatment,andresourceutilizationin2010US-dollar value.State-specicutilitieswerederivedfrompatient-levelEQ-5DusingUS-specicweights.EffectivenesswasexpressedinLYsandQALYs.Costsandeffectivenesswerediscountedat3%.ICERwascalculatedforVMPvsMPR-Rover20-year horizon(lifetime).Inthebasecase,theMPR-RvsMPhazardratio(HR)forPFSwas 0.499andthatforOSwas1,asthesurvivalbenetwithMPR-RvsMPwasnot observedinMM-015.One-waysensitivityanalyseswereconductedforkeyparameterstoassessthegeneralrobustnessofthemodel. RESULTS: Base-caseresultsfor VMPvs.MPR-Rshowed:$119,102vs$248,358fordirectmedicalcosts;4.187yearsvs 3.409yearsforLYs;costofRmaintenancewas$107,047.VMPwasassociatedwith reducedcostsandbetteroutcomescomparedwithMPR-R;VMPcostsapproximately50%lessthanMPR-RandseemstoprovideslightlymoreQALYs(0.567)on average.SensitivityanalysessupportedtherobustnessofmodelndingsandidentiedtheMPR-RvsMPHRforOSasakeydriver;onlywhenthisHRwas 0.24did MPR-Rbecomecost-effectivevsVMPat$100,000perQALY. CONCLUSIONS: VMP hadlowercostsandbetterhealthoutcomescomparedtoMPR-R.Rmaintenance showedlittleadditionalbenet. PSY20 ECONOMICEVALUATIONOFSEQUENTIALANALGESICTREATMENTINTHE MANAGEMENTOFMODERATEANKLESPRAININMEXICANADULTSArreola-OrnelasH1,Rosado-BuzzoA2,Garcia-MollinedoM2,Muci–o-OrtegaE3, Mould-QuevedoJF4 ,Galindo-SuarezRM3 1Fundaci—nMexicanaparalaSaludAC,MexicoCity,Mexico,2Links&LinksS.A,deC.V.,Mexico City,Mexico,3Pzer,MexicoCity,Mexico,4Pzer,NewYork,NY,USAOBJECTIVES: Anklespraincauseswalkinglimitation,painandconsequently,economiclosses.Thepurposeofthisstudywastoestimatethecost-effectiveness fromasocietalperspectiveofsequentialuseofparecoxib/celecoxibforpainmanagementassociatedwithanklespraininadults. METHODS: AMarkovmodelwas performedtoestimatehealthandeconomicconsequencesduringatimehorizon of45days(3-daycycles).Effectivenessmeasureswere:percentageofpatientsreportingpainreduction,percentageofpatientsreporting 50%painreductionfrom baseline,reductionoflengthofstay(LOS)andreductioninthenumberoftreatmentretirementsdueadverseevents.Transitionprobabilitieswereobtainedfrom ameta-analysisemployinginternationalpublishedliterature.Dosesofcomparatorswere:diclofenac(75mgbid)followedbydiclofenac(100mgbid);parecoxib (40mgbid)followedbycelecoxib(200mgbid)andketorolac(30mgqid)followedby ketorolacpo(10mgtid)asthereference.Parenteralandoralformswereadministeredfortwoandvedays,respectively.ResourceusewasobtainedfromSocial SecurityMexicanInstitutedatabases(n 1,395records).Directcostswereextracted frominstitutionalofcialsourcesandindirectcostsfromavalidatedsurveyappliedtopatients.Costsincluded:hospitalization,drugs,medicalprocedures,imaging,adverseeventsmanagement,disabilitybenets,productivitylosses,and out-of-pocketexpenses.Probabilisticsensitivityanalyseswereperformedemployingbootstrappingtechniques.Acceptabilitycurveswereconstructed. RESULTS: Parecoxib/celecoxib,ketorolacanddiclofenaccostsperpatientwere:US$440.77 [95%CIUS$425.65-US$455.88],US$526.08[US$509.97-US$542.20]andUS$815.43 [US$790.18-US$840.68],respectively(p 0.05).Parecoxib/celecoxibexhibitsthelowestLOS(0.69days[0.670.71](p 0.05))andnumberoftreatmentdiscontinuations (16/1000patients).Acceptabilitycurvesshowedthatparecoxib/celecoxibwillbe cost-effectivewith90%ofcondenceatawillingnesstopayclosertoUS$0. CONCLUSIONS: Parecoxib/celecoxibisthelesscostlytreatmenttomanagepain associatedwithmoderateanklespraininMexico;aswellitrepresentsacostsavingalternativeinLOSreductionandtreatmentdiscontinuationregardingcompetingalternatives. PSY21 COST-EFFECTIVENESSANALISYSOFTREATMENTWITHAMFEPRAMONE (DIETHYLPROPION)INOBESITYINMEXICORamirezRamirezM1,Soto-MolinaH1,Rizzoli-CordobaA2,Delgado-GinebraI3, Pizarro-CastellanosM2 1ItelinessConsulting,MexicoCity,Mexico,2HospitalInfantildeMŽxicoFedericoG—mez,Mexico City,Mexico,3PanamericanUniversity,MexicoCity,MexicoOBJECTIVES: ThetreatmentguidelinesofobesityattheMexicanInstituteforSocial Security(IMSS)considerthepharmacologicaltreatmenttogetherwiththechange inhealthhabitsaproperoptionincasesofobesepatients(BMI 30kg/m2).With thewithdrawalofSibutramineofthemexicanmarket,Diethylpropion(Amfepramone)emergeasagoodcandidatetollthevoidofaschemetocontrolandreduceA62VALUEINHEALTH14(2011)A1–A214

PAGE 71

weight.Themainobjectiveistoconductapharmacoeconomicanalysistondout thecosteffectivenessofDiethylpropionwithdietandexercise.(DEP DaE),comparedagainstDietandExercise(DaE)inMexico. METHODS: Thepointofviewofthe studywasthesupplierofIMSShealthservices.Thetargetpopulationweremen andwomenover18yearswithBMI 30kg/m2.Outcomemeasureswerethereduc tionofweightinkgandQualityAdjustedLifeYears(QALYs).Thedirectcostsinthe treatmentofobesitywereassessed,treatmentofadverseeventsandcomplications(Type2DiabetesMellitusandcardiovasculardisease).WeusedaDynamic, StochasticDiscreteModel(DSD)andaunivariateandprobabilisticsensitivityanalysiswasperformed.AllthequantitiesexpresedinMexicanpesos(MXP) RESULTS: DEP DaEpresentedalowercostandimprovedtheutilityandeffectivenesswhen comparedDaE.Incrementalcostwas$18,361MXPinmalesand$16,285MXPin women.Incrementaleffectivenessandutilitywas4.19kgand0.10QALYsinmen and3.77kgand0.08QALYsinwomen.ICERpointedtheabsolutedominancefor DEP DaE.Estimatedsavingsper100patientsintheIMSScanbe$1,697,027MXPif complementedbya100%changehealthhabitsusingDEP DaE. CONCLUSIONS: ThecombinationofDEP DaEprovidesacosteffectiveimprovementtothetreatmentofpatientswithariskproleforobesity. PSY22 ANALYSINGTHEBENEFITSANDCOSTSAVINGOFORALVERSUSIV FLUDARABINEFORMANAGEMENTOFB-CELLCHRONICLYMPHOCYTIC LEUKAEMIA(CLL)ZinzaniPL1,ConnorsJM2,KnightAD3,CroweAM3,JonesR4,TuckerB5 1UniversityofBologna,Bologna,Italy,2BCCACentreforLymphoidCancer,Vancouver,BC, Canada,3CorvusCommunicationsLimited,Buxted,UK,4GenzymeTransplantandOncology, Oxford,UK,5GenzymeOncology,SanAntonio,TX,USAOBJECTIVES: Fludarabine(F)isamainstaytreatmentforCLL.Despitetheavailabilityofanoralformulationwithequivalentefcacyandimprovedpatientacceptability,manycentresuseanIVgenericformulationassumingitischeaper.A cost-minimisationanalysiswasundertakenonthetwoformulations. METHODS: ThemodelassessedoralvsIVFandcombinationwithoral/IVcyclophosphamide (FC).Itincluded:Europeanacquisitionvaluesdefaultedtogenericprices;body surfacearea1.75m2;mg/m2dosagesoralF40,IVF25,oralC150,IVC250;dosage days/cycl eF5/FC3for6 cycles/patient.Publishedadverseeventsrateswere equivalentexceptforIVadministrationcomplications(8%default)anddiarrhoea (oralF34.6%vsIVF11.3% Grade1/2and3.8%vs0% Grade3/4) 2%managed inhospital.Equipmentandclinicalresourcecostsweredefaultedtopublished ratesandauthors'centres.Asensitivityanalysisassessingminimumandmaximumpotentialcostswasapplied. RESULTS: Acquisitioncostspertreatment coursewerehigherfororalvsIV(F5,334vs 1,313FC3,213.16vs 894).However IVcostsincreasedwithadverseevents(oralvsIVcomplicationsmean 0vs 40, diarrhoea 1vs 0)andhospitalresourcecosts(oralvsIVF18vs 4,200,FC18 vs 2,520).DirectoralvsIVtreatmentcostsperpatientwereF5,353vs 5,553 (range 5,3345,353and 5,5135,577),FC3,232vs 3,454(range 3,2143,233 and 3,4143,478).Modellingoraladoptionin100patients-80%I VF/20%IVFC,a 100%and90%switchrespectivelyresultedin 19,943meancostsavings,releasing fundingandresourcesforimprovedcareandpatientthroughput. CONCLUSIONS: OraludarabineacquisitioncostsarehigherthangenericIV,butcost-minimisationmodellingdemonstratedreduceddirectcostswithoraludarabine,whilealso beingpreferredbypatients. PSY23 ECONOMICEVALUATIONOFERYTHROPOIESISSTIMULATINGAGENTSIN CRITICALLYILLTRAUMAPATIENTSChuiBKY,HazelM,DongJ,KlarenbachS UniversityofAlberta,Edmonton,AB,CanadaOBJECTIVES: Recentrandomizedtrials(RCT)havesuggestederythropoiesisstimulatingagents(ESA)reducemortalityincriticallyilltraumapatients;howeverESA arecostly.Wesoughttodeterminecost-effectivenessofESAinthispatient population. METHODS: Adecisionanalyticmodelwasconstructedtocomparethe useofESAtostandardcareintraumapatientsadmittedtoanintensivecaresetting.BasecasecostsandbenetsatoneyearwereestimatedusingmortalityestimatesfromavailableRCTs.Onewayandprobabilisticsensitivityanalyseswere conductedforcomparisonofthebasecasescenariowith10and25yeartime horizonsinMarkovmodels. RESULTS: ESAusewasassociatedwithacostperQALY gainedof$74,500to$81,748comparedwithstandardcareatoneyear.One-way sensitivityanalysesindicatedresultsweresensitivetochangesinmortalityrisk, riskofthrombosis,relativeriskofmortality,relativeriskofthrombosis,andquality oflifeestimates.CosteffectivenessacceptabilitycurvesgeneratedfromprobabilisticsensitivityanalysisindicatedthattheprobabilityESAwouldbeconsidered attractiverangedfrom35%to80%overtherangeofWTPof$60,000to$120,000. ConsiderationoflongertimehorizonsreducedthecostperQALYgainedto$9,338 $5,957. CONCLUSIONS: WhilethecostperQALYgainedwithESAusefallsintothe rangeofcurrentlyacceptedhealthcaretechnologiesfundedinCanada,signicant uncertaintyexits,particularlylong-termsurvivalbenetwithESAuse.Further researchintotheefcacyandsafetyofESAuseincriticallyilltraumapatientsis requiredpriortowidespreaduse. PSY24 COST-UTILITYANALYSESOFDIFFERENTERYTROPOIETINSINANEMIC PATIENTSONDIALYSISINRUSSIAVorobyevP1,BezmelnitsynaL1,BorisenkoO1,EliseevaE2,BaevV3,VezikovaN4 1RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,Russia,2Vladivostok StateMedicalUniversity,Vladivostok,Russia,3AbakanCityHospital,Abakan,Russia,4PetrozavodskStateUniversity,FacultyofMedicine,Petrozavodsk,RussiaOBJECTIVES: Toexaminecost-utilityofdifferenterythropoietins(epoetinalfaand epoetinbeta)inRussiananemicpatientsondialysis. METHODS: Healthsurveyof patientsindialysisdepartmentsof23regions.About1400questionnaireswere distributedinMay-June2010.Health-relatedqualityoflifewasassessedwith self-administratedvalidatedversionofRussianversionofEuroqol-5Dquestionnaire.Inclusioncriteriawereageolderthan18years,terminalstageofchronic renalfailureandtreatmentwithdialysismorethen4months.Allpatientshave beendividedonfoursubgroupsdependingonthenameofreceivederythropoietin. Directcostcontainsofcostoferythropoietin,administrationcost,costofiron supplementionwascalculated. RESULTS: 1395completedquestionnaireswerereceivedbySeptember,2010.Detailedanalysisofantianemictherapyhasshownthat 85%patientsreceivederytropoetins.Therewerefourthemostuseddrugs:Recormon(epoetinbeta)(42%),Eralfon(epoetinalfa)(18%),Epocrinum(epoetinalfa) (16%),Eprex(epoetinalfa)(9%).Groupswerenotdifferaccordingtothelevelof hemoglobin.AverageHRQL,assessedwithVAS,wasnotdiffersignicantlybetween4groups:0.62forEralfon,0.59forRecormon,0.585forEprex,0.56forEpocrinum.PeryearcostoftreatmentwithEprexwas5382USD,withEralfo n4463 USD,withRecormo n4204USD,withEpocrinum1652USD.Lifelongcostsof treatmentwithEprexwas50017USD,withEralfon49098USD,withRecormon 48785USD,withEpocrinum46596USD.ThecostperQALYwas85499USDper QALYforEprex,83505USDperQALYforEpocrinum,82547USDperQALYfor Recormonand75596USDperQALYforEralfon. CONCLUSIONS: PresentcostutilityanalysesrevealedthatinRussiananemicpatientsondialysisthemost cost-effectivestrategyofantianemictherapyamong4commonerythropoietins wasEralfonwith75596USDperQALY. PSY25 RELATIONSHIPBETWEENMONTHLYOUT-OF-POCKETHEALTHCARECOSTS ANDWORKPRODUCTIVITYAMONGPATIENTSWITHINFLAMMATORYBOWEL DISEASENaimA1,DiBonaventuraM2,LoandJ1,FreedmanD3 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2KantarHealth,NewYork,NY,USA,3KantarHealth,Princeton,NJ,USAOBJECTIVES: InammatoryBowelDisease(IBD)isanimmune-mediatedinammatorydisorderwhichincludesCrohn'sdiseaseandulcerativecolitis.Therelationshipbetweenout-of-pocket(OOP)healthcarecostswithproductivityisnotwell knowninIBD.ThisstudyexplorestherelationshipofOOPhealthcarecostson productivity. METHODS: Participantsself-identiedwithIBDandaged 18completedaninternet-basedcross-sectionalsurveyin1Q2009.Self-reportedproductivitylossincludingabsenteeism,presenteeism,overallworkimpairmentandactivityimpairmentexpressedasimpairmentpercentages,weremeasuredbythe WorkProductivityandActivityImpairmentquestionnaire.Participantswereasked aboutIBD-relatedandoverallmonthlyOOPhealthcarecosts.Negativebinomial regressionmodelsweredevelopedtoadjustforage,gender,ethnicity,income, education,andmaritalstatus. RESULTS: 534participated,67%werefemale,55% wereemployed,and13%wereonshort-andlong-termdisability.Morethanhalf (53%)reportedannualhouseholdincomesbetween$50-100K,40%werecollege educated.OverallOOPcostswere:$0-$50for52.6%,$51-$100for18.2%,$101-$200 for15.5%,and $200for13.7%.Amongemployed,ratesofabsenteeism,presenteeism,andoverallworkimpairmentwere9.97%,16.3%and20.0%,respectively. 43.2%reportedimpairmentindailyactivities.Signicantlygreaterproductivity loss(presenteeism,b 2.79,p 0.0004)andoverallworkimpairment(b 2.78,p 0.0006)wereobservedamongpatientswithoverallOOPcosts $200,ascompared tothosewithOOPcostsbetween$0-50afteradjustingforpatientdemographics. CONCLUSIONS: HighermonthlyOOPcostswereassociatedwithhigherworkimpairmentandlowerproductivity.Shiftinghealthcarecoststoemployeesmaybe associatedwithproductivityloss.FurtherresearchisneededtodeterminelongertermimpactofOOPhealthcarecostsonworkproductivity. PSY26 IMPACTOFHYPONATREMIAONPATIENTOUTCOMESANDHEALTHCARE RESOURCEUTILIZATIONINHOSPITALIZEDPATIENTSDeitelzweigS1,AminA2,LinJ3,BelkK4,BaumerD4,LoweT5 1OchsnerHealthSystem,NewOrleans,LA,USA,2UniversityofCalifornia,Irvine,Irvine,CA, USA,3NovosysHealth,Flemington,NJ,USA,4Premier,Inc.,Charlotte,NC,USA,5Premier HealthcareAlliance,Charlotte,NC,USAOBJECTIVES: Hyponatremia(HN)istheleadingelectrolyteabnormalityamong hospitalizedpatients.Intheabsenceofsymptoms,HNisoftenoverlookedasa conditionthatwarrantsaggressiveintervention.ThoughHNiscommon,littleis knownregardingtheinuenceofHNonpatientoutcomesandhealthcareresource utilization.ThisstudyassessedtheimpactofHNoninpatienttotalandintensive careunit(ICU)cost,totalandICUlengthofstay(LOS),likelihoodofICUadmission, and30-,90-,and180-dayreadmission. METHODS: Premier'sPerspective¨databasewasutilizedtoconstructastratiedsampleofhospitalinpatientsdischarged betweenJanuary1,2007andJune30,2009.HNpatientswereidentiedusingprimaryorsecondarydiagnosisofHN(n 558,815)andwerematchedtoanon-HN controlgroup(n 558,815)usingexactmatchingofage,gender,providerregion, and3MAPR-DRGassignment.Matchingwasfurtherrenedusingpropensity scoresbasedonadditionalpatientandhospitalcharacteristicsandpatientcomorbidities.Costwasanalyzedusingamultivariategenerallinearregressionmodel withgammadistributionandlog-link,LOSwithnegativebinomialdistributionand log-link,andICUadmissionandhospitalreadmissionwithmultivariatelogistic regression. RESULTS: Incontrasttonon-HNpatients,patientswithHNhadsignificantlyhighertotalinpatientcost(9.49%,CI 8.99-10.00;p .0001),ICUcost(9.12%; CI 8.04-10.23;p .0001),totalLOS(12.24%,CI 11.85-12.64;p .0001),andICULOS (11.06%,CI 9.60-12.54;p .0001).HNpatientsweresignicantlymorelikelytobeA63VALUEINHEALTH14(2011)A1–A214

PAGE 72

admittedtotheICU(OR 1.027,CI 1.016-1.037;p .0001),andreadmittedat30 (OR 1.129,CI 1.118-1.141;p .0001),90(OR 1.185,CI 1.175-1.196;p .0001),and 180(OR 1.163,CI 1.153-1.173;p .0001) dayspostdischargecomparedto non-HNpatients. CONCLUSIONS: HNrepresentsasignicantclinical,utilization,andeconomicburdenduetoincreasedcost,LOS,ICUadmissionsand hospitalreadmission. PSY27 REAL-WORLDEVALUATIONOFHEALTHCARERESOURCEUTILIZATIONAND COSTSINPATIENTSWITHNEUROPATHICPAINASSOCIATEDWITHDIABETIC PERIPHERALNEUROPATHY(PDPN)TREATEDWITHPREGABALINOR GABAPENTINUdallM,MardekianJ,HarnettJ Pzer,Inc.,NewYork,NY,USAOBJECTIVES: Toevaluateandcomparechangesinhealthcareresourceutilization andcostsassociatedwiththeinitiationofpregabalinorgabapentininpDPNpatientsinareal-worldsetting. METHODS: Aretrospectivecohortstudyutilizingthe MarketScanCommercialClaimsandEncountersDatabase(2007-2009).Patients withanewprescriptionforpregabalinorgabapentinin2008(dateoftheprescriptiondenedasindexdate)and 1healthcareencounterwithanICD-9codefor diabeticperipheralneuropathy(DPN;250.6or357.2)within30dayspriortotherst prescriptionforpregabalinorgabapentinwereidentiedandpropensityscore matched.Bothcohortswerecontinuouslyenrolledforthe12monthpre-andpostindexperiodsduringwhichhealthcareutilizationandcostswereassessed.Pre-to post-indexchangeswerecomparedbetweenpregabalinandgabapentinusinga difference-in-difference(DID)approach. RESULTS: 910pregabalinpatients(48.6% female;meanage63.3 12.1years)werematchedto910gabapentinpatients(48.8% female;meanage63.3 12.1years);otherdemographicandclinicalcharacteristics werealsocomparablebetweencohorts,withtheexceptionofUSregion.Thepre-to post-indexDIDinresourceutilizationdidnotdifferbetweenthetwocohortsincluding:numberofofcevisitsperpatient(P 0.66),numberofERvisits(P 0.78), numberofinpatientstays(P 0.92),averageinpatientlengthofstayperhospitalization(P 0.79),numberoftotalprescriptionslled(P 0.37).TheDIDoftotal healthcarecostsperpatientwerenon-signicantwithpre-topost-indexincreases of$3,081inpregabalinpatientsand$4,683ingabapentinpatients(P 0.50). CONCLUSIONS: PatientswithpDPNinitiatingpregabalinorgabapentinexperiencedcomparablechangesinhealthcareresourceutilizationandcosts.Theseresultssuggestoverallcostneutralitybetweenpregabalinandgabapentin. PSY28 FACTORSLEADINGTOHIGHHEALTHCAREEXPENDITUREAMONGOBESE INDIVIDUALSINTHEUNITEDSTATESPatelJG,AparasuRR,RajanSS,JohnsonML UniversityofHouston,Houston,TX,USAOBJECTIVES: Toidentifythefactorsassociatedwithhighhealthcareexpenditure amongobeseindividualsintheUnitedStates. METHODS: Analyticalsampleconsistedofobeseadults20yearsorolderwith(BMI 30).Totalyearlyexpendituresfor obeseadultswereobtainedfrom2007MedicalExpenditurePanelSurveydata. Observationswithtotalyearlyhealthcareexpendituregreaterthan90thpercentile wereclassiedashighhealthcareexpendersandcodedyes/notobeusedasthe dependentvariable.IndependentfactorsselectedaccordingtotheAnderson'sBehavioralModelwereage,gender,race,MSA,region,education,povertycategory, sourceofpayment,usualsourceofcare,MSA,education,generalhealthstatus, mentalhealthstatusandmortalityriskstatus.Surveylogistic'regressionwasused todeterminethepredictorsofhighexpenditures. RESULTS: Descriptiveanalysis showthatadultswithhighhealthcareexpenditurewerewhites(74.23%),hadmidschooleducation(60%),livedinMSA(79%),hadusualcareprovider(95%),private insurance(67%)andbelongedtoveryhighmortalityriskgroup(46%).Inadjusted analysis,age(OR:1.019CI:1.009-1.029),usualsourceofcare(OR:1.901CI:1.1963.020),private(OR:3.609CI:1.992-6.539)andpublicinsurance(OR:3.251CI:1.7815.936),veryhigh(OR:3.342CI:2.327-4.798)andhigh(OR:2.213CI:1.666-2.938) mortalityriskandgeneralhealthstatus(OR:2.493CI:2.148-2.893)wereassociated withhighhealthcareexpenditure. CONCLUSIONS: Severalenablingandneedfactorsseemtobeassociatedwithhighhealthcareexpenditure.Bothprivateand publichealthinsurancehaveasignicantburdenofhealthcareexpenditureassociatedwithobesity.Individualswithusualsourceofcarearemorelikelytoincur highhealthcareexpenditure.Comorbiditieshaveasignicantimpactonhealthcareexpenditureamongobeseindividualsandhighlighttheneedtodelivercomprehensivediseasestatemanagement.Thendingsfromthisstudycanhaveimportantimplicationsinunderstandingincreasedutilizationofhealthservices amongobeseadultsintheUnitedStates. PSY29 ACOMPARISONOFDAILYAVERAGESUPPLY(DASUP)OFOXYCODONEAND OXYMORPHONELONG-ACTINGORALTABLETSRubinoM1,SummersKH1,OhsfeldtR2,Ben-JosephR1,PuenpatomRA1 1EndoPharmaceuticals,Inc.,ChaddsFord,PA,USA,2TexasA&MHealthScienceCenter,College Station,TX,USAOBJECTIVES: AlthoughbothoxycodoneCRandoxymorphoneERareapprovedfor twicedailydosing,itisunclearwhetheractualutilizationisconsistentwiththis. Theobjectivewastoassesswhetherthereexistdifferencesintheutilizationof oxycodoneCRandoxymorphoneER. METHODS: Weusedprescriptionclaimsdata fromacommercially-insuredpopulation(i3InVisionDataMartdatabase),during theperiodJanuary1,2007toMarch31,2010.Aclaimforatleastone30-daysupply ofstudydrugonemonthpriortotheobservationperiodwasnecessarytoavoid capturingutilizationassociatedwithtitration.Subsequently,allsubjectshada claimforatleasta90-daysupplyoftherespectiveopioidanalgesic.Patients'claims wereexcludedfromthisanalysisiftherewasevidenceofaswitchfromonetothe otherstudyopioidduringthe90-dayobservationperiod.Therewerenolimitations ontheuseofotheropioids,duringthebaselineorobservationperiods.Thenal cohortconsistedof6,567oxycodoneCRpatientsand796oxymorphoneERpatients. Inadditiontounivariateanalysis,multivariateanalyseswereconductedusing generalizedlinearmodels(GLM)toadjustfortheobservedheterogeneityamong patientsintheobservationaldatabase. RESULTS: Theunadjusteddailyaverage supply(DASUP)meanvaluesforthehigheststrengthsofoxycodoneCR(80mg)was 3.9comparedto2.9foroxymorphoneER(40mg),3.0vs.2.4foralllowerstrengths, and3.1vs.2.5forallstrengths.(allp 0.001).Afteradjustingforbaselinedifferences,arisk-adjustedoverallmeandifferenceinDASUPremainedwithoxycodone CRpatientsreceiving,onaverage,0.6tabletsmoreperdaythanthosedispensed oxymorphoneER(p 0.001).Thedirection,magnitudeandstatisticalsignicanceofthesedifferenceswereessentiallyunchangedinsensitivityanalyses. CONCLUSIONS: Inmanagingthepharmacybenet,decision-makersmaywantto considerthenancialimplicationsoftheseDASUPdifferences. SystemicDisorders/Conditions–Patient-ReportedOutcomes&Preference-Based Studies PSY30 AREAL-WORLDEVALUATIONOFADHERENCEANDPERSISTENCEOFWEIGHTLOSSPHARMACOTHERAPYINTHEUNITEDSTATES:2007-2010CharlandSL,AgatepBC,AubertRE,LazzaroJ,StanekEJ MedcoResearchInstitute,MedcoHealthSolutions,Inc,Bethesda,MD,USAOBJECTIVES: ThereisarapidlygrowingoverweightandobesityepidemicintheUS, andahighdemandforlong-termsafeandeffectiveweight-lossagents.While3 newantiobesitytherapiesunderwentFDAreviewin2010,littleisknownaboutthe utilizationofavailableweight-lossdrugs.Theobjectiveofthisanalysisistodescribethereal-worldprescriptionpatterns,adherence,andpersistenceofweightlosspharmacotherapyintheUnitedStates. METHODS: Aretrospectivecohort analysiswasconductedusingMedco'sintegratedclaimsdatabasetoevaluate adultpatientsinitiatingweight-losspharmacotherapybetweenMay2007-October 2010.Eligibilitycriteriaincludednewweight-lossdrugprescriptionclaims(no weight-losstherapyprescriptions6monthspriortoindexclaimdate)andcontinuouseligibilityfor6monthspre-and14monthspost-indexclaimdate.Patientson drugswith 1000distinctpatientcountswereanalyzedforadherence(annual MedicationPossessionRatio;MPR),persistency(allowinga45daygap),andconcomitanttherapy. RESULTS: Analysesincluded91,160patientsreceivingve drugs:phentermine(N 67434),sibutramine(N 13438),orlistat(N 8047),phendimetrazine(N 4631),and,diethylpropion(N 4350).Mean SDagewas44 12 years(96%,18-64y/o),82%werefemale,46%residedintheSouth,and91%lled prescriptionsatretailonly.Amongobesitydrugprescriberswithknownspecialty, 71%werefamily/internalmedicinepractitioners.Patientsreceivedameanof3 3 concomitantchronicmedications,with38%,31%,22%,and11%onantihypertensive,antidepressant,dyslipidemia,ororalantidiabetictherapy,respectively.The meanadherencerangedfromMPR 0.20(phendimetrazine)toMPR 0.26(phentermine).Persistenceat3,6,and12monthsrangedrespectivelyfrom26%-38%(lowphendimetrazine,high-phentermine),9%-16%(low-phendimetrazine,high-sibutramine),and3%-6%(low-phendimetrazine,high-sibutramine). CONCLUSIONS: Weight-losspharmacotherapiesintheUnitedStateswereprescribedbyprimary carephysicianstopredominantlyyounger,femalepatientsonconcomitanttherapyforcommonobesity-relatedconditions.Adherenceandpersistencetotherapy islow,evenovershort-termexposure,althoughtreatmentdurationmayextend beyondrecommendationsinsomecases(e.g.,phentermine). PSY31 HEMOPHILIAA:PATIENTIMPACTANDECONOMICBURDENOFTHEDISEASEStrandberg-LarsenM1,GaterA2,ThomsonA2 1NovoNordisk,Inc.,Sborg,Denmark,2MAPIValues,Bollington,Cheshire,UKOBJECTIVES: HemophiliaAisararebutseriousbleedingdisordercausedbyablood clottingfactorVIII(FVIII)deciency.ToexaminetheimpactofhemophiliaAand replacementFVIIItherapiesfromahumanisticandeconomicperspective,atargetedliteraturereviewwasconducted. METHODS: Searcheswereconductedin MEDLINE¨andtheNationalHealthServiceEconomicEvaluationDatabase,using disease,patient-reportedoutcomeandeconomic-relatedkeywords,limitedto EnglisharticlespublishedfromJanuary2000-January2010(inclusive).From653 abstractsretrieved,34full-textarticleswereselectedfordetailedconsideration. RESULTS: FindingsrevealedincreasedmortalityratesanddecreasedlifeexpectancyamongpeoplewithhemophiliaA,comparedwiththegeneralpopulation. Thisislargelyattributedtothetransmissionofblood-borneviruses(e.g.HIVand Hepatitis-C)duetouseofplasma-derivedFVIII(pdFVIII)concentrates.ImprovementsinviralattenuationprocessesplusdonorscreeningpracticesforpdFVIII productshavereducedtherisksofviraltransmission,butriskfromnon-enveloped virusesandotherunknownpathogensstillexists.Newlydevelopedrecombinant FVIIItherapiesminimizetheserisks;however,suchtherapiesarenotcurrently widelyavailableglobally.AllavailableFVIIItherapiesusedemandingregimens requiringregulartime-consuminginjections,whichcanbedetrimentaltopatients' health-relatedqualityoflifeandlimitadherence.Non-adherenceisassociated withdiminishedproductefcacy,poorerhealthoutcomesandincreasingeconomicexpenditure.TreatinghemophiliaAiscostly,primarilyduetothehigh acquisitioncostofreplacementFVIIIproducts.However,indirectcosts(e.g.patient disability)alsocontributetoeconomicburden.Prophylactictreatment,whileini-A64VALUEINHEALTH14(2011)A1–A214

PAGE 73

tiallymoreexpensivethanon-demandtreatment,isassociatedwithgreaterclinicalefcacyandimprovedlong-termoutcomes,whichmayleadtocostsavings overapatient'slifetime. CONCLUSIONS: HemophiliaAisassociatedwithconsiderablehumanisticandeconomicburden.Substantialunmetneedsremainamong hemophiliaApatientswithregardtothesafety,convenienceandglobalaccessto FVIIItherapy. PSY32 HOWDOMINIMALLYIMPORTANTDIFFERENCESVALUEOUTCOMES?SuhJK,DoctorJ USCSchoolofPharmacy,LosAngeles,CA,USAOBJECTIVES: Toassessthefunctionalformoftheminimallyimportantdifference (MID) METHODS: Subjectswereshownabasebodyweight,y1,andacurrentbody weightafteratreatment,y2.Ifasubjectendorsedy2asanimportantdifference overy1,wedenotedapair(y2,y1).AnMIDisthey1andy2differencewhenthe probabilityofdiscriminatingy2fromy1wasgreaterthan0.5.Thestandardformula ofthisdiscriminationprobabilityisgivenbyFechner'sLaw:P(y2,y1) F(U(y2)U(y1))whereFisanincreasingfunctionandUisaMIDfunction.Whentheprobabilityofdiscriminationisconstantoverproportionalchangesinweight(i.e.,Weber'sLaw:P(y2,y1) P(y2,y1)),theutilityfunctionforhealthstatesisalog function.ToassessthefunctionalformofU,wetestedahypothesisofP(y2,y1) P(y2,y1) 0.5.Anon-linesurveywasconductedwiththreebaseweights:29,36, and43lboverweight.One-wayrepeatedmeasureanalysisofvariance(Wilk'stest) andone-waywithinsubjectsanalysiswereused. RESULTS: Atotalof131subjects participated.ThemeanMIDvalueswere9.391(SD 4.001),11.195(SD 5.236),and 13.073(SD 6.491)for29,36,and43lboverweight,respectively.Thenormalized meanMIDvalueswere0.323(SD 0.138),0.311(SD 0.145),and0.304(SD 0.151) for29,36,and43lboverweight,respectively.TheresultsofWilk'stestwasnot statisticallysignicant(p 0.2807)Moreover,one-waywithinsubjectsresultedin noeffectofthedifferentbodyweightsontheMIDsatthesignicancelevel 0.01. CONCLUSIONS: ThenormalizedmeanMIDsfromthreedifferentbodyweights werenotstatisticallydifferent.WithinthelimitedrangestudiedMIDstalog utilitymodel.Thus,MIDsmaybeproblematicasaprescriptivepolicytool. PSY33 EXPLORINGTHEFACTORSTRUCTUREOFTHEPROMISFATIGUEITEMBANKIN PATIENTSWITHFIBROMYALGIASchillingS1,BlumSI2,TourkodimitrisS2,WilliamsDA1 1UniversityofMichigan,AnnArbor,MI,USA,2ForestResearchInstitute,JerseyCity,NJ,USAOBJECTIVES: TheNIHroadmapproject"Patient-ReportedOutcomesMeasurement InformationSystem"(PROMIS)providesabasisformeasuringsymptoms-includingfatigue;associatedwithchronicdiseases.However,informationonuseofPROMISwithFibromyalgia(FM)patientsislimited.Thisstudysoughttoevaluatethe factorstructureofthe95-itemPROMISFatigueitembankamongindividualswith FM.Itisunknownwhethertheconceptualstructureoffatigueexperience(intensity,frequency,andduration)andimpact(physicalandmental)willbesupported inindividualswithFM. METHODS: Aninternet-basedsurveywasconductedwith individualswithFM.Participantshadtobe 18yearsanddiagnosedwithFMbya physician.Respondentswererandomizedintooneofthreecohorts,eachofwhich completedone-thirdofthe95-itemsfromtheentirePROMISv1.0FatigueItemBank (www.nihpromis.org).Itemswerebalancedbetweencohorts.Itemfactoranalysis wasperformedusingbothapolychoriccorrelationmatrixwithMplusandafullinformationapproachusingORDFAC.ComparisonsweremadeusingtheAkaike InformationCriteria(AIC)andBayesianInformationCriterion(BIC).Modelschosen basedonBICproducedinterpretablefactorsforassessmentofthePROMAXrotated factorloadings. RESULTS: 1,207respondentscompletedthesurvey.Meanageof therespondentswas52.1years.MeantimesinceFMdiagnosiswas8.7years.All groupsproducedfactorstructuresthatincluded3factors.Therstconsisted mainlyofFatigueExperienceitems.Thesecondconsistedmainlyofitemsrelated toMentalImpactandthethirdconsistedofGeneralImpactitems.Anadditional factorrelatedtoSocialImpactwasidentiedinGroupBonly. CONCLUSIONS: For individualswithFM,thePROMISFatigueitembankappearstocontainafactor structurethatisconsistentwiththeoriginalPROMISconceptualmodel.Subscales fromthePROMISFatigueitembankcanlikelybeusedtoseparateFatigueExperiencefromGeneralandMentalimpactsofFatigue. PSY34 CALCULATINGPROMISFATIGUEITEMBANKIRTPARAMETERESTIMATES SPECIFICTOPATIENTSWITHFIBROMYALGIASchillingS1,BlumSI2,TourkodimitrisS2,WilliamsDA1 1UniversityofMichigan,AnnArbor,MI,USA,2ForestResearchInstitute,JerseyCity,NJ,USAOBJECTIVES: TheNIHroadmapproject"Patient-ReportedOutcomesMeasurement InformationSystem"(PROMIS)providesabasisformeasuringsymptomsincludingfatigue;associatedwithchronicdiseases.However,informationonuseofPROMIS withFibromyalgia(FM)patientsislimited.Thisstudysoughttore-estimatethe ItemResponseTheory(IRT)itemparameterestimatesforeachofthe95-items includedinthePROMISFatigueitembankusingresponsesfromindividualswith FM.ItisunknownwhethertheparameterestimatesdifferamongasampleofFM patientsfromthatofthebroaderpopulationusedintheinitialPROMISitem calibrations. METHODS: Aninternet-basedsurveywasconductedwithindividuals withFM.Participantshadtobe 18yearsanddiagnosedwithFMbyaphysician. Respondentswererandomizedintooneofthreecohorts,eachofwhichcompleted one-thirdofthe95-itemsfromtheentirePROMISv1.0FatigueItemBank(www.nihpromis.org).IRTmodelttingofSamejima'sGradedResponseModel(GRM)was conductedusingMULTILOG7todetermineitemparameterestimatesandcomparedwithstandardPROMISparameterestimates. RESULTS: Atotalof1,207respondentscompletedthesurvey.IRTitemparameterestimates(slopeanditem characteristiccurvethresholds)anditeminformationfunctionswereobtainedfor all95-items.MeanscoreacrossallitemsfortheFMsamplewas3.68comparedto thePROMISmeanof2.12.Maximuminformationfunctionwas 2for22/95itemsin theFMsample,comparedtoonly2/95itemsinthePROMISsample,indicating differencesinbothdiscriminateabilityaswellasthedifcultyofitemsforaFM specicpopulation. CONCLUSIONS: DiseasespecicitemlevelcalibrationsofPROMISitemscanbeestimated.UsingthesameGRMmethodologyasPROMISwewere abletocalculatereviseditemparameterestimatesandinformationfunctions. Theseestimatescanbeusedtoconstructimprovedefcientshort-formsorfor applicationincomputeradaptivetesting. PSY35 VALIDATIONOFTHELUPUSIMPACTTRACKERJollyM1,KosinskiM2,GarrisCP3,JhingranPM3,MikolaitisRA1,DennisG4,WallaceDJ5, ClarkeA6 ,ParkeA7,DooleyMA8,StrandV9,Alarc—nGS10 1RushUniversity,Chicago,IL,USA,2QualityMetricIncorporated,Lincoln,RI,USA,3GlaxoSmithKlineR&D,ResearchTrianglePark,NC,USA,4HumanGenomeSciences,Rockville, MD,USA,5UCLA,WestHollywood,CA,USA,6McGillUniversity,Montreal,QC,Canada,7UniversityofConnecticutHealthCenter,atSt.FrancisHospitalandMedicalCenter,Hartford, CT,USA,8UniversityofNorthCarolinaatChapelHill,ChapelHill,NC,USA,9Stanford University,PortolaValley,CA,USA,10UniversityofAlabamaatBirmingham,Birmingham,AL, USAOBJECTIVES: ToevaluatethereliabilityandvalidityoftheLupusImpactTracker (LIT),apatient-basedassessmentofsystemiclupuserythematosus(SLE)disease impact. METHODS: Reliabilityandvalidityofscalescoresusingthe10selectedLIT itemswereevaluatedusingtwodatasets(cross-sectional,n 253andlongitudinal, n 307)whichcollectedinformationonqualityoflife(SF-36),diseaseactivity(SLEDAItotalscore,revisedSELENAareindex)andpatients'overallgeneralhealth rating(GHR)fromLupusPRO.Internalconsistencyreliability(ICR)wasestimated inbothstudiesusingCronbach'salpha.Inthelongitudinaldataset,reliabilitywas alsoevaluatedbycomputingtheintra-classcorrelationbetweenLITscalescoresat baselineandfollow-up.ConvergentvaliditywasevaluatedbycorrelatingLITscale scorewithSF-36,diseaseactivityandGHR.Constructvaliditywasexaminedby known-groupsvalidity(comparingmeanscalescoresacrossgroupsknowntodifferonconceptuallyrelatedcriterionmeasures).AnalysisofvariancewasconductedtocompareLITscalescoresacrosspatientsstratiedbyeachcriterion measure.Responsivenesswasstudiedusingknown-groupsvalidity.Specically, themagnitudeanddirectionofmeanscorechangesontheLITscaleagainst changesinGHR(0-100scale)wereevaluated. RESULTS: The10LITitems(concentration,medicationsideeffects,fulllingfamilyresponsibilities,wokeupfeeling wornout,painandachinginbody,limitedactivitiesduetopain/fatigue,anxiety, depression,selfconsciousaboutappearance,abilitytoplanactivities)showed goodICR(0.89)andtest-retestreliability(0.87).LITitemscorrelatedmoderately (r 0.5)withSF-36domainsandsummarymeasuresandtheGHR.Meanscale scoresdifferedsignicantly(p 0.05)inthehypothesizeddirectionacrossgroups stratiedbydiseaseactivityandGHR. CONCLUSIONS: TheLupusImpactTracker, a10-itemPRO,isreliableandvalidandoffersapracticalwayofassessingthe impactoflupusonapatient'slife. PSY36 DEVELOPMENTOFTHELUPUSIMPACTTRACKER:ATOOLFORPATIENTSAND PHYSICIANSTOASSESSANDMONITORTHEIMPACTOFSYSTEMICLUPUS ERYTHEMATOSUS(SLE)JollyM1,KosinskiM2,GarrisCP3,JhingranPM3,MikolaitisRA1,DennisG4,WallaceDJ5, ClarkeA6 ,ParkeA7,DooleyMA8,StrandV9,Alarc—nGS10 1RushUniversity,Chicago,IL,USA,2QualityMetricIncorporated,Lincoln,RI,USA,3GlaxoSmithKlineR&D,ResearchTrianglePark,NC,USA,4HumanGenomeSciences,Rockville, MD,USA,5UCLA,WestHollywood,CA,USA,6McGillUniversity,Montreal,QC,Canada,7UniversityofConnecticutHealthCenter,atSt.FrancisHospitalandMedicalCenter,Hartford, CT,USA,8UniversityofNorthCarolinaatChapelHill,ChapelHill,NC,USA,9Stanford University,PortolaValley,CA,USA,10UniversityofAlabamaatBirmingham,Birmingham,AL, USAOBJECTIVES: ToderiveashortformquestionnairefromtheLupusPRO,forusein dailypracticetoassesstheimpactofSLE. METHODS: LupusPROisa44-itemvalidatedqualityoflifeinstrument.Toidentifythesubsetofitemsthatwouldta unidimensionalstructure,one-factorconrmatoryfactoranalysis(CFA)and itemto-factorloadingsandgoodnessoftstatistics(ConrmatoryFitIndex[CFI],TuckerLewisIndex[TLI])wereconducted.Backwardstepwiseregressionmethodswere usedtoeliminateitemsthatshowedtheweakestrelationshiptoclinicalcriterion measuresofdiseaseactivity(SLEDAItotalscore,revisedSELENAareindex)and patients'overallgeneralhealthratings(fromLupusPRO).Fourfocusgroupswere conducted:patientswereaskedtorateeachitemofLupusPROusinga4-point importancescaleandtoidentifythe"top15"mostimportantitems. RESULTS: BasedonCFA,21itemstaunidimensionalstructure.Bothgoodnessoftstatistics(CFI,TLI)exceededtherecommendedthresholdof0.9.Allitemsshowedhigh factorloadingsrangingfrom0.76to0.91.Backwardstepwiseregressionanalyses resultedin12itemsmosthighlyrelatedtopatientandclinicalcriterionmeasures. Offocusgroupparticipants,80%ratedthetop15itemsas"moderately"or"very" important(importanceratings 2.3).Intotal7itemswereselectedacrossthe3 approaches(CFA,backwardregression,focusgroups).Another14wereselected basedonresultsfromatleast2approaches.Itemsmeasuringsymptomsandprocreationwereexcluded:symptomitemsdidnottwiththeintentofthetool; procreationitemswouldnotapplytoallpatients.Finally,changesintheitem selectionbasedonclinicalconsensusweremade,withoutlosingpsychometric properties. CONCLUSIONS: Usingquantitativeandqualitativemethods,10itemsA65VALUEINHEALTH14(2011)A1–A214

PAGE 74

wereselectedforinclusionintheLupusImpactTracker,aPROintendedforclinical practiceusetomonitortheimpactofSLEonpatients'livesandwellbeing. PSY37 THEPAINMANAGEMENTISALWAYSAPUBLICHEALTHCHALLENGEBertinP1,RahhaliN2,AugesM2,TaiebC2 1ComitŽLUttecontrelaDouleur,Limoges,France,2PFSA,BoulogneBillancourt,FranceOBJECTIVES: Adailyassessmentofthespeedofactionandeffectivenessoftreatmentofacombinationofparacetamolandcodeineinpatientssufferingfrom intensepain,whichhasprogressedsincelessthan7days. METHODS: AmulticentrelongitudinalobservationalprospectivestudycarriedoutinmetropolitanFrance usingdatacollectedbygeneralpractitionerswhoagreedtoparticipate. RESULTS: A totalof574patientstreatedbyaparacetamol-codeinecombination(600mg/50mg and400mg/20mg)wereincluded.Theseverityofpainmeasuredatinclusionusing avisualnumericscalewas7 1.28.Theseverityofpainmeasuredafterhalfadayof treatmentwas5.61 1.87and5.28 1.86attheendoftherst24hoursoftreatment.Asignicantimprovementinpainwasobservedfromthersthalf-day (p 0.001).Theseverityofpainonthe2nd,4thand7theveningswasrespectively 4.08 1.85;2.76 1.76and1.79 1.69OnD1,69.7%declaredtreatmenttobeeffective,61.04%weresatisedwiththeirtreatmentand79.26%didnotobserveanyside effectstothetreatment.OnD7,96.15%(D3,91.3%)declaredtreatmenttobeeffective,87.10%weresatisedwiththeirtreatmentand88.96%didnotobserveanyside effectstothetreatment.8outof10patientsdidnotcomplainaboutsideeffects relatedtothetreatment. CONCLUSIONS: Areductioninpainwithintherst12 hoursshowedthepertinenceoftreatmentusingaparacetamol-codeinecombination.Thispertinencewasconrmedbytwo-thirdspatientswhodeclaredthetreatmenttobeeffectivefromthe1stday,and91%ofthemdeclaredthisonthe3rdday PSY38 PATIENT-REPORTEDOUTCOMESINSUBJECTSWITHPAINFULDIABETIC PERIPHERALNEUROPATHY:PAINDESCRIPTIONANDQUALITYOFLIFEKoKS1,ChaBY2,KimCH3,KwonHS2,LeeJH4,ParkTS5,WonJC1,KoSK6,ParkHJ6 1InjeUniversitySanggyePaik-Hospital,Seoul,SouthKorea,2TheCatholicUniversityofKorea,St. Mary'sHospital,Seoul,SouthKorea,3SejongGeneralHospital,Bucheon-Si,Gyeonggi-do,South Korea,4DaeguCatholicUniversityMedicalCenter,Daegu-City,SouthKorea,5ChonbukNational UniversityHospital,JeonjuSi,Jeollabuk-Do,SouthKorea,6PzerPharmaceuticalsKoreaLtd., Seoul,SouthKoreaOBJECTIVES: Thediabeticperipheralneuropathy(DPN)isthemostcommon chroniccomplicationofdiabetes.Neuropathicpaintendstobesevere,chronic,and hasahugelynegativeimpactonthephysicalandemotionalfunctioning,andoverallwell-beingofthepatient.Thisstudywasperformedtoevaluatethescopeof painandqualityoflifeamongpatientswithpainfuldiabeticperipheral neuropathy. METHODS: Across-sectional,observationalstudywascarriedouton type2diabeticpatientsfrom40medicalcentersthroughoutKorea.ThestandardizedquestionnairessuchasMichiganNeuropathyScreeningInstrument(MNSI), BriefPainInventory-shortform(BPI-sf),MOS6-itemsSleepScaleandEuroQoL (EQ-5D)wereusedtodiagnosisandestimatethequalityoflifeinpeoplewithDPN. ApatientwithDPNwasdenedasthepatientswhoshowedpositivepatients medicalrecordsorneuropathicsymptoms(MNSI 2)andpositiveresponseonthe basicneurologicalexamination(Monolamenttest). RESULTS: Among4,000diabeticpatients,33.5%hadDPN.Ofthose1,338DPNpatients,874(65.3%)tookpain medicationsorhad 1worstpainscores(BPI-sf,0-10scales).Patientswhotook painmedicationsorhad 1worstpainscoreswere63 10.6yearsold(57.9%female),haddiabetesfor13 8.0yearsandDPNfor4.0 3.3years.AverageandWorst Painscorewere2.9 2.3and4.9 3.1.Paininterferenceindexwerehigherinpatientswhotookpainmedicationsorhad 1worstpainscores(2.4 2.6and0.2 0.8, respectively, P 0.0001).Painsubstantiallyinterferedwithgeneralactivity,mood, walking,normalwork,relationships,sleep,andenjoymentoflife.Thesepatients reportedgreatersleepproblemscomparedwithDPNpatientswithoutpain.The meanEQ-5Dutilityscorewas0.6 0.2. CONCLUSIONS: Thisstudyfoundthatpain associatedwithdiabeticneuropathyisasignicantmedicalissuethathasasubstantialimpactonthequalityoflifeofmanypeoplewiththiscondition. PSY39 EPIDEMIOLOGYOFHEREDITARYCOAGULOPATHIESINKAZAKHSTAN: PATIENT-REPORTEDDATAVorobyevP1,BorisenkoO1,ZhulyovY2,RybalovaT3,KrasnovaL4,BezmelnitsynaL1 1RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,Russia,2Russia HemophiliaSociety,Moscow,Russia,3KazakhstanHemophiliaSociety,Almaty,Kazakhstan,4MoscowStateMedicalUniversitynamedafterI.M.Sechenov,Moscow,RussiaOBJECTIVES: Toassesshealthstatus,treatmentpatternsandqualityoflifein patientswithhereditarycoagulopathiesinKazakhstan. METHODS: Postalandtelephonehealthsurvey.QuestionnairesweredistributedinMay-August2009.The questionnairecontainedquestionsonclottingfactorlevel,numberofbleeding, numberofinjectionsofclottingfactors,waysofreceivingmedications,numberof ambulancecallsandhospitalizations,andthewayofadministrationofmedicines. Thepatients'educationlevelandemploymentdatawascollected.Analysiswas performedwith2andStudent'scriteria. RESULTS: 626completedquestionnaires werereceivedbySeptember,2009.HemophiliaAwasreportedby486responders (77.6%).426patients(68%)wereadults.Meanagewas25.4years.60patientswith hemophiliaAandB(11.6%)hadalevelofcoagulationfactorslowerthen1%,i.e. theyhadasevereformofdisease.Thepresenceofantibodieswasdetectedin68 patients(11%).87.8%ofpatientsexperiencedbleedinginthelastmonth.Themedianofanumberofbleedingamongallbledpatientswas2.84.4%ofpatientsused clottingfactorsinthelastmonth.Themediannumberofinjectionsofclotting factorVIIIinpatientswithhemophiliaAwas2(standardfrequencyis12).7.2%of patientscalledforanambulanceinalastmonth(mediumfrequencyofcallsinlast month2).7.4%ofpatientswerehospitalizedduringlastmonth(mediumfrequencyofhospitalization2).564(90.2%)ofpatientsperformtheinjectionsof clottingfactorthemselves.12.8%ofpatientsreporteduseofprophylaxisschemeof treatmenthemophilia,70.6%ofpatientsreporteduseof"ondemand"scheme. CONCLUSIONS: Resultsofsurveyshowedverylowrateoftreatmentofhemophilia withproperprophylacticscheme(lessthan13%ofpatients).Therateofambulance callsandhospitalizationsiscomparativelylow.Themajorityofpatientsintroduce thefactorsthemselves. PSY40 EPIDEMIOLOGYOFHEREDITARYCOAGULOPATHIESINUKRAINE:PATIENTREPORTEDDATAVorobyevP1,BorisenkoO1,ZhulyovY2,KrasnovaL3,BezmelnitsynaL1 1RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,Russia,2Russia HemophiliaSociety,Moscow,Russia,3MoscowStateMedicalUniversitynamedafter I.M.Sechenov,Moscow,RussiaOBJECTIVES: Toassesshealthstatus,treatmentpatternsandqualityoflifein patientswithhereditarycoagulopathiesinUkraine. METHODS: Postalhealthsurvey.QuestionnairesweredistributedinMay-August2009.Thequestionnaire containedquestionsonclottingfactorlevel,numberofbleeding,injectionsofclottingfactors,waysofreceivingmedications,numberofambulancecallsandhospitalizations,andthewayofadministrationofmedicines.Analysisofexperimental datawasperformedwithsuchstatisticalparametersas2andStudent'scriteria. RESULTS: 154completedquestionnaireswerereceivedbySeptember,2009.HemophiliaAwasreportedby103responders(67%).126(81.8%)patientswereadults.The detectionofalevelofcoagulationfactorsduringlastyearwasdonein50(40.4%) patientswithhemophiliaAandB.24(19.4%)patientswithhemophiliaAandBhad alevelofcoagulationfactorslowerthen1%,i.e.theyhadasevereformofdisease. Thepresenceofantibodieswasdetectedin27patients(17.5%).80.6%ofpatients experiencedbleedingduringthelastmonth(median-4).51.4%ofpatientswith hemophiliaAusedclottingfactorVIIIduringlastmonth.Themediannumberof injectionsofclottingfactorVIIIinpatientswithhemophiliaAwas4(standard frequencyis12).36.4%ofpatientscalledforanambulanceduringlastmonth (mediumfrequency1).49.3%ofpatientswerehospitalizedduringlastmonth (mediumfrequency1).Only23(15%)ofpatientsperformedtheinjectionsof clottingfactorthemselves.Only1.3%ofpatientsreporteduseofprophylaxis schemeoftreatmenthemophilia,85.7%ofpatientsreporteduseof"ondemand" scheme. CONCLUSIONS: Resultsofsurveyshowedverylowrateoftreatmentof hemophiliawithproperprophylacticscheme(aboutthan1.3%ofpatients).The rateofambulancecallsandhospitalizationsisextremelyhigh.Themajorityof patientsintroducethefactorsinhospitals. PSY41 HEALTH-RELATEDQUALITYOFLIFEINKAZAKHSTANPATIENTSWITH HEREDITARYCOAGULOPATHIESVorobyevP1,BorisenkoO1,ZhulyovY2,RybalovaT3,KrasnovaL4,BezmelnitsynaL1 1RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,Russia,2Russia HemophiliaSociety,Moscow,Russia,3KazakhstanHemophiliaSociety,Almaty,Kazakhstan,4MoscowStateMedicalUniversitynamedafterI.M.Sechenov,Moscow,RussiaOBJECTIVES: Toassesshealthstatus,treatmentpatternsandqualityoflifein patientswithhereditarycoagulopathiesinKazakhstan. METHODS: Postalandtelephonehealthsurvey.QuestionnairesweredistributedinMay-August2009.The questionnairecontainedquestionsonclottingfactorlevelandpresenceofantibodiestoit,numberofbleedinginlastmonth,numberofinjectionsofclotting factorspermonth,namesofusedmedications,waysofreceivingmedications, numberofambulancecallsandhospitalizations,andthewayofadministrationof medicines(self-administrationofadministrationataambulatory/hospital).The patients'educationlevelandemploymentdatawascollected.Health-relatedqualityoflifewasassessedwithself-administratedvalidatedversionofRussianversionofEuroqol-5Dquestionnaire.Comparisonwasmadewithhealth-relatedqualityoflifeofRussianpatientsdata(P.Vorobeyvet.al.,2008,dataof1003patientswas usedinanalyses). RESULTS: 626completedquestionnaireswerereceivedbySeptember,2009.Health-relatedqualityoflifewasassessedforpatientsolderthan11 years(n 514).MorethanhalfofpatientsreportedproblemswithineachofEQ-5D dimensionsofhealth.Thus63.9%ofpatientsreportedofproblems(moderateand severe)withmobility(63.9%ofRussianpatients,p 0.05);65.2%ofpatientsinformedofdifcultieswithself-care(35.5%ofRussianpatients,p 0.05);70.1%of patientshaddifcultieswithusualactivity(61.9%ofRussianpatients,p 0.05);74% ofpatientsreportedofpresenceofpainordiscomfort(78.4%ofRussianpatients, p 0.05);27.3%ofpatientshadananxietyordepression(54.2%ofRussianpatients, p 0.05).Theaveragevalueofqualityoflifeevaluatedwithvisual-analogscale (VAS)was0.53(SD0.18),median-50. CONCLUSIONS: Resultsofthestudyshown highrateofproblemswithmobility,usualactivityandhighrateofpain/discomfort andlowrateofproblemwithanxiety/depression. PSY42 EPIDEMIOLOGYANDQALITYOFLIFEOFANEMICPATIENTSONDIALYSISIN RUSSIA:PATIENT-REPORTEDDATAVorobyevP1,BezmelnitsynaL1,BorisenkoO1,EliseevaE2,BaevV3,VezikovaN4 1RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,Russia,2Vladivostok StateMedicalUniversity,Vladivostok,Russia,3AbakanCityHospital,Abakan,Russia,4PetrozavodskStateUniversity,FacultyofMedicine,Petrozavodsk,RussiaOBJECTIVES: Toassessepidemiologyofanemiacasedbychronicrenalfailurein patientsondialysisandpatient'squalityoflife. METHODS: Healthsurveyofpatientsindialysisdepartmentsof23regions.About1400questionnairesweredistributedinMay-June2010.Theformcontainedfollowingquestions:personal information(name,age,patients'educationandemployment),medicaldata(di-A66VALUEINHEALTH14(2011)A1–A214

PAGE 75

agnosis,durationofdisease,clinicalandlaboratorysymptoms,proceduresofdialysisandantianemictherapy).Health-relatedqualityoflifewasassessedwith self-administratedvalidatedversionofRussianversionofEuroqol-5Dquestionnaire.Inclusioncriteriawereageolderthan18years,terminalstageofchronic renalfailureandtreatmentwithdialysismorethen4months. RESULTS: 1395 completedquestionnaireswerereceivedbySeptember,2010.Meanagewas42 years.45%patientsreportedchronicglomerulonephritisasamainreasonof chronicrenalfailure,22%-chronicpyelonephritis,10%-diabetes,8%-polycystic kidneydisease,otherdiseases1,5%.Averagelevelofhemoglobinwas107 16.2 g/l.80%ofpatientsreportedthattheygot12-14proceduresofdialysisduringlast month.Theaveragelengthofoneprocedurewas4-5hoursjustfor17%ofpatients, inothercases(80%)lengthofdialysiswasshorter.Analysisofantianemictherapy hasshownthat85%goterytropoetinsand70%-ironsupplements.Morethanhalf ofpatientsreportedproblemswithineachofEQ-5Ddimensionsofhealth.Thus 42.5%ofpatientsreportedproblemswithmobility,20%-difcultieswithself-care; 58.7%-difcultieswithusualactivities;67%-reportedaboutpainordiscomfort; 47%-reportedaboutananxietyordepression.Theaveragevalueofqualityoflife evaluatedwithvisual-analogscale(VAS)was0.59(SD0.17),median0.6. CONCLUSIONS: Analysesofqualityoflifehavedemonstratedhighrateofproblem withusualactivityandpainanddiscomfortandlowrateofproblemwithselfservice. PSY43 HEALTH-RELATEDQUALITYOFLIFEINUKRAINEPATIENTSWITHHEREDITARY COAGULOPATHIESVorobyevP1,BorisenkoO1,ZhulyovY2,KrasnovaL3,BezmelnitsynaL1 1RussianSocietyforPharmacoeconomicsandOutcomesResearch,Moscow,Russia,2Russia HemophiliaSociety,Moscow,Russia,3MoscowStateMedicalUniversitynamedafter I.M.Sechenov,Moscow,RussiaOBJECTIVES: Toassesshealthstatus,treatmentpatternsandqualityoflifein patientswithhereditarycoagulopathiesinUkraine. METHODS: Postalhealthsurvey.QuestionnairesweredistributedinMay-August2009.Health-relatedquality oflifewasassessedwithself-administratedvalidatedversionofRussianversionof Euroqol-5Dquestionnaire.Comparisonwasmadewithdataabouthealth-related qualityoflifeofRussianpatients(P.Vorobeyvet.al.,2008,dataof1003patientswas usedinanalyses).Analysisofexperimentaldatawasperformedwith2criteria. RESULTS: 154completedquestionnaireswerereceivedbySeptember,2009. Health-relatedqualityoflifewasassessedforpatientsolderthan11years(n 142). MorethanhalfofpatientsreportedproblemswithineachofEQ-5Ddimensionsof health.Thus88.8%ofpatientsreportedofproblems(moderateandsevere)with mobility(63.9%ofRussianpatients,p 0.05),57.7%ofpatientsinformofdifculties withself-care(35.5%ofRussianpatients,p 0.05);81%ofpatientshaddifculties withusualactivity(61.9%ofRussianpatients,p 0.05);90.9%ofpatientsreportedof presenceofpain/discomfort(78.9%ofRussianpatients,p 0.05);60.5%ofpatients hadananxietyordepression(54.2%ofRussianpatients,p 0.05).Theaveragevalue ofqualityoflifeaccordingtovisual-analogscalewas0.44(SD0.22). CONCLUSIONS: ThestudyofqualityoflifeinpatientswithhereditarycoagulopathieswasperformedforthersttimeinUkraine.Resultsofthestudyshownhighrateofproblemsallaspectsofhealth-relatedqualityoflife.Ukrainepatientswithhemophilia haveworsequalityoflifethanRussianpatients. PSY44 PAINMANAGEMENT:IMPACTONQUALITYOFLIFEBertinP1,RahhaliN2,AugesM2,TaiebC2 1ComitŽLUttecontrelaDouleur,Limoges,France,2PFSA,BoulogneBillancourt,FranceOBJECTIVES: Theghtagainstpain,whichrepresentsapublichealthchallenge andacriteriaofthequalityanddevelopmentofahealthsystem,aboveallaspires tomeetobjectivesofhumanity,ethicsandhumandignity.Physicalpainandmentalsufferingexperiencedacrossallagegroupsrendersthoseaffectedbyillness evenmorevulnerable.Paincausesincapacity,handicapandconsiderabledeteriorationinthequalityoflife.Fightingagainstpainmeetsalegitimateexpectationof everyperson. METHODS: Toassessthequalityoflifeinpatientssufferingfrom intensepainwhichhasprogressedsincelessthan7daystreatedbyacombination ofparacetamolandcodeine.Amulti-centrelongitudinalobservationalprospective studycarriedoutinmetropolitanFranceusingdatacollectedbygeneralpractitionerswhoagreedtoparticipate. RESULTS: 574patientstreatedbyaparacetamolcodeinecombination(600mg/50mgand400mg/20mg)wereincluded;atinclusion thequalityoflifeassessedusingSF-12wasaffectedasmuchintermsofthemental component(41.83 7.92)asthephysicalcomponent(35.88 11.68)thenormof thescoresforeachcomponentisequalto50onD7,thequalityoflifeassessedin asimilarmannerusingSF-12was43.31 9.89forthementalcomponentand40.93 7.92forthephysicalcomponent.Astatisticallysignicantimprovementwasnoted foreachofthe2mental(p 0.0091)andphysical(p 0.0001)componentsbetween therstdayoftreatmentandthe7thday. CONCLUSIONS: Theimprovementin qualityoflifeobserveddirectlythroughSF-12wasalsoconrmedbypatientsatisfaction:fromtherstday,61%ofpatientsdeclaredthemselvestobesatised.On the7thdayoftreatment,87.10%weresatisedwiththeirtreatment. PSY45 CHARACTERISTICSOFCAREGIVERSANDTHEIREXPERIENCEASSOCIATED WITHCONGENITALHEMOPHILIAOFINHIBITORPATIENTSWisniewskiT1,PetrillaAA2,DeKoevenM2,LeeWC2,CooperDL1,HolotN1, vonMackensenS3 1NovoNordisk,Inc.,Princeton,NJ,USA,2IMSHealth,FallsChurch,VA,USA,3University MedicalCentreHamburg-Eppendorf,Hamburg,GermanyOBJECTIVES: Congenitalhemophiliaisachronicdiseasediagnosedearlyinlifein severecases.Developmentofalloantibodyinhibitorstoexposuretoreplacement factorVIIIandIXconcentratesisthemostserioustreatmentcomplicationamong patients.Itisestimatedthat900-1,200patientshavecongenitalhemophiliawith inhibitors(CHwI)intheUnitedStates(US).Limitedresearchexiststodocumentthe experienceoftheircaregivers. METHODS: Aspartofacross-sectionalsurveyevaluatingpreservationofjointhealthandqualityoflifeamongCHwIpatientsinthe US,caregivers'experiencewasassessedviatheHAEMO-QoLinstrument,containinganopen-endedquestioncapturingcaregivers'burden.Responseswereassessedfortrendsandconsolidatedintocategories.Descriptiveanalyseswereemployedtoevaluatecaregivers'demographiccharacteristicsandtheirburdenwith thedisease. RESULTS: Questionnaireswerereceivedfrom60caregivers.Meanage ofpatientsundercarewas10.6years(SD 10.3).91.7%ofpatientsundercarehad HemophiliaAand68.3%hadhigh-titerinhibitors.Unemploymentasaresultof caregiverresponsibilitieswasreportedby28.1%.31.7%reportedhavinghealth insurancewithalifetimecap.Amongthese,73.7%were"somewhat"to"veryconcerned"aboutreachingthecap.Among55HAEMO-QoLrespondents26.3%were "considerably"to"verymuchso"botheredbythepatient'sdisease.Atotalof159 open-endedresponseswereprovidedby51caregiversconcerningtheirexperience aboutlivingwithhemophiliaanditstreatment:22.6%highlightedemotional stress,15.1%emphasizednancialburdenincludingworktime,12.0%involvedlost timeexcludingwork,and11.3%involvedmedicalmanagementissues. CONCLUSIONS: Onequarterofcaregiverswere"considerably"to"verymuch" botheredbythepatient'scondition.Theyreportedemotionalstressandnancial burdenasnegativeexperiencesdealingwithhemophiliaanditstreatment.Medicalmanagementisalsoanareaofconcern.Acaregiver'snegativeexperiencemight impactchildren'sHRQoL,whichshouldbeanalyzedinfutureresearch. PSY46 DISEASESYMPTOMS,TREATMENTSATISFACTION,ANDCOPINGSTRATEGIES INPATIENTSWITHLUPUSLerstrmK1,CrimmingsM2,GovoniM3,Nika•E3,MerrillJ4,IsenbergD5 1LupusEurope,Romford,UK,2LupusFoundationofAmerica,Washington,DC,USA,3UCB, Brussels,Belgium,4OklahomaMedicalResearchFoundation,OklahomaCity,OK,USA,5UniversityCollegeLondonHospitals,London,UKOBJECTIVES: Theimpactoflupusonpatients'everydaylivesisimportanttounderstand.Wethereforeinvestigatedtheperceptionsoflupuspatientsondisease symptomsandcopingstrategies. METHODS: Participants(USAandEurope)completedapatient-ledInternet-basedquestionnairebetweenAprilandMay2009. RESULTS: 914respondentsreportedhavinglupus(95%female,80%2050yearsof age).Themostcommonlupussymptomsreportedwerefatigue(92%)andarthritis, musclepain/weaknessortendinitis(89%).Antimalarialswerethemostcommonly usedlupustreatment(53%),followedbycorticosteroids(51%).Lessthanhalfofthe respondentsweresatised/verysatisedwiththeircurrenttreatment(49%).Approximatelyone-thirdofrespondentsstatedthatlupushadahighlysignicant effectontheircareers(37%),ontheirphysicalwell-being(34%)andoneveryday things(30%).Furthermore,approximatelyone-fthofrespondentsfeltthatlupus hadahighlysignicanteffectontheirsociallife(23%)andmentalhealth(20%), while11%reportedithadahighlysignicanteffectontheirrelationshipwiththeir family.Mostpatientsreportedtalkingtofamily(93%)andfriends(89%)aboutlupus orsharingwiththemtheyhadlupus(97%and95%,respectively);79%reported talkingtootherswithlupus.77%reporteditwaseasy/veryeasy'totalktotheir doctorsabouttheirsymptoms.Themostdifcultsymptomstoexplainweredepression(45%)andCNSproblems(34%).Themosthelpfulformsofsupportwere understandingfromfamilyandfriends(73%),abilitytospeakwithhealthcareprofessionals(63%),andbeingconnectedtootherlupuspatientsviatheInternet(50%). CONCLUSIONS: Lupusimpactsgreatlyonpatient'severydayliveswiththemost commonsymptomreportedbeingfatigue.Manypatientsremainunsatisedwith theirtreatment.Communicationwithfamilymembers,friends,healthcareprofessionalsandotherlupuspatientsareimportantstrategiestofacilitatecopingwith lupussymptoms. PSY47 WILLINGNESSTOPAY(WTP)FORWEIGHTLOSSCOACHING:RESULTSFROM THEPOWERTRIALAlaviR1,AppelL1,BrancatiF1,ClarkJ1,MohrP2,DaumitG1 1JohnsHopkinsUniversity,Baltimore,MD,USA,2CenterforMedicalTechnologyandPolicy, Baltimore,MD,USAOBJECTIVES: Efcaciousmedicalorbehavioralweight-lossprogramsarenotroutinelycoveredbyinsurance.Understandinghowpatientsvalueweightlossinterventionsbyassessingtheirwillingnesstopay(WTP)iscriticaltotranslatingndingsfromeffectivenesstrialsintopractice.Uponcompletionofarandomized clinicaltrialcomparinganin-personbehavioralinterventiontoaphone/email/ web-basedinterventionforweight-loss,westudiedparticipants'WTPandcharacteristicsassociatedwithWTPforcontinuingtheseprograms. METHODS: POWER HopkinsisanNHLBI-funded24-monthrandomizedclinicaltrialofweightloss comparingin-personvs.callcenterdirectedcoachinginterventionspluswebbasededucationandtrackingsupport.Attheend-of-studyvisit,weinterviewed participantsanduseddoublebounddichotomous-choicetoassesstheirWTP.Usinglinearregression,weexaminedbaselinecharacteristics(age,sex,income,race, education,andBMI)associatedwithWTP. RESULTS: Ofthe234adultscompleting thetrialtodate,206(88%)reportedonWTP.Meanagewas57years,61%were women,56.3%wereWhite,meanBMIatbaseline(2yearsearlier)was36.4kg/m2, 33%hadgraduateorprofessionaldegrees,61%hadannualfamilyincome $75,000.Participantsthoughttheirinterventionwasworth$70.9(95%CI60.3-A67VALUEINHEALTH14(2011)A1–A214

PAGE 76

81.4),andtheywerewillingtopay$49.6/month(95%CI44.8-54.3).Multivariate analysisidentiedraceasapredictorofWTP:WhitesreportedmeanWTP$18.2(CI 8.0-28.3)lessthanAfricanAmericans. CONCLUSIONS: Aftercompletingatwoyear trial,themajorityofparticipantsinbothgroupswerewillingtopaysimilarratesto othercommerciallyavailableweightlossproducts.ThesurprisingracialdifferencesinWTPseeninthishighSESgroupneedtobefurtherexaminedwhentrial weightoutcomesareavailable.Afterafreeorfullysubsidizedperiod,itmightbe reasonabletoimplementadirecttoconsumerorcostsharingmechanismtobetter translateeffectiveevidence-basedweightlossinterventionsintopractice. SystemicDisorders/Conditions–HealthCareUse&PolicyStudies PSY48 AWORKPLACEHEALTHPROGRAMFORBRITISHCOLUMBIAPUBLICSERVICE AGENCY(CANADA)TarrideJE1,HarringtonK2,BalfourR3,SimpsonP4,FoordL5,AndersonL6,LakeyWK7 1McMasterUniversity,Hamilton,ON,Canada,2AbbottCanada,StLaurent,QC,Canada,3Pacic BlueCross,Vancouver,BC,Canada,4GlaxoSmithKline,Mississauga,ON,Canada,5AstraZeneca, Mississauga,ON,Canada,6BritishColumbiaPublicServiceAgency,Victoria,BC,Canada,7BritishColumbiaPublicServiceAgency,Vancouver,BC,CanadaOBJECTIVES: ToevaluatetheMyHealthMatters!(MHM)program,amultifaceted workplaceinterventionrelyingoneducationandawareness,earlydetectionand diseasemanagementwithafocusonriskfactorsformetabolicsyndrome. METHODS: TheMHMprogramwasofferedto2,000publicservantsworkingin morethan30worksitesinBritishColumbia,Canada.TheMHMprogramincludeda healthriskassessmentcombinedwithanopportunitytoattendanon-sitescreeningandface-to-facecallbackvisitsandrelatedon-siteeducationalprograms.Clinicalandeconomicoutcomeswerecollectedovertimeinthisone-yearprospective studycoupledwithadministrativeandsurveydata. RESULTS: Fortythreepercent ofemployees(N 857)completedtheonlineHRAand23percent(N 447)attended theinitialclinicalvisitwiththenurse.Riskfactorsformetabolicsyndromewere identiedinmorethanhalfofthoseattendingtheclinicalvisit.Thenumberofrisk factorssignicantlydecreasedby15percentoversixmonths(N 141).Thecostper employeecompletingtheHRAwas$205whilethecostperemployeeattendingthe initialclinicalvisitwas$394.Eighty-twopercentofemployeeswouldrecommend theprogramtootheremployers. CONCLUSIONS: Thisstudysupportsthatworkplaceinterventionsarefeasible,sustainableandvaluedbyemployees.Assuch, thisstudyprovidesanewframeworkforimplementingandevaluatingworkplace interventionsfocussingonmetabolicdisorders. PSY49 PREDICTORSOFOBESITYTREATMENT(COUNSELINGORPHARMACOTHERAPY) INAMBULATORYSETTINGSMehtaHB,PatelJ,ParikhR,AbughoshS UniversityofHouston,Houston,TX,USAOBJECTIVES: OnethirdofUSadultsareobeseanditisprojectedthatby2030halfof USadultswillbeobese.Thestudyaimedtoidentifypredictorsofobesitytreatment inambulatorycaresettings. METHODS: NationalAmbulatoryCareSurvey (NAMCS)2006-07,across-sectionalnationallyrepresentativedata,wasusedforthe study.Aretrospectivecohortstudydesignwasemployed;obeseadults,age 20 yearsandBMI 30kg/m2orhavingobesitydiagnosis(ICD-9-CMcode:278),were includeinthecohort.Obesitytreatmentincludedeitherobesitycounseling(diet/ nutrition,exercise,weightreduction)orpharmacotherapy(anorexiantorlipase inhibitor).Predisposing,enablingandneedcharacteristicsasperAnderson'sbehavioralmodelwereincludedaspredictorvariables.Descriptivestatisticsand multivariatelogisticregressionwereconductedtoidentifyobesitytreatmentpredictorswhilepreservingcomplexsurveydesignofNAMCS. RESULTS: Totalof214 millionvisitsoccurredduring2006-07byobeseadults;ofwhich,32.66%visitsresultedinobesitytreatment.Factorspredictingobesitytreatmentwerereasonfor visit,preventivevisit(OR 2.23;95%CI 1.50-3.32)andchronicvisit(OR 1.93;95% CI 1.46-2.55)comparedtoacutevisit;timespentwithphysician, 24minutes (OR 2.67;95%CI 1.81-3.94)and13-24minutes(OR 1.89;95%CI 1.26-2.82)comparedto0-12minutes;highcomorbidity(OR 1.46;95%CI 1.13-1.89);morbidly obeseadultsi.e.BMI 40(OR 1.88;95%CI 1.52-2.34)andvisittoprimarycare physician(OR 2.38;95%CI 1.69-2.36)comparedtospecialist.Olderadultsaged 65(OR 0.98;95%CI 0.97-0.99)andsmokers(OR 0.52;0.39-0.69)hadlesslikelihoodofreceivingobesitytreatment.Gender,race,regionandinsurancestatus werenotsignicantpredictorsofobesitytreatment. CONCLUSIONS: Onlyone thirdvisitsresultedintoobesitytreatment.Reasonforvisit,timespentwithphysician,comorbidity,BMI 40,providerspecialty,ageandnonsmokingstatuswere signicantpredictorsofobesitytreatment.Futureresearchshouldidentifyreasons fortheseobserveddifferencesandeffortsshouldbetakentodeliverequitable access. PSY50 PRESCRIPTIONMONITORINGPROGRAMS’UTILIZATIONFlemingML,BarnerJC,ChandwaniH,OkoroT,WeberS UniversityofTexasatAustin,Austin,TX,USAOBJECTIVES: Prescriptionmonitoringprograms(PMPs)havebeenpurportedtobe aneffectivetooltocombatprescriptiondrugabuse.However,utilizationratesof PMPdatabyhealthcarepractitionersisrelativelylow.Theobjectivesofthisstudy weretodetermine1)theratesofPMPrequestbyprofessionalafliationper100,000 population;2)determinedifferencesinratesofrequestsbasedonPMPaccessibility (onlinevs.other);3)ifdifferencesexistinprofessionalafliationratesofrequests per100,000populationbyPMPhousingagency(lawenforcementagencyvs.health profession);4)collectdataonannualoperatingcostsofPMPsper100,000 population. METHODS: Thiswasacrosssectionalstudyemployingawebbased survey.Thesurveywasemailedtothe33operationalstatePMPadministrators duringDecember2010.DescriptivestatisticswereusedtodescribePMPstructure andutilizationrates.T-testsandManovawereusedtodeterminetheassociations betweenratesofrequestandPMPfeatures. RESULTS: Thepreliminaryresponse ratewas27%.PrescribersaccessedPMPdatamostfrequentlyamongallauthorized users,meanrequests 1,764,SD 3,106.Pharmacistmeanrequests 171,SD 220. TheT-testsresultsindicatethatthereisastatisticallysignicantlowerrequestrate forPMPdatawhenhousedbylawenforcementadministration,(t 2.5,p .04).On averagePMPannualcostswere$8,146per100,000population. CONCLUSIONS: Basedonpreliminaryresults,thePMPhousingentityhasanimpactonhealthcare professionals'utilization.Onlineaccessibilityisalsoassociatedwithanincreasein requesteddatareportsbyhealthcareprofessionals.Moreresearchisneededto determineotherfactorsassociatedwithPMPutilizationbyprescribersandpharmacists. PSY51 EFFECTIVENESSOFFDA’SNEWOVER-THE-COUNTERACETAMINOPHENLABEL WARNINGREQUIREMENTSINIMPROVINGCONSUMERRISKPERCEPTIONOF LIVERDAMAGEGoyalRK1,RajanSS2,EssienEJ2,SansgiryS2 1UniversityofNorthCarolina,ChapelHill,NC,USA,2UniversityofHouston,Houston,TX,USAOBJECTIVES: Thepurposeofthisstudywastoevaluatetheeffectivenessofthenew FDAmandatedover-the-counter(OTC)acetaminophenorgan-speciclabelwarnings,on:1)consumerriskperceptionofliverdamageassociatedwithacetaminophenuseand2)behavioralintentiontoperformprotectivebehavior. METHODS: In thiswithin-subjectexperimentalstudy,English-speakingadultsvisitingOTCsegmentsofselectedpharmacystoresinHoustonwereconvenientlyrecruited.Participantswererandomlyexposedtobotholdandnewlabelwarningsandtheir respectiveriskperception(measuredonavisualanalogscale,0%,norisk,100%, extremerisk)andbehavioralintention(measuredona7-pointLikertscale)were recordedusingavalidated,self-administeredquestionnaire.Descriptivestatistics andnon-parametricWilcoxonsigned-ranktestswereperformedusingSASstatisticalsoftware(v9.2)atapriorisignicancelevelof0.05. RESULTS: Atotalof200 responseswerecollectedwitharesponserateof56.81%.Meanageofthesample was42.68(SD15.30)years;48.5%ofrespondentsweremaleand52.7%werewhites. Amajorityofrespondents(74.4%)werenotawareofthenewwarnings;however,a majority(67.8%)hadpriorknowledgeoftherisk.Themeanriskperceptionscore forthenewlabelwarningswasfoundtobesignicantlyhigher(72.2%vs.65.9%, p .0001)ascomparedtothatoftheoldlabelwarnings.Similarly,theaverage intentionscoreforthenewlabelwarningswasfoundtobesignicantlyhigher(5.06 vs.4.86,p .0001)thanthatoftheoldlabelwarnings. CONCLUSIONS: Theresultsof thisstudyindicatethatthenewlabelwarningsmandatedbyFDAmaybeeffective inimprovingconsumerriskperceptionofpotentialliverdamageandmayencourageprotectivebehavior.Futurestudiesareessentialtoidentifytheimpactasactualchangesinconsumerbehaviorandsubsequentreductioninacetaminophenrelatedmorbidityandmortality. PSY52 POTENTIALDRUG-DRUGINTERACTIONS(DDIS)WITHPAINMEDICATIONS AMONGPATIENTSWITHBACKANDNECKPAINDIAGNOSESCATEGORIZED INTONOCICEPTIVE,NEUROPATHICORMIXEDPAINCOHORTSKozmaC1,BensonC2,PatelA2,ProvenzanoD3,SlatonT4,BiondiD2,KimM2 1IndependentResearchConsultant/AdjunctProfessor,UniversityofSouthCarolina,St.Helena Island,SC,USA,2Ortho-McNeilJanssenScienticAffairs,LLC,Raritan,NJ,USA,3OhioValley GeneralHospital,Pittsburgh,PA,USA,4TLSStats,WestColumbia,SC,USAOBJECTIVES: Todescribethepotentialfordrug-druginteractions(DDIs)among patientswithneckandbackpaindiagnosescategorizedintoneuropathic(NEURO), neuropathicandnociceptive(MIXED),nociceptiveonly(NOCI)orosteoarthritis(OA) cohorts. METHODS: ThePharMetricsUSNationalmanagedcaredatabasewasused toidentifycommerciallyinsuredpatients18to63yearsofagewithatleastone claimforanopioidanalgesicandapre-existingstudy-relateddiagnosis,whowere continuouslyeligibleforservicesfrom9/1/2006-8/31/2008.Patientswhohadnursinghomecareclaims,drug/alcoholabuse,malignancy,andspineprocedures,or pregnancy-relateddiagnoseswereexcluded.Overthe2-yearstudyperiod,thefrequencyofpatientswithatleast10daysofsimultaneousavailabilityofpainor pain-relatedprescriptionsandmedicationsthatareknowninhibitorsorinducers ofthecytochromeP450(CYP)metabolicpathwaywasexamined. RESULTS: The analysisidentied2,375NEURO,37,019MIXED,39,496NOCI,and6,124OApatients. Ahighprevalenceofcoexistingmedicalconditionswasfoundinallcohortswith 40%-74%ofpatientshavingdiagnosesinatleast8differentdiseasecategories. Basedonthe10-daysimultaneousdrugavailabilitycriterion,thepotentialforDDIs wereidentiedin26%ofallpatientsduringthe2-yearobservationperiod.This percentagewashighestintheMIXEDcohort(31%)andlowestintheNOCIcohort (20%).Overall,potentialinhibitorinteractionswerefoundin20%ofpatientsand potentialinducerinteractionswerefoundin11%.TheCYP-2D6substratestramadolandoxycodonewerethemostfrequentpotentialinhibitorinteractions(5.6% and5.9%ofpatients,respectively).PotentialinducerDDIsweremostcommonly foundintheCYP-1A2pathway(8%ofpatients). CONCLUSIONS: PotentialDDIsare commonamongpatientstakingpainmedication.Coexistingmedicalconditions andtheirtreatment,andvariationsinthemetabolismofdifferentpainmedicationscontributetothecomplexityofselectinganalgesicsthatwouldbeexpectedto effectivelytreatpain.A68VALUEINHEALTH14(2011)A1–A214

PAGE 77

PSY54 APPLICATIONOFTHENON-LINEARBLINDER-OAXACADECOMPOSITIONTO STUDYRACIAL/ETHNICDISPARITIESINANTI-OBESITYMEDICATIONUSEMehtaHB,RajanSS,AparasuRR,JohnsonML UniversityofHouston,Houston,TX,USAOBJECTIVES: TheBlinder-Oaxaca(BO)decompositionhelpstoidentifythecauses ofdisparitybetweentwogroupsbasedonobservedcharacteristics,andquanties theresidualdisparityduetounobservedheterogeneity.ThisstudyusesthenonlinearBOdecomposition,anextensionoftheoriginalBOdecomposition,toidentifyandquantifytheobservedcausesbehindtheracial/ethnicdisparityinantiobesitymedicationuseamongUSadultswhometthemedicalcriteriatoreceive thesemedications. METHODS: TheMedicalExpenditurePanelSurveydata,20022007,wasusedinthisretrospectivecross-sectionalstudy.Themedicalcriteriawas denedasadults(age 20years)withbodymassindex(BMI) 30,orwithBMI 27 andcomorbiditiessuchashypertension,cardiovasculardiseases,diabetesorhyperlipidemia.Thenon-linearBOdecompositionmethodwasusedtoidentifythe contributionofeachobservedsocio-demographicandhealthcharacteristicinexplainingtheracial/ethnicdisparityandtoestimatetheresidualunexplained disparity. RESULTS: Atotalof395,319,750adultsmetmedicalcriteriafrom20022007;ofwhich,0.61%usedanti-obesitymedications.Theproportionofmedically eligibleBlacksusinganti-obesitymedicationwas33%lowerthanWhites.ThenonlinearBOdecompositionmethodrevealedthatonlymaritalstatusexplainedthe racial/ethnicdisparityamongtheobservedcharacteristics.Blackswereyounger, hadahigherproportionofwomeninthesample,andhadahigherBMI,andthese characteristicswerethechiefdeterminantsofincreasedanti-obesitymedication use.Thus,thedecompositionestimatedthatiftheblackshadsimilarcharacteristicsasthewhitestheobserveddisparityinthesamplewouldhavebeen58% higher. CONCLUSIONS: Thedecompositionrevealedinterestingbutdisconcerting ndings.Noneofthedifferencesintheobservedsocio-demographicandhealth characteristics(exceptmaritalstatus)successfullyexplainedtheracialdisparity. Thus,furtherexploratorystudiesexaminingracialdifferencesinindividualbeliefs, behavioralpatternsandproviderprescriptionpatternsarevitaltounderstand thesedisparities. PSY55 NATIONALTRENDSINANTI-OBESITYMEDICATIONUSEFROM2002TO2007MehtaHB,RajanSS,AparasuRR,JohnsonML UniversityofHouston,Houston,TX,USAOBJECTIVES: TheUSranksrstamongalldevelopedcountriesforobesity/overweightprevalence.Anti-obesitymedicationsmayhelpaddressthispublichealth problem,butnationalutilizationdataforanti-obesitymedicationsislacking.The studyobjectivewastoexaminetrendsinanti-obesitymedicationuseamong adultswhometmedicalcriteria(MC)andunderstandtheextentofuseofsuch medicationsamongadultswhodidnotmeetMC. METHODS: AretrospectivecrosssectionalstudywasperformedusingtheMedicalExpenditurePanelSurveydata from2002-2007.Thestudycomprisedallnon-pregnantadults( 20yearsofage; N 210,829,728).MeetingtheMCforanti-obesitymedicationuseforanindividual wasdenedashavingBodyMassIndex(BMI) 30(clinicallyobese)orhavingBMI 27withcomorbiditiessuchashypertension,cardiovasculardisease,diabetes andhyperlipidemia.Anti-obesitymedicationswereidentiedfromprescription medicinelesusingnationaldrugcodesanddrugnames.ProportionsofantiobesitymedicationusersamongadultswhometMCwerecalculatedforeachyear andlogisticregressionwasperformedtotestutilizationtrends.Descriptiveanalysiswasconductedtoassessproportionofadultsusinganti-obesitymedications whodidnotmeetMC. RESULTS: OftheadultswhometMC(N 65,886,625),0.52% weretakinganti-obesitymedications.Analysisofmultiyeardatashowednostatisticallysignicanttrend( pvalue 0.25)inanti-obesitymedicationuseamong adultswhometMCfrom2002to2007.OfallUSadultswhoweretakinganti-obesity medications(N 542,297),31.98%didnotmeetMCforuse. CONCLUSIONS: Less thanonepercentofeligibleadultsusedanti-obesitydrugsineachyearandantiobesitymedicationusedidnotchangefrom2002to2007.Almostonethirdof anti-obesitymedicationusersdidnotmeetMC.Giventheassociatedside-effects forthesemedications,furtherresearchisneededtounderstandtheuseofantiobesitymedicationsamongadultswhodonotmeetMC. PSY56 CONTROLLEDSUBSTANCEPRESCRIBINGANDUTILIZATIONOFPRESCRIPTION DRUGMONITORINGPROGRAMSTalbertJ,PerinN,BlumenscheinK,FreemanPR UniversityofKentucky,Lexington,KY,USAOBJECTIVES: Over40stateshaveprescriptiondrugmonitoringprograms(PDMP)to monitortheutilizationofcontrolledsubstances(CS)withthegoalofreducingdrug abuseanddiversion.LittleisknownabouthowPDMPsareactuallyusedbyhealth careprofessionals.Thepurposeofthispaperistoreviewhowprescribersutilize PDMPsandtoassesstheimpactofaPDMPonCSprescribing. METHODS: Thepaper usesdatafromtheKentuckyAllSchedulePrescriptionElectronicReportingProgram(KASPER)from2002-2009.WereviewCSprescribingandKASPERutilization usingdescriptivestatistics,visualization,andt-test,controllingforcharacteristics ofKASPERusers,thenumberofKASPERrequestsforeachusertype,andgeographicdistribution.StatisticalanalysisisconductedinSTATAv11.0. RESULTS: CS prescriptionsissuedbyKentuckyprescribersincreasedfrom8.4millionin2002to over11millionin2009.Duringthisperiod,KASPERsystemrequestsincreasedfrom 100,000in2002tomorethan530,000in2009.Yet,mostCSprescribers(64%)donot haveKASPERaccounts.ComparingKASPERusersacrossgeographicregions,rural areashaveahigher(49%)rateofKASPERusethanurbanareas(32%,p .05).ForCS prescribing,thetypicalKASPERaccountholderissued1,665prescriptionsayearin 2009,comparedto250prescriptionsissuedbynon-accountholders.ThetopCS prescriberin2009issuedover41,000prescriptions,anincreaseof60%fromthetop prescriberin2005.Finally,CSprescribinghasbecomeskewed,with10%ofprescribersissuing95%ofallCSprescriptions. CONCLUSIONS: Since2004,CSprescribinginKentuckyhasincreasedby40%andKASPERsystemrequestshaveincreasedby400%.ThemostinterestingndingisthatthemajorityofgrowthinCS prescribingisfromarelativelysmallnumberofprescribers.Whetherthisndingis adirectresultofthePDMPisaquestionforfurtherresearch. PSY57 DOESCLINICALDIAGNOSISOFOBESITYINCREASETHELIKELIHOODOF RECEIVINGOBESITYRELATEDCOUNSELING?A2005-2008NAMCS-NHAMCS ANALYSISKadakiaA,MehtaH,RajanSS,AbughoshS UniversityofHouston,Houston,TX,USAOBJECTIVES: Studieshaveshownthatdespitethenationalguidelinesrecommendingcounselingforobesepatients,counselingratesduringoutpatientvisitsremain low.Wedeterminedthetrendsforobesitycounseling(diet,exerciseandweight reduction)fromyear2005to2008andtheassociationbetweenbeingclinically diagnosedwithobesityandbeingcounseled,forobeseadultsduringtheiroutpatientvisits.Wealsodeterminedifclinicaldiagnosisofobesityvariedwithobesity classwhichisclassiedbyBMI. METHODS: Weusedthedatafrom2005to2008 NationalAmbulatoryMedicalCareSurvey(NAMCS)andNationalHospitalAmbulatoryMedicalCareSurvey(NHAMCS).ThestudyincludedalladultshavingBody MassIndex(BMI) 30.Theoutcomeofinterestwaswhetherthepatientreceived obesitycounselingornot. RESULTS: 478millionpatientshadBMI 30ofwhich only6%wereclinicallydiagnosedwithobesity,and31%receivedcounseling.There wasnostatisticallysignicantchangeincounselingforobeseadultsfromtheyear 2005to2008.Patientsclinicallydiagnosedwithobesityhad7.7times(95%CI:5.311.1)greateroddsofreceivingcounselingascomparedtoundiagnosedobesepatientsaftercontrollingforsociodemographicandclinicalfactors.Moreover,classII (35.5 BMI 39.9)andclassIII(BMI 40)obesepatientshad1.8(95%CI:1.4-2.4)and 4.5(95%CI:3.4-6.0)timesgreateroddsofbeingdiagnosedwithobesityascompared toclassI(30 BMI 34.4)obesepatients. CONCLUSIONS: Beingclinicallydiagnosedwithobesityisanindicatorofaccesstocare,andthepatient'sinitiativesto seekobesity-relatedcare.Ourndingoflowratesofcounselingfordiet,exerciseor weightreductionduringvisitsbyobesepatientsindicatesthatobesityiscommonly under-diagnosed.Interventionsshouldfocusonpatientawarenessstrategiesto increasetherateofclinicaldiagnosisandhencecounseling. PSY58 DISPARITYINOBESITYMANAGEMENT:UNDER-TREATMENTAMONGELDERLY POPULATIONShahAJ,PatelJG,RajanSS,AbughoshS,MasilamaniS UniversityofHouston,Houston,TX,USAOBJECTIVES: Obesityisamajorcauseofmorbidityandmortalityandarising epidemicworldwide.Despiteguidelinerecommendations,manyobesepatients failtoreceiveadequatemedicalcare.Theobjectivesofthestudyweretoexamine predictorsofobesitytreatmentandassessdifferencesinreceivingtreatmentbetweenvariousagegroups. METHODS: Thestudyuseddatafrom2006-2008NationalAmbulatoryMedicalCareSurveyandNationalHospitalAmbulatoryMedical CareSurvey.Adultswithage 20yearswereincludedandobesitywasdenedas havingaBMI 30.Obesitytreatmentwasdenedasthereceiptofmedication(like Sibutramine,Orlistat,etc.)and/orobesitycounselingintermsofweightreduction, exercise,ordietcounseling.Splineswereconstructedtoassesschangesinobesity treatmentwithage.Descriptivestatisticsandmultivariatelogisticregressionwere usedtodeterminepredictorsofreceivingobesitytreatment. RESULTS: Outof3.1 billionvisits,29.2%(95%CI:28.130.3%)wereobese.Amongthose,31.4%(95%CI: 28.234.6%)betweenages20-39;32.5%(95%CI:29.635.4%)betweenages40-64and 25.4%(95%CI:22.228.7%)above64yearsreceivedtreatment.Inmultivariatemodel, aunitincreaseinageincreasedthelikelihoodtoreceivetreatmentby1.8%(95%CI: 0.13.6%)withintheagegroup20-39years;decreasedthelikelihoodby1.4%(95%CI: 2.30.4%)within40-64years;anddecreasedthelikelihoodby3.4%(95%CI:5.01.7%) above64years.Primarycarephysicians(OR1.66,95%CI:1.38-1.99),wereapproximately70%morelikelytoprovidetreatmentcomparedtospecialists.Individuals withcomorbiditiesweremorelikelytoreceivetreatment(OR-1.37,95%CI:1.171.60).Asignicantdifferenceacrossgeographicalregionwasalsoobserved. CONCLUSIONS: Studysuggestsdifferencesinobesitymanagementacrossage groups.Olderadultswerelesslikelytoreceiveobesitytreatment.Furtherresearch isneededtoensurethatqualitycareisprovidedacrossallagegroups. PSY59 IMPACTOFOBESITYONAGE-SPECIFICPREVALENCEOFCHRONICDISEASESIN THEUNITEDSTATESPatelJG,ShahA,RajanSS,AparasuRR,JohnsonML UniversityofHouston,Houston,TX,USAOBJECTIVES: Toexamineprevalenceofchroniccomorbiddiseasesamongobese adultsintheUnitedStates METHODS: MedicalExpenditurePanelSurveyData from20052007wasusedfortheanalysis.Adultswithage 20yearsandBMI 30, weredenedobese.Unadjustedage-specicprevalenceofhypertension,diabetes, dyslipidemia,CHD,osteoarthritisandasthmawereobtainedamongobese,nonobeseandentirepopulationusingdescriptiveanalysis.Adjustedprevalenceestimateswereobtainedusingregressionanalysisaftercontrollingforsocio-demographicfactors.Age-specicprevalenceratios(relativerisk)wereobtainedbyA69VALUEINHEALTH14(2011)A1–A214

PAGE 78

takingratiosofpredictedprevalenceratesforobeseversusnon-obeseindividuals. Bootstrapped95%condenceintervalsweregeneratedforprevalenceratios. RESULTS: Amongobeseadultstheunadjustedprevalenceofhypertensionwas (34.40%),followedbydyslipidemia(21.87%),diabetes(16.34%)andasthma(6.92%). Adjustedprevalenceofchronicdiseaseswasalwayshigheramongobeseascomparedtonon-obeseandtheentirepopulation.Theprevalenceratiofordiabetes was3.06(95%C.I:2.823.30)attheageof20andwas2.20(95%C.I:2.092.31)at70 years.Atanyage,obesityincreasesthelikelihoodoftheseconditionsbyatleast 50%ascomparedtonon-obeseindividuals. CONCLUSIONS: Prevalenceratiosindicatethatobesityhashighestimpactonprevalenceofdiabetes;followedbyhypertension,osteoarthritis,dyslipidemia.Studyndingssuggestthatobesityisnot onlyadisease,butmayalsobeacauseforotherchronicdisorders.Thereisaneed todevelopeffectiveinterventionstocombatobesityandthusminimizeitsimpact onotherdiseasesintheUnitedStates. PSY60 ECONOMICCONSEQUENCESOFUNDER-UTILIZATIONWITHTUMORNECROSIS FACTORINHIBITORSINRHEUMATOIDARTHRITISPATIENTSCarterC1,GunnarssonC2,RizzoJ3,BolgeS1,InghamM1 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2S2StatisticalSolutions,Inc., Cincinnati,OH,USA,3StonyBrookUniversity,StonyBrook,NY,USAOBJECTIVES: AdherencetoFoodandDrugAdministrationrecommendedadministrationwithadalimumab,etanercept,oriniximabmaybevariable.Suboptimal adherencepatternsmaycontributetohealthcarecostincreases.Thisstudyestimatedincrementalhealthcarecostsoftumornecrosisfactorinhibitor(anti-TNF) under-utilizationfromamanagedcareperspective. METHODS: Medical/pharmacy claimsfromtheLifeLinkHealthPlandatabasewereused.Inclusioncriteriaincluded:indexanti-TNFstartedJanuary1,2004-December31,2007,age 18years,2 pre-indexrheumatoidarthritisdiagnosiscodes(ICD-9code714.xx), 365daysof indexdrug,continuousenrollment6/12monthspre-/post-index.Exclusioncriteria included:switchingbiologicspost-indexandselectedinammatoryconditions. Under-utilizationeventsweredenedasprescriptions/infusionswith10%less thanrecommendeddosingforadalimumab(40mgeveryotherweek/weeklywith/ withoutmethotrexate),etanercept(50mgweekly),oriniximab(3mg/kgdoseand maintenanceinfusioninterval 56days).Incrementalincreasesinhealthcare costsforpatientswithunder-utilization,comparedtoreceivingrecommended dosing,wereestimatedusingcostregressionmodelscontrollingforrell/infusion intervals.Modelswereestimatedfora12-monthtimehorizonanduntilindexdrug discontinuationorlossofenrollment. RESULTS: Atotalof4,586RApatientsreceivingadalimumab(N 1,255;27,540prescriptions),etanercept(N 2,242;48,517prescriptions),oriniximab(N 1,089;19,656infusions)wereincluded.Meanlengths oftime(days)patientsreceivedadalimumab(856),etanercept(881),andiniximab (903)werecomparable.Proportionofunder-utilizationeventswere16%,39%,and 2%foradalimumab,etanercept,andiniximab,respectively.Adalimumaboretanerceptunder-utilizationwassignicantlyassociatedwithincrementalincreasedhealthcarecosts($2,352and$879;p 0.01)for12monthsandthroughend ofdata($4,677and$3,806;p 0.01).Iniximabunder-utilizationwasinfrequentand notassociatedwithincreasedhealthcarecosts. CONCLUSIONS: Inthisanalysis, adalimumaboretanerceptunder-utilizationwasassociatedwithincreasedtotal healthcarecosts;however,iniximabunder-utilizationdidnothaveasimilarresult.Additionalresearchassessingclinicalconsequencesofunder-utilizationis warranted. PSY61 CHARACTERISTICSOFGOLIMUMABUTILIZATIONANDCOSTSINASPECIALTY PHARMACYPROVIDER(SPP)SETTINGEllisL1,TandonN1,ModyS1,DoshiD1,BaekC2,GunnarssonC3 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2DiplomatPharmacy,WartzCreek, MI,USA,3S2StatisticalSolutions,Inc.,Cincinnati,OH,USAOBJECTIVES: Golimumabisa50milligram(mg),oncemonthly,injectedanti-tumor necrosisfactoralphatherapyfortreatmentofrheumatoidarthritis,ankylosing spondylitisandpsoriaticarthritis.Thisstudycalculatedexpectedcostsofannual golimumabtherapybaseduponobserveddosingpatternswithinaSPPpopulation duringtherst16monthsofgolimumabavailability. METHODS: Pharmacyand correspondingeNAVIGATORTMpatientcaremanagementdatawerecollectedby DiplomatSpecialtyPharmacy,Flint,MIforadultpatientswithagolimumabprescriptionbetween4/24/2009and8/24/2010.Datawereanalyzedstatisticallyandare reportedasn,%,mean standarddeviation(SD)andmedian.Costsweremodeled inUSdollarsusingthewholesaleacquisitioncost(WAC;effective6/9/2010)of $1,731.48per50mg. RESULTS: Thestudyincluded89patients.Themajoritywere female(65%);age 45years(69%);andreportedpriorbiologicuse(56%).A50mg golimumabdosewasdispensedin100%ofpatientsand100%ofalldoses.The mean( SD)intervalbetweengolimumabdoseswas32.0 14.1daysandthemedianwas28days.Themeangolimumabdosingintervalinpatientsreportingbiologicusepriortogolimumabinitiationwas32.9 15.9days(mean SD)andwas similartothemeandosingintervalobservedinpatientsreportingnobiologicuse priortogolimumabinitiation(mean SD:31.1 12.1days;p 0.15;NS).Basedupon modelingoftheseearlyobservations,theaveragegolimumabpatientwillutilize approximately11.4dosesofgolimumabannuallyatacostof$19,739(WAC). CONCLUSIONS: InthisSPPpopulation,allpatientsreceived50mgofgolimumab. Themeanandmediantimesbetweendistributionofgolimumabdoseswere32 daysand28days,respectively.Baseduponthedosinganddistributionpatterns observed,theestimatedaverageannualperpatientcostofgolimumabwouldbe $19,739.Golimumabutilizationmaybesimilarforpatientsregardlessofprioruse ofbiologictherapies. PSY62 PERCEIVEDBENEFITSANDDISADVANTAGESOFINTRAVENOUS(IV)BIOLOGIC THERAPYAMONGPATIENTSWITHIMMUNOLOGYCONDITIONSBolgeS,VanderpoelJ,EldridgeH,ModyS,LoandJ,InghamM CentocorOrthoBiotechServices,LLC,Horsham,PA,USAOBJECTIVES: Toidentifyperceivedbenetsanddisadvantagesofintravenousinfusion(IV)biologictherapyamongpatientswithimmunologyconditionscurrently treatedwithIVbiologicmedication. METHODS: Semi-structuredtelephoneinterviewswereconductedwithpatientsself-reportingadiagnosisofankylosingspondylitis,Crohn'sdisease,psoriasis,psoriaticarthritis,rheumatoidarthritis,orulcerativecolitisandcurrentlyreceivingIVbiologictherapy.Studyprotocoland questionnairewereapprovedbyanindependentinstitutionalreviewboard.PatientsratedsatisfactionwithcurrentIVmedicationona7-pointLikertscale,where 1 notatallsatisedand7 verysatised.Patientsalsodiscussedbenetsand disadvantagesofIVbiologictherapyandreasonsforIVpreference. RESULTS: 405 interviewswereconducted.Meansatisfactionwas6.1;77%ratedsatisfactionas6 or7.ThemostfrequentlydescribedbenetsofIVtherapyrelatedtohealthcare professionalmonitoringandoversightattimeofinfusion.MorethanhalfofpatientsalsoexperienceasocialbenetofIVadministration,includingtalkingto otherpatientsaboutexperiences(56%)andtying-inotheractivitieswithinfusion facilityvisits(55%).Mostcommonlydescribeddisadvantagesofinfusionweredurationofinfusion(41%)andschedulingissues(23%).OfcurrentIVusers,most(82%, n 332)preferanIVmedicationtoasubcutaneousinjection.Themostcommon reasonsforIVpreferencewere:notwantingtoself-inject(43%),lessfrequentdosing(34%),andpreferenceforhealthcareprofessionaladministration(24%).Satisfactionwithmedicationandperceivedbenetsvariedsomewhatbydemographics,immunologiccondition,andfactorsrelatedtotreatment. CONCLUSIONS: CurrentIVbiologicusersarehighlysatisedwiththeirmedications.Patientsperceivetheadditionalopportunityforhealthcareproviderinteractionatinfusion facilitiesasabenetofthismodeofadministration.Theseresultssupporttheneed forcontinuedpatientaccesstoIVtherapeuticoptionsandshareddecision-making betweenpatientsandphysicianswhenselectingbiologictreatment. PSY63 LONGITUDINALANALYSISOFINFLIXIMABDOSINGANDINFUSIONINTERVALS ACROSS30INFUSIONSCarterC1,HaasS2,GunnarssonC2,RizzoJ3,InghamM1 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2S2StatisticalSolutions,Inc., Cincinnati,OH,USA,3StonyBrookUniversity,StonyBrook,NY,USAOBJECTIVES: Iniximab(IFX)isaninfusibleanti-tumornecrosisfactor(anti-TNF) drugusedinthetreatmentofrheumatoidarthritis(RA),withFoodandDrugAdministration(FDA)recommendedadministrationsof3mg/kgatweeks0,2,6,and every8weeksthereafter.Doseincreasesupto10mg/kgordecreasesininfusion intervalstoevery4weeksmaybeappropriatebasedonindividualresponse.Limiteddataareavailablepresentingweight-baseddosing,totalquantityadministered,andinfusionintervalssimultaneouslyoverthecourseof30infusions.The objectiveofthisstudywastocalculateweight-baseddosing,totalquantityinfused, andinfusionintervalsforRApatientsreceivingIFX. METHODS: Anevent-level analysiswasconductedusingmedical/pharmacyclaimsfromtheIMSLifeLink HealthPlandatabase.Inclusioncriteriaincluded:IFXinitiationJanuary1,2004December31,2007(i.e.,indexdate);patientage 18yearsold;2RAdiagnosiscodes (ICD-9714.xx);and365daysofIFXpersistence(i.e.,numberofdaysbetweenrst andlastIFXtreatment).Patientswereexcludediftheyhad:psoriaticarthritis (ICD-9696.0),psoriasis(ICD-9696.1),ulcerativecolitis(ICD-9556.xx),Crohn'sdisease(ICD-9555.xx),orankylosingspondylitis(ICD-9720.0);evidenceofanyantiTNFduringthe6monthspriortoindexdate;orevidenceoftakingabataceptor rituximabwhileonIFX. RESULTS: Therewere19,656IFXinfusionevents(N 1,089) identied.Themedianweight-baseddosesspanned3.0-4.2mg/kg.Overallmedian quantityinfusedateachinfusionspanned330-477mg.Medianinfusionintervals spanned50-56daysforinfusions4-20.Themedianinfusionintervalsspanned 44-50daysforinfusions21-30. CONCLUSIONS: TheobservedIFXadministration schedulewasconsistentwithFDA-approvedprescribingofweight-baseddosing andinfusionintervalsoverthecourseof30infusions.Thesedatacontributetothe publishedliteraturebydescribingaconsistentreal-worldadministrationschedule overalongerperiodoftimecomparedtootherpublishedstudies. PSY64 GECENTRICITYELECTRONICMEDICALRECORDSSTUDY:COMORBIDITIES ANDBIOLOGICEXPERIENCEAMONGPATIENTSRECEIVINGGOLIMUMABTandonN1,CarterC1,HaasS2,GunnarssonC2 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2S2StatisticalSolutions,Inc., Cincinnati,OH,USAOBJECTIVES: Golimumabistherstandonlymonthlysubcutaneousfullyhuman anti-TNFapprovedinthetreatmentofrheumatoidarthritis(RA),psoriaticarthritis (PsA),andankylosingspondylitis(AS).Thisstudyevaluatedcomorbiditiesand pre-indexbiologicuseamongpatientswhoreceivedgolimumabintheGECentricity¨electronicmedicalrecords(EMR)database. METHODS: LongitudinalEMRdata werecollectedfromoutpatientpracticesintheUnitedStates.Itcontainsclinical informationonover15millionpatientsthroughSeptember2010.Patientswere selectedbasedonthefollowingcriteria: 1ICD-9diagnosisofRA,PsA,orAS; 18 yearsofageatthetimeoftherstdiagnosis;data6-monthspriorand3-months afteraprescriptionrecordoftherstbiologic,andatleastoneprescriptionrecord forgolimumab.Biologicexperiencewasdenedasuseofabiologicwithin6 monthspriortotherstgolimumabrecord. RESULTS: Atotalof153[118(RA);20 (PsA);15(AS)]patientsreceivinggolimumabwereidentiedasmeetingallthe inclusioncriteria.Themeanagewas49yearsand75%werefemale;101(66%)A70VALUEINHEALTH14(2011)A1–A214

PAGE 79

patientswerebiologic-experienced,and52(34%)werebiologic-na•vebeforeinitiatinggolimumab.Ahigherpercentageoffemalepatientswereinthebio-experiencedcategory(70%vs.55%).Osteoarthritis(27%),hypertension(24%),dyslipidemia(17%),anddepressivedisorders(14%)werethemostcommoncomorbidities priortothestartofgolimumab.Ahigherrateofdepressivedisorderwasobserved inthebiologic-experiencedgroup.BaselinemeanC-reactiveproteintestvalues werealsohigherinthebiologic-experiencedgroup(3.69vs.0.97).Biologic-experiencedpatientsongolimumabwereswitchedmostlyfromadalimumab(n 42)and etanercept(n 25). CONCLUSIONS: InthislongitudinalEMR,patientsreceiving golimumabweremorelikelytohavepriorbiologicexperience.Biologic-experiencedpatientsappearedtohavehigherC-reactiveproteintestvaluesandgreater ratesofdepressivedisordersthantheirbiologic-na•vecounterparts. PSY65 CHARACTERISTICSOFGOLIMUMABUTILIZATIONINALARGENATIONAL PAYERDATABASETandonN1,SaulnierA2,GunnarssonC2 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2S2StatisticalSolutions,Inc., Cincinnati,OH,USAOBJECTIVES: Golimumab(GLM)isamonthlyself-injectedanti-tumornecrosisfactoralphatherapyprovidingonce-monthlydosingforpatientswithrheumatoid arthritis(RA),ankylosingspondylitis(AS),andpsoriaticarthritis(PsA).Thisstudy assessesthebaselinecharacteristicsandutilizationpatternsofpatientswhoreceivedGLM. METHODS: WeperformedaretrospectivedatabaseanalysisofThe MarketScan¨ResearchDatabasefromThomsonReuters.Thisdatabasecontains individual-level,de-identied,healthcareclaimsinformationfromemployers, healthplans,hospitals,Medicare,andMedicaid.Atotalof29,774patientsinthis databasehadadiagnosisofeitherRA,PsAorASandatleastonebiologiconrecord andmetthefollowinginclusioncriteria: 18yearsofageatthetimeoftherst diagnosis.Fromthissample,atotalof174patientshadatleastoneprescription recordforGLM. RESULTS: Atotalof174patientsreceivingGLMwereidentiedas meetingalltheinclusioncriteria;with128(RA),30(PSA),and16(AS).Themeanage was48yearsand75%ofthesamplewasfemale.Atotalof155(89%)patientswere bio-experiencedand19(11%)werebio-na•vebeforeinitiatinggolimumab.Atotalof 111patientsreceivedatleasttwoGLMdoses.OfthepatientswithtwoormoreGLM doses,themedianandmean SDdosingintervalwas29.5daysand33.65 15.56 days.Whenlookingatbiologicna•vepatientsthemedianandmean SDdosing intervalwas30daysand35.37 17.63daysversusbiologicexperiencedpatients withadosingintervalof29daysand33.15 15.00days. CONCLUSIONS: Inthe MarketScandatabase themajorityofpatientswithaprescriptionforGLMwas femaleandhadpriorbiologicexperience.GLMmedianandmeandoseswere29.5 and33.37daysrespectively.Previousbiologicsexperiencedidnotsignicantly changetheGLMdosingpatterns. PSY66 BIOLOGICEXPERIENCEANDDOSINGOFGOLIMUMABPATIENTSINMANAGED CARECarterC1,TandonN1,SmithD2 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2IMSHealthIncorporated, Watertown,MA,USAOBJECTIVES: Golimumab,aneweranti-tumornecrosisfactoragentusedinpatients withrheumatoidarthritis,psoriaticarthritis,andankylosingspondylitis,hasrecommendeddosingof50mgoncemonthly.Theobjectiveofthisstudywastodescribethe biologicexperienceanddosingforpatientsusinggolimumabinthemanagedcare setting. METHODS: TheIMSLifeLinkHealthPlandatabase( 100managedcare plans)wasutilizedtoidentifypatientsaged 18yearsatindexandhavinganindex golimumabpharmacyclaimstartedbetween4/24/2009(productapproval)and10/31/ 2009.Patientswererequiredtohave24monthspre-and 2monthspost-index continuousenrollmentwith 1RA,PsA,orASICD-9diagnosiscode.Biologicexperiencewasassessedforthepre-indexperiod.Dosingwasassessedthroughendofdata orlossofenrollment. RESULTS: Atotalof282patientsreceivinggolimumabwere identied;72%werefemale;meanagewas52years.Themajority(73%)ofpatientshad pre-indexbiologicexperience.Amongthebiologic-experienced,60%received1 uniquebiologic,33%received2uniquebiologics,and7%received3 uniquebiologics beforegolimumab.Golimumabpatientshadexperiencewithvariouscombinationsof abatacept,adalimumab,certolizumab,etanercept,andiniximab.Adalimumabalone andetanerceptaloneweremostfrequentlyusedpriortogolimumab.Thegolimumab doseateachoftherstsixprescriptionllswas50mgforover97%ofpatients.The mean(median)daysbetweenllsspanned29-33(29-30)days. CONCLUSIONS: The majorityofpatientsreceivinggolimumabwerebiologic-experienced.Observeddosing wasconsistentwithprescribingrecommendations.Consistencyindosingwasobservedovertherstsixprescriptions.Despitethehistoryofbiologicuse,golimumab patientsdidnothaveanapparentincreaseddoserequirementuponinitiation.Further researchisnecessarytoconrmthesendingsinalargersamplesizeoveralonger durationoffollow-up. PSY67 REASONSFORINITIATINGINTRAVENOUSBIOLOGICTHERAPYAMONG PATIENTSWITHIMMUNOLOGYCONDITIONS:SUBSETANALYSISOFPRIOR SUBCUTANEOUSINJECTION(SQ)USERSANDIMPLICATIONSFORSHARED DECISIONMAKINGBolgeS,VanderpoelJ,EldridgeH,ModyS,LoandJ,InghamM CentocorOrthoBiotechServices,LLC,Horsham,PA,USAOBJECTIVES: Tounderstandthereasonsforinitiatinganintravenous(IV)biologic therapyamongpriorsubcutaneousinjection(SQ)users,andpatientsatisfaction beforeandafterswitching. METHODS: Semi-structuredtelephoneinterviewswere conductedwith405immunologypatientscurrentlyreceivingIVbiologictherapy. Patientsratedtheirlevelofsatisfactionwithcurrentorpriormedicationona 7-pointLikertscale(7 Verysatised;1 notatallsatised)andreportedreasons forswitchingfromSQtherapy. RESULTS: Morethanathird(37%)ofsurveyedIV biologicpatientspreviouslyusedSQadministration.Ofthesepatients,overall meansatisfactionwithSQwas3.8,with32%ratingtheSQexperienceasa1or2. CurrentIVsatisfactionratingsamongpriorSQuserswas6.2(vs.6.1forallpatients surveyed).OfpriorSQusers,26%didnotself-administertheirinjections,most frequentlydueto:dislikeofneedles;lackofcondenceinownabilitytoadminister theinjectioncorrectly;and/orphysicalinabilitytohandlethesyringe.Ofthosewho didnotself-administerSQ,39%wenttoaphysician'sofceforadministration. PrimaryreasonsforswitchingfromSQtoIVadministrationincludedlackofefcacy(81%),sideeffects(21%),cost(10%)anddislikeofselfinjections(9%). CONCLUSIONS: Inthisanalysis,currentIVbiologicusersappearhighlysatised withthismedication.DissatisedSQusersthatdidnotself-administermayoffset potential"convenience"advantagesofaSQ.TheyalsoswitchedtoIVmostlyfor lossofefcacyandappeartobesatisedwithanIV.Giventhepotentialforpatient preferencedifferences,thereappearsaneedforexpandedtreatmentchoices.InvolvingthepatientinshareddecisionmakingandprovidingaccesstobothIVand SQmodesofadministrationmaybeimportanttooptimizingpatientsatisfaction. PSY68 DEFINITIONSOFANTI-TNFDISCONTINUATIONMAYIMPACT UNDERSTANDINGOFREAL-WORLDUTILIZATIONPATTERNSSchmeichel-MuellerC1,BuysmanE2,BolgeS1,InghamM1,McKenzieRS1 1CentocorOrthoBiotechServices,LLC,Horsham,PA,USA,2i3Innovus,EdenPrairie,MN,USAOBJECTIVES: Toevaluatetreatmentpatternswhendifferentdenitionsofdiscontinuationareused.Anti-tumornecrosisfactortherapiesincludingadalimumab(ADA), etanercept(ETA)andiniximab(IFX)areusedtotreatrheumatoidarthritis(RA).Examiningdiscontinuationratescanhelpunderstandingofreal-worldtreatmentpatterns.Differentdenitionsofdiscontinuationrateshavebeenreportedelsewhere. METHODS: Databetween01/2005-06/2006wereextractedfromthei3Innovusdatabase.Inclusioncriteriawereaged 18, 1claimforRA,andnoevidenceofpre-index biologicuseinthesixmonthsprior.Patientswerefollowedfor24months.Discontinuationwasdenedasagapintherapy 60or 365daysfollowingthelastdayssupply. RESULTS: Totalof1,780patientswereanalyzed:ADA 601(33.8%);ETA 785(44.1%); IFX 394(22.1%).Ifdiscontinuationwasdenedasagapintherapyof 60days,57.2% ofpatientstreatedwithADAdiscontinued,57.5%ofpatientstreatedwithETAand 37.6%ofIFXpatientsdiscontinued.Ifdiscontinuationwasdenedasagapintherapy of 365days,22.5%ofADApatientsdiscontinued,17.3%ofETApatientsand18%of IFXpatientsdiscontinued.Ofthosedenedasdiscontinuers'afteragapoftherapyof 60days,49.7%ofADApatients,62.1%ofETApatientsand12.8%ofIFXpatients restartedtheirindextherapyonaverage146.8,146.5,and302.4daysfromthetimeof deneddiscontinuation'.SignicantlyfewerIFXpatientsrestartedtheirindextherapy(P 0.0001)andthetimefromdeneddiscontinuationtorestartwaslonger (p 0.0001). CONCLUSIONS: Thisanalysisdemonstratesthatdifferentdiscontinuationratesareobservedwhendifferentdenitionsofdiscontinuationareemployed. Thismayimpacttheunderstandingofreal-worldprescribingpatterns.Thedataalso suggeststhatpatientstreatedwithADAandETAexperiencelengthygapsintherapy. Futureresearchisneededtoexaminegapsintherapyonclinicalandhealtheconomic outcomes. PSY69 MODELINGTHEIMPACTOFREFILLORADMINISTRATIONGAPSONPATIENT DRUGLEVELSOFTHETNF-ALPHAINHIBITORSETANERCEPTANDINFLIXIMABInghamM,CarterC,BolgeS,ComisarC CentocorOrthoBiotechServices,LLC,Horsham,PA,USAOBJECTIVES: Prioretanerceptdrugutilizationstudieshavereported,thatamong patientswithagapinobservedrelltimeexceedingtherecommendedrelltime, 31%-44%ofpatientsexperiencedgapsgreaterthan2weeks.Meangaps(assessedat eachindividualrellperiod)rangedfrom19to37days.Limitedcomparativedata existforiniximab.Theobjectiveofthismodelinganalysiswastoassessthe impactondruglevelsfromgapsinprescriptionrellbehaviorthatmayresultin under-dosing. METHODS: Steadystateconcentrationmodelsforetanerceptand iniximabweredevelopedtosimulatetheeffectofmisseddosing(i.e.,gapin etanerceptprescriptionrelloriniximabadministrationinterval).Resultswere expressedasapercentofsteadystatetroughlevels(SSTL)ing/mL,andwere presentedtosimulategapsthatvariedfrom1-5weeks.Populationpharmacokineticmodelsforetanerceptandiniximabwereusedtosimulateoutsteadystate concentrationproles.Theparametersusedinthesemodelswereextractedfrom recentpublications. RESULTS: Afterarelloradministrationgapofoneweek, etanerceptandiniximabconcentrationswereat13%and73%ofSSTLrespectively.Attwoweeks,SSTLwere4%and55%respectively.Etanerceptlevelswere effectivelynon-existentafterthreeweeks,whereasiniximabwasat40%ofSSTL. Clinicalimplicationswerenotsimulated. CONCLUSIONS: SSTLofetanerceptdeclinesrapidlywithgapsinrellsanddisappearswithgapsasshortasthreeweeks. Iniximabdemonstratesamuchmoregradualdecline.Poorpatientrellbehavior mayhaveconsequencesthatgobeyonddecliningdruglevels,andpatientadherenceriskshouldbeanintegralpartofanydiscussionduringsharedtreatment decisionmaking. PSY70 ASSOCIATIONOFPROVIDERCONTINUITYWITHHOSPITALIZATIONAMONG FLORIDAMEDICAIDENROLLEESWITHSICKLECELLDISEASE(SCD)MaQ,KaufT,HallAG UniversityofFlorida,Gainesville,FL,USAA71VALUEINHEALTH14(2011)A1–A214

PAGE 80

OBJECTIVES: Providercontinuityisdenedasseeingthesamehealthcareprovider overtime.Previousstudiesindicatedthathighprovidercontinuityimproves healthcareoutcomesandtheefciencyofhealthcaredelivery.Theimpactof providercontinuityinsicklecelldisease(SCD)care,however,isunknown.This studyexaminedtheassociationbetweenprovidercontinuityandtheriskofSCDrelatedhospitalizationandre-hospitalizationwithin30daysofdischarge. METHODS: AretrospectivecohortstudywasconductedusingFloridaMedicaid claimsdatafrom2001-2005.PatientswithclaimscontainingSCD-relatedICD-9 codes(282.41-282.42,282.6-282.64,282.68-282.69)wereselected.Tobeeligible, individualshadtobeaged 65years,continuouslyenrolledinMedicaid,andhave made 2ambulatoryvisits.Modiedmodiedcontinuityindex(MMCI)scoreswere calculatedtoquantifyprovidercontinuity.Coxproportionalhazardmodelingwas usedtoexaminetherelationshipbetweenMMCIandhospitalizationand30-day re-hospitalizationcontrollingforbasicdemographics,priorutilization,SCDtreatmentsandcomplications. RESULTS: Atotalof2422patientswithmeanageof10.2 (SD 11.9)and47.2%malewereincluded.AverageMMCIscorewas0.60(SD 0.28), 53.47%had 1SCD-relatedhospitalization;ofthose,18.8%werere-hospitalized within30days.Aftercontrollingforpatient-levelfactors,patientswithhigher providercontinuitywerelesslikelytobehospitalized(HR 0.53,95%CI:0.44-0.65), butMMCIwasnotsignicantlyassociatedwith30-dayriskofre-hospitalization (HR 0.89,95%CI 0.55-1.43). CONCLUSIONS: HigherprovidercontinuitywasassociatedwithalowerriskofSCD-relatedhospitalizationintheFloridaMedicaidSCD population,butdidnotaffecttherateof30-dayre-hospitalization.Policiesthat improveandmaintaincontinuitybetweenSCDpatientsandasingleprovidermay leadtolowerhospitalizationratesandpossiblylowerhealthcarecosts. PSY71 HEALTHANDFUNCTIONALSTATUSFORINDIVIDUALSWITH THROMBOCYTOPENIAINTHEUNITEDSTATES:EVIDENCEFROMNHANESKaufT1,SchelfhoutJ1,NelsonDR1,WangPF2 1UniversityofFlorida,Gainesville,FL,USA,2GlaxoSmithKline,Philadelpha,PA,USAOBJECTIVES: Thrombocytopenia,orlowbloodplateletcount,isassociatedwith diseaseprogressionandtreatmentforseveralchronicconditions,includingliver disease,cancer,andHIV.Yet,littleinformationdescribingtheimpactofthrombocytopeniaonpatientsisavailable.Thisstudycomparedindicatorsofhealthand functionalstatusamongindividualswithandwithoutthrombocytopenia. METHODS: 1999-2008NationalHealthandNutritionExaminationSurvey (NHANES)datawereusedtoidentifyindividualsaged 20yearswithnormaland lowplatelets,usingathresholdof150x109/L.Indicatorsofhealthand/orfunctional statusincludedself-reportedgeneralhealth[1-5scale,1 excellent],numberof healthcarevisitsandovernighthospitalstays,daysphysicalhealthwasnotgood, inactivedaysduetophysicalhealth,andworklimitations.Weightedmeansand frequenciesweresummarizedandcomparedusingpairedt-testsandchi-squared tests,respectively. RESULTS: 22,959/25,772participantsaged 20yearshadinformationonplatelets.Ofthese,526(1.9%)hadlowplateletsatthetimeofexamination.Individualswiththrombocytopeniawereolder(51.7vs37.2years)andmore likelytobemale(67.7%vs48.6%),infectedwithhepatitisC(7.4%vs1.4%),andever diagnosedwithcancer(14.2%vs8.0%)oralivercondition(12.3%vs3.1%),p .05for all.Fewerparticipantswiththrombocytopeniareportedexcellenthealthcompared tothosewithnormalplatelets(12.5%vs16.5%,p .0001).Participantswiththrombocytopeniareportedmorehealthcareprovidervisits(2.35vs2.02,p .0001)and weremorelikelytoreportanovernighthospitalstay(0.19%vs0.11%,p .001). Participantswiththrombocytopeniahadmoreinactivedaysduetophysicalhealth andweremorelikelytofaceworklimitations(p .01forall). CONCLUSIONS: Thrombocytopeniaisassociatedwithlimitationsonfunctionalstatusandincreasedhealthcareutilization,someofwhichmayberelatedtopatients'underlyinghealthconditions.Furtherstudyisneededtodeterminetheincremental impactofthrombocytopenia. PSY72 VARIATIONBYAGEINHEALTH-RELATEDQUALITYOFLIFEOFPATIENTS INITIATINGTREATMENTFORMYELODYSPLASTICSYNDROMESPashosC1,GrinblattDL2,KomrokjiRS3,SekeresMA4,NarangM5,SullivanKA6, StreetTK7 ,KhanZM7 1UnitedBioSourceCorporation,Lexington,MA,USA,2NorthShoreUniversityHealthSystem, Evanston,IL,USA,3MofttCancerCenter,Tampa,FL,USA,4ClevelandClinic,Cleveland,OH, USA,5AllianceHematologyOncology,Westminster,MD,USA,6CelgeneCorporation,Overland Park,KS,USA,7CelgeneCorporation,Summit,NJ,USAOBJECTIVES: Minimalhealth-relatedqualityoflife(HRQOL)datahavebeenpublishedonmyelodysplasticsyndromes(MDS)patientsintheUnitedStates.This analysischaracterizedHRQOLbyagegroupofMDSptstreatedwithVidaza¨inthe AVIDA¨registry. METHODS: DatawerecollectedfromAVIDA,aVidazatreatment registry,initiatedinUScommunityclinics.Cliniciansprovidedpatientdemographic,treatmentpattern,andclinicaloutcomedata.PatientsreportedHRQOL (EORTC-QLQ-C30)within2monthsoftreatmentinitiation.Mean(StandardDeviation[SD])scoresonglobalhealthstatus(GHS),5functionalscalesand9symptom/ otherscaleswereanalyzedbyage: 65,65-74, 75years.Statisticalsignicance wasascertainedbyANOVAusingSAS9.1. RESULTS: Datawerereportedfrom99 clinicson427pts: 65(n 68),65-74(n 127),and 75(n 232)years.Mean(SD)GHS was54.8(1.2)overall,andsimilaramongagegroups:53.9(2.9),54.7(2.1),and55.2 (1.6),p 0.9263.Similarly,physical,role,andsocialfunctioningwerecomparable. Emotionalandcognitivefunctioning,however,wereworseamongthose 65versustheoldergroups:66.2(3.5),78.6(1.8),79.5(1.5),p 0.0001foremotional,and72.4 (3.5),82.7(1.7),80.3(1.4),p 0.0067forcognitive.The 65cohortreportedsignicantlyworsescoreson4of9symptom/otherscores:nancialdifculties(34.8(4.5), 15.3(2.4),13.0(1.7),p 0.0001);insomnia(39.7(4.0),27.5(2.7),27.3(2.2),p 0.0141); appetiteloss(29.9(3.8),18.4(2.5),25.3(2.1),p 0263);andfatigue53.9(3.1),44.3(2.3), 48.5(1.8),p 0.0452). CONCLUSIONS: AVIDAregistryndingsindicatebaseline HRQOLamongMDSpatientsinrealworldsettingsdiffersbyageincertaindomains.FindingssuggestHRQOLissimilarorbetterinelderlyMDSpatientsthan youngerpatients.Thesedifferencesinemotionalandcognitivefunctioning,fatigue,insomnia,appetiteloss,andnancialdifcultiesrequireconsiderationby cliniciansinmanagingMDSpatients. PSY73 PAINMEDICATIONUSEANDDETERMINANTSOFOPIOIDSPRESCRIBINGINTHE UNITEDSTATESOUTPATIENTSETTINGSRasuR,CunninghamL,SohrabyR,KnellM UniversityofMissouri-KansasCity,KansasCity,MO,USAOBJECTIVES: ChronicpainisamajorpublichealthconcernintheUS.Established guidelinesareavailableformanagementofnon-malignantchronicpain,including opioiduse.However,discrepanciesinopioidprescribingpatternsduetophysician misconceptionsremainconcerning.Therefore,thisstudyevaluatedpainmedicationuseandinvestigateddeterminantsofopioidanalgesicprescribingintheUS outpatientsettingsforcommonnon-malignantchronicpainindications. METHODS: Thiscross-sectionalstudyanalyzedtheNationalAmbulatoryMedical CareSurvey(NAMCS)datafrom2002-2007onpatients18yearsandolderwith non-malignantchronicpaindiagnosisbasedonICD-9-CMcodesidentiedasreasonforvisits.PainmedicationsprescribedwereretrievedusingNAMCSdrugcodes. Multivariatelogisticmodelsexamineddeterminantsofopioidprescribingamong chronicpainpatients. RESULTS: Approximately69millionweightedoutpatient visitswerereportedfornon-malignantchronicpainbetween2000-2007intheUS. Themeanageforpatientswas53(range18-100)andthemajoritywere women(63%).Neuropathicpainwasreportedfor2.39%visitswhile16.24%hadan inammatorypaindiagnosis.Non-medicationtreatmentwasprescribedduring 26%ofthesevisits.Whilemost(95%)visitsreportedprescribingNSAIDs,29%reportedreceivingprescriptionsformorethanvemedications.Primarycarephysicians(PCPs)were1.74timesmorelikelytoprescribeopioids[OR(oddsratio):1.74, CI:1.42-2.14]thanotherspecialtyphysicians.Patientsreceivingmorethanve medicationswere2.80timesmorelikelytoreceiveopioids(OR:2.80,CI:2.28-3.44) thanthosewithlessthanvemedications.Patientsfromthesouthernregionofthe USwere1.43timesmorelikelytoreceiveopioids(OR:1.43,CI:1.06-1.94)thanpatientsfromthenortheastregion. CONCLUSIONS: Accordingtoourstudy,visits withPCPs,morethanveprescriptions,establishedpatients,andphysicianvisits inthesouthernregionoftheUSweresomeofthedeterminantsofopioidprescribing.Increasedawarenessofopioidprescriptionguidelinesforpainmanagement mayeliminateprescriptiondiscrepanciesandimprovepatientcare. PSY74 TREATMENTCHOICEFORPAINMANAGEMENTINNURSINGHOMEHOSPICE/ PALLIATIVECARERESIDENTSINTHEUNITEDSTATESAgrawalR,MhatreSK,SansgiryS UniversityofHouston,Houston,TX,USAOBJECTIVES: Thisstudydeterminedfactorsassociatedwithtreatmentchoicefor painmanagementinUSnursinghomehospice/palliativecareresidents. METHODS: Thisisacrosssectionalstudyofdatafromtheresidentleofthe2004 NationalNursingHomeSurvey(NNHS).Residentsassignedtoabedinahospice specialtyunitorreceivingservicesfromaspecialprogramforhospice/palliative carein1174systematicallyselectednursinghomesandhavingpainin7days beforethesurveywereincluded.Treatmentchoicesincludedstandingorder(SO)of analgesics,as-neededorders(PRN)ofanalgesicsandnon-pharmacologicaltreatment(NPT).Factorsevaluatedweregender,maritalstatus,age,ethnicity,eating limitation,toiletlimitation,urinaryincontinenceandbowelincontinence. Weighteddescriptiveanalysis,bivariateanalysisusingchi-squaretestsandmultivariateanalysisusinglogisticregressionwereconductedusingSASversion9.2to determinefactorsassociatedwithpainmanagement. RESULTS: Overall22.6% (338,029/1,492,207,weighted)patientsmetthestudyinclusioncriteria.Mean[SD] agewas79.1[13.3]years,74.7%werefemalesand89.8%werewhite.78.6%,49.8%, and29.3%receivedPRN,SO,andNPT,respectively.Logisticregressionshowed higherlikelinesstoreceiveSOforpatientsofageatorabove65years(OR 1.5195% CI 1.184-1.948P 0.001),patientswithbowelincontinence(OR 1.2195% CI 1.020-1.437P 0.028)andMedicaidpatients(OR 1.595%CI 1.258-1.799 P 0.0001).LikelinesstoreceivePRNwaslowerforpatientsofageatorabove65 years(OR 0.6095%CI 0.428-0.841P 0.0031),patientswithbowelincontinence (OR 0.6795%CI 0.533-0.859P 0.0014),Medicaidpatients(OR 0.6595%CI 0.5250.806P 0.0001)andHispanics(OR 0.32795%CI 0.204-0.527P 0.001).Likeliness toreceiveNPTwashigherforpatientswithtoiletinglimitation(OR 2.90595% CI 1.135-7.434P 0.026)andlowerforpatientswithurinaryincontinence(OR 0.02295%CI 0.003-0.172P 0.0003). CONCLUSIONS: Factorsaffecting treatmentchoicewereage,bowelandurinaryincontinence,toiletinglimitations, ethnicityandinsurancetypeandcanbetargetedinfuturestudiesaimingatmanagementofpaininnursinghomes. SystemicDisorders/Conditions–ResearchonMethods PSY75 PREVALENCERATIOSAGAINSTODDSRATIOSASEFFECTMEASURESINA CROSS-SECTIONALSTUDYOFOBESITYANDITSCHRONICCOMORBID CONDITIONSPatelJG,RajanSS,AparasuRR,JohnsonM UniversityofHouston,Houston,TX,USAA72VALUEINHEALTH14(2011)A1–A214

PAGE 81

OBJECTIVES: TocomparetheestimatesofPrevalenceRatio(PR)andOddsRatios (OR)aseffectmeasuresintheanalysisofcrosssectionaldataforobeseindividuals sufferingwithchroniccomorbidconditions. METHODS: MedicalExpenditurePanel Surveydatalesfrom2005-2007wasutilizedfortheanalysis.Obeseadultswere denedasindividualswithage 20yearsandhavingaBMI 30.Prevalenceratios wereestimatedfromlogisticregressionsbydividingpredictedprevalenceestimatesofcomorbiditiesamongobeseindividualstothepredictedprevalenceof comorbiditiesinnon-obeseindividualsafteradjustingforsocio-demographicfactors.Ninety-vepercentboot-strappedcondenceintervalsweregenerated aroundtheprevalenceratios.OddsratiosweregeneratedbyusingtheSurvey Logistic'syntaxinSASafteradjustingforthesamesocio-demographicfactors. RESULTS: Bothoddsratiosandprevalenceratiosindicatethatobeseindividuals haveagreaterburdenofchronicconditionsintheUnitedStates.Obeseindividuals hadthegreatestoddsofhavinghypertensionandosteoarthritisfollowedbycoronaryheartdisease,diabetesanddyslipidemia.Prevalenceratioswerehighestfor diabetes,followedbyhypertension,osteoarthritisanddyslipidemia.Oddsratios werealmostalwaysgreaterthantheprevalenceratiosforallchronicconditions. Thecondenceintervalsderivedaroundoddsratioswerealsowiderthanthe condenceintervalsaroundtheprevalenceratios. CONCLUSIONS: Therelative importanceofobesityupondifferentdiseaseconditionsdiffereddependingon whetherPRorORwereestimated.Thisstudyalongwithpreviousliteratureshows thatoddsratiosmaybeoverestimatingthetrueeffectofadiseaseconditiononthe prevalenceofacomorbiddisease.Althoughitiseasiertoderiveoddsratios,prevalenceratiosmaybemorerealisticestimatesofthetruepublichealthburden.Thus prevalenceratiosshouldideallybeusedaseffectmeasuresinsteadofoddsratios especiallyforcommonlyoccurringdiseaseconditions. PSY76 TREATMENTOFCRYOPYRIN-ASSOCIATEDPERIODICSYNDROMES(CAPS) KNOWLEDGESTUDY(TOCKS):NOVELDATACOLLECTION,VIEWINGAND DYNAMICREPORTINGMECHANISMArnoldRJ1,GroesserK2 1MountSinaiSchoolofMedicine,NewYork,NY,USA,2ArnoldConsultancy&TechnologyLLC, NewYork,NY,USAOBJECTIVES: Theobjectiveofknowledgestudy(TOCKS)wastoquantifythepatient burdenandtocharacterizepatientsymptomatologyandacceptanceoftreatment withanakinrainCryopyrin-associatedPeriodicSyndrome(CAPS),anorphandisease.Additionally,chartreviewandpatientrecallofsymptoms,adverseevents, andresourceusewerecompared.Theaforementionedobjectiveswerecompleted throughtheuseofauniqueonlineapproachtogatheringdataandrapidlydisplayingtheresults. METHODS: Thisretrospectivemedicalchartreviewandconcurrent onlinepatientsurvey,conductedinCentersofExcellenceforCAPSinEurope,was accomplishedbyanalyzingdataenteredviainternet-basedcasereportforms (CRFs).DatawereenteredintotheCRFsbybothstudygroupsfortwodistinct collectionperiodspriortoandduringthemostrecent12monthsofanakinra treatment. RESULTS: Foursites(50patientstotal)participatedinTOCKS.Priorto enteringdata,usersviewedanonlinetutorialtohelpincompletingthesurvey.No patientswerelosttofollowupandthedataweresuccessfullycollected,analyzed, andreviewedviatheonlineCRFanddynamically-generateddatatables.Thisinnovativeonlineapproachalloweduserstoenterde-identieddata,inmultiple sessions,eveninthemostremotedestinations.OnepatientsuccessfullycompletedtheirsurveyintheAmazon.Inaddition,auniqueidentierwasrandomly generatedthatlinkedchartreviewwithpatientrecalldataonline.Siteinvestigatorswerealsoabletoviewthedynamictablesandreportsonlineandbenchmark themselvesagainsttheother,encrypted,sites. CONCLUSIONS: Thisnovelinternet technologyallowedforefcientdatacollectionfrommultiplesitesandmultiple sourceswithinashorttimeframe.Theuniqueidentiermadeiteffortlesstocollectandcomparetheentriesfromthechartandfromthepatientsforthesame patient. PSY77 COMPARISONOFOBESITY-ASSOCIATEDCOMORBIDITIESBETWEENEMRAND CLAIMSDATABASESBrixnerD1,BronM2,BellowsBK1,YeX1,HarikrishnanV2,OderdaG1 1UniversityofUtah,SaltLakeCity,UT,USA,2TakedaPharmaceuticalsInternational,Inc., Deereld,IL,USAOBJECTIVES: Tocompareratesofobesity-associatedcomorbiditiesbetweenthe GeneralElectric(GE)CentricityEMRandThomsonReutersMedStatMarketScan commercialclaimsdatabasesinpatientswitheitherarecordedBMI 25(GEEMR) orICD-9codeforoverweight/obesity(MarketScan). METHODS: FromtheGEEMR, subjectsaged20-64withatleastoneBMIvalue 25and 2yofEMRactivityprior toBMIindexdatewereincluded.FromMarketScan,subjects 20y(mostly 65y), withaclaim(ICD-9code)foroverweight,obesity,ormorbidobesity,and 365days ofclaimsactivityafterselectedcomorbiditywasrstrecorded,wereincluded. SelectedcomorbiditieswereidentiedbyICD-9codesinbothdatabasesandstratiedbyBMIoroverweight/obesityclaims. RESULTS: IntheGEEMR,109,885subjects(36.7%)wereoverweightorobese(BMI 25).Ofthese,51%hadBMI25-29.9, 28%hadBMI30-34.9,and22%hadBMI 35.IntheMarketScandatabase,246,261 subjects(0.7%)hadoverweight,obesity,ormorbidobesityclaims.Ofthese,4%were overweight,66%wereobese,and30%weremorbidlyobese.Thethreemostprevalentcomorbiditiesinbothdatabases,acrossallweightcategories,werehypertension,hyperlipidemia,andchronicbackpain,exceptdiabetesreplacedbackpainin themorbidobesitygroupinMarketScan.Forallcomorbidities,prevalencewas higherinMarketScan,exceptpolycysticovarysyndromeprevalencewashigherin GEEMRBMI 35groupcomparedtomorbidobesitygroupinMarketScan. CONCLUSIONS: Despitedifferentmethods(BMIvs.ICD-9codes)forcapturingobesitybetweentwodatabases,prevalencerankingsofcomorbiditiesweresimilar. ObesityistherapeuticallyclassiedbyBMI;however,mostlargeclaimsdatabases captureobesitybyICD-9codes.Thismayresultinvariationinweightgroupdistributionsandsignicantunder-reportingofobesityclaimsinMarketScan.Our ndingshighlighttheneedforfurtherresearchinoptimalsourcesfordatainobese patients. PSY78 ESTIMATINGTHEBMI-MORTALITYRELATIONUSINGFRACTIONAL POLYNOMIALSWongE1,WangBCM2,GarrisonL2,Alfonso-CristanchoR2,FlumD2,ArterburnD3, SullivanSD2 1DepartmentofVeteransAffairs,Seattle,WA,USA,2UniversityofWashington,Seattle,WA, USA,3GroupHealthResearchInstitute,Seattle,WA,USAOBJECTIVES: Thisstudytestsaexiblemodelingapproach,whichendogenously estimatesthenon-linearandasymmetricfunctionalformforbodymassindex (BMI),toexaminetherelationshipbetweenmortalityandobesitymeasuredasBMI 30. METHODS: ThisstudyusedtheNationalHealthInterviewSurvey(NHIS), between1997and2000.RespondentswerelinkedtotheNationalDeathIndexwith mortalityfollow-upthrough2005.Weestimatedthe5-yearprobabilityofdeath usingthelogisticregressionmodeladjustingforBMI,ageandsex.Themultivariablefractionalpolynomials(MFP)procedurewasemployedtodeterminethebest ttingfunctionalformforBMIandcomparedtoalternativefunctionalformsusing achi-squaredtest.Expectedyearsoflifelostduetoobesitywerebasedonadjusted deathprobabilitesandcomputedusingstandardlifetablefunctions. RESULTS: The bestttingadjustmentmodelcontainsthepowers 1and 2forBMI.Achisquaredtestshowsastatisticallysignicantimprovementinmodeltcompared tootherBMIpolynomialfunctions.Theestimatedrelationshipbetween5-year probabilityofdeathandBMIexhibitsaJ-shapedpatternforwomenandaU-shaped patternformen.TheBMIassociatedwithminimummortalityis27.53formales and25.19forfemales.A40-year-oldfemalewithaBMIof40hasanestimated5.82 feweryearsofexpectedlifecomparedtoananalogousfemalewithaBMIof25.For acomparablechangeinBMIina40-year-oldmale,theexpectedyearsoflifelostis 5.20. CONCLUSIONS: TheBMI-mortalityrelationisataroundtheminimum,but especiallyhighmortalityisassociatedwiththemorbidlyobese.TheMFPapproach providesarobustalternativetoestimatingmortalitybyallowingthedatatodeterminethebestttingmodel.TheapproachisalsousefulinestimatingtherelationshipbetweenthefullspectrumofBMIvaluesandotherhealthoutcomes. PSY79 CONTENTVALIDITYOFTHEMULTIPLESCLEROSISINTERNATIONALQOL (MUSIQOL)QUESTIONNAIREINIRAN,EGYPT,MOROCCO,SAUDIARABIAAND TUNISIATrevinA1,SalmassiL1,AuquierP2 1MAPIInstitute,Lyon,France,2TimoneUniversityHospital,Marseille,FranceOBJECTIVES: TheMusiQOLquestionnairewasco-developedin15countriestoassessthequalityoflifeofpatientswithMultipleSclerosis.Theobjectiveofthisstudy wastotesttheculturalrelevanceoftheinstrumentin4Arabiccountries(Egypt, Morocco,SaudiArabiaandTunisia)andIran,notinvolvedintheinitialdevelopment,andtoassesstheconceptualequivalenceofthetranslationswiththeUK originalusedasabasisfortranslation. METHODS: Ineachcountry,thetranslation processwasconductedbyalinguisticexpert,usingeitherthestandardforward/ backwardmethodologyortheadjustedprocess(adaptationfromtheSaudiArabia version),includingcognitiveinterviewswith6patients.Thebasisfordiscussion wastheconceptlistdevelopedincollaborationwiththeauthor. RESULTS: Linguisticandculturalissuesemergedduringthetranslationprocess.First,theacronym "MS"usedthroughouttheoriginalversionfor"MultipleSclerosis"wasreplacedby thefullnameinArabiccountriesforclarity,butnotinIranwherethepatients preferredtheabbreviation.Second,usingeuphemisticexpressionsfortabooconceptssuchassexlifeprovednecessarytoensurehomogenousresponseacrossall languageswithoutanynegativeconnotation.Finally,leisureactivities(e.g.shopping,goingouttoamovie,gardening)describedintheoriginalhadtobeadaptedto thereligiousandsocialcontextinthetargetcountries. CONCLUSIONS: The5languageversionsoftheMusiQolwereestablishedfollowingaprovenstandardized methodology,onthebasisofaconceptlistworkedoutwiththeauthor,toallow internationaldatapoolingandminingwhilstaddressingthespecicchallengesof regionaladaptation.Theprocessasawholesupportstheadvantageofintegrating continuousanddiverseinternationalfeedbackonwordingduringthelinguistic validationprocess. PSY80 PRIORITIZATIONANDWEIGHTINGOFPATIENT-RELEVANTENDPOINTS(PRES) ASPARTOFTHEIQWIGSEFFICIENCYFRONTIERMETHODINGERMANYMŸhlbacherA1,BridgesJF2,BethgeS1,DintsiosCM3,SchwalmA3,NŸblingM4 1DukeUniversity,Durham,NC,USA,2JohnsHopkinsUniversity,Baltimore,MD,USA,3Institute forQualityandEfciencyinHealthCare(IQWiG),Cologne,Germany,4GesellschaftfŸr empirischeBeratungmbH,Denzlingen,GermanyOBJECTIVES: TheIQWiG'smethodprovidesaconceptforthedrawingofanefciencylimitanditsextrapolationtotheassessmentofthecost-benetrelation. Herebenetsaretobeassessedintermsofasetofclinicalendpoints.PREsmight includedifferentissuesofmortality,morbidityandhealth-relatedqualityoflifeas wellasotherimportantfactorsidentiedasbeingpatient-relevant.Efciencyfrontierscannotbedrawnforallpossibleoutcomes,itisimportanttodeterminewhich factorsaremostrelevanttopatients.IQWiGallowsdifferentefciencyfrontiers. Howcancontradictoryresultsbeinterpreted?Theproblemwithintheutilitymea-A73VALUEINHEALTH14(2011)A1–A214

PAGE 82

surementwithmorethanoneendpointisthegenerationofacardinalindex/score. Withouttheprioritizationandweightingofmultipleendpointsadeductionof recommendationsmightbequestionable. METHODS: Apilotstudywasconducted toelicitpatients'preferencesaboutantiviraltherapyofchronichepatitisC.Forthe Discrete-Choice-Experiment(DCE),7attributeswereselectedwith3Levelseach. Thereforeanorthogonal,balancedandefcientdesignwasusedandresultswere analysedwithrandomeffectslogitmodels. RESULTS: Patientsandexpertsprioritizedtherespectiveendpointsinalmostthesameorder,butweightedthemdifferently.Sustained-Virological-Responsereceivedthehighestweightfollowedbyfrequencyofapplication(patients)ordurationoftherapy(experts). CONCLUSIONS: Aim wastodemonstratehowDCEscanbeusedtoempiricallydeterminewhichPREs shouldbeincludedintheefciencyfrontieranalysis.Furtheritisdemonstrated howsuchmethodscanbeusedtoprioritizeacrosssuchmultipleefciencyfrontiers.ThesurveydemonstratedhowDCEscouldbeusedtoempiricallydetermine whichPREsareimportantinantiviraltreatmentofchronichepatitisC.Theresults couldbeusedforthedevelopmentofinnovativetherapeuticschemesandnew drugswhichcouldmeetpatients'needs.ForIQWiGpurposestheweightsofPREs areincludedinthehealtheconomicevaluation. PSY81 PERCEPTIONSABOUT’RAPID’WARFARINREVERSALWITHFRESHFROZEN PLASMATOREDUCETHEINR:ANOVELSURVEYMETHODOLOGYPetrozzinoJ1,HoescheJ2,JonesC2 1ComparaBiomedical,Orlando,FL,USA,2CSLBehring,KingofPrussia,PA,USAOBJECTIVES: Todeterminehealthcareproviderperceptionsabouttimingof'rapid' warfarinreversal-relatedpatientcareeventsusingnovelsurveymethodology. METHODS: Forty-eightadultandpediatrictraumacenterswerecontactedtoparticipateinadirect-to-provider(DTP)survey.Participantswereaskedtoprovide aggregateinformationaboutpatientsreceivingfresh-frozenplasma(FFP)foracute warfarinreversal. RESULTS: Nineteento25healthcareprofessionalsfrom18centersprovidedinformationbysurvey.Averageperceptionsoftimeneededtoinfuse FFPunderthissetting(mean4.6hrsfromtimeoftriage;95%CI1.08.2hrs)are consistentwithactual,publishedvalues.Incontrast,averageperceptionsoftime neededforinitialInternationalNormalizedRatio(INR)normalizationusingFFP (mean5.8hrs;95%CI2.88.8hrs)underestimateactual,publishedvaluesb y626 hrs.Healthcareprovidersperceivedthatrelativelylittlecumulativetimelapses (1.6hrs,ontheaverage)forcompletingtherstFFPinfusion.ThereislittleperceivedtimelagbetweenorderingandbeginningtherstFFPinfusion(0.3hrs,on theaverage),consistentwithactual,publishedvalues.Thereissubstantialreportedtime(anadditional3.0hrs,ontheaverage)neededtocompletesubsequent FFPinfusions,amountingto52%ofallperceivedtimelapsingforinitialINRnormalizationinthissetting. CONCLUSIONS: DTPsurveymethodologyappearstobe anefcientmethodforgatheringclinicalinformationforresearchpurposes. Healthcareprovidersmayhaveperceptionsthataredifferentfrompublishedstudies,includinginaccurateperceptionsofthedelaybetweenthetimeFFPinfusions arecompletedandtimeofinitialINRnormalization.Toourknowledge,thisisthe rststudytoshowthatsubsequent-to-rstFFPdoseinfusiontimesaccountforthe majorityofperceivedINRnormalizationtime.Delaystotreatmentcompletionmay presentseriousdownstreamconsequences.Suchperceptionsmayinuenceclinicaldecisionmaking,andwarrantfurtheranalysisandinvestigation. Urinary/KidneyDisorders–ClinicalOutcomesStudies PUK1 INCIDENCEAND30-DAYMORTALITYOFCOMMUNITYACQUIREDPNEUMONIA (CAP)INTHEMEDICAREFEE-FOR-SERVICE(FFS)POPULATIONRubinJL1,LiS2,YuH3,DunningS2,SatoR3 1DaVitaClinicalResearch,Minneapolis,MN,USA,2ChronicDiseaseResearchGroup, Minneapolis,MN,USA,3Pzer,Collegeville,PA,USAOBJECTIVES: ToestimateCAPepisodeincidenceand30-daymortalityratesinthe MedicareFFSpopulation 65yearsofage. METHODS: ClaimsdatafromtheMedicare5%samplewereusedtoidentifypneumoniadiagnosesbetweenJuly1,2007to June30,2008.WedenedpneumoniahospitalizationsaseitherPartA(PA)primary dischargediagnosis(pdd)ofpneumonia,orsepsisorrespiratoryfailureaspddplus pneumoniaassecondarydiagnosis.OutpatientpneumoniawasdenedasapneumoniadiagnosisfromPAoutpatientornon-hospitalsourcedPartB,withclaimfor chestx-raywithin14daysofpneumoniadiagnosis.CAPepisodesweredenedas beingindexedontherstpneumoniaclaimdatefollowedthroughthelastpneumoniaclaimdateprecededandfollowedbya90daycleanperiod,respectively. Episodesweredenedasoutpatient,unlessbeginningwithorcontainingaCAP hospitalization,inwhichcasetheywereconsideredinpatientCAPepisodes.We furtherexcludedevidenceofhospitalization,long-termcareandmechanicalventilatorusefromclaimswithin14daysoftheCAPindexdate.Incidenceandmortalitywerestratiedbyinpatientstatusandage. RESULTS: Atotalof56,262CAP episodeswereidentied;38.4%wereinpatient.Theaverageinpatientandoutpatientepisodelengthwas32.7days(SD46.3)and12.8days(SD27.6),respectively. Correspondingincidencerateswere1,767/100,000person-yearsforinpatientand 2,837/100,000foroutpatientCAP.Overallincidenceriseswithincreasingage,rangingfrom2,786/100,000person-yearsforthose65-69to8,449/100,000forthose85 Overall30-daycase-fatalitywas5.6%;8.5%forinpatientand3.8%foroutpatient CAP.Case-fatalityratesalsoincreasewithincreasingage. CONCLUSIONS: The overallincidenceofCAPintheMedicareFFSpopulationof4,604/100,000personyearsissubstantial.Theseresultssuggestanestimated1.13millioncasesofCAP and63,000CAP-relateddeathsannuallyamongthispopulation.CAPremainsan importantpublichealthburdenintheUnitedStates. PUK2 DUTASTERIDEFORBENIGNPROSTATICHYPERPLASIA:ASYSTEMATICREVIEW ANDMETA-ANALYSISParkT,ShamliyanTA UniversityofMinnesota,Minneapolis,MN,USAOBJECTIVES: Thisstudysystematicallyreviewsexistingevidencetodeterminethe efcacyandadverseeventsofdutasteride,aloneorincombination,versusaplaceboorcontrol,usedforthetreatmentofbenignprostatichyperplasia(BPH). METHODS: MedlineandtheCochraneLibraryweresearchedforrandomizedcontrolledtrialsthatwerelongerthansixmonthsinduration.Thestudypopulation wasmen,aged50orolder,withmoderateandseveresymptomsofBPHasdeterminedbytheInternationalProstateSymptomScore(IPSS)andclinicalparameters includingmaximumurinaryowrate(Qmax)andprostatevolume.Ameta-analysis witharandomeffectsmodelwasconductedtosynthesizethendingsofmultiple studies.HeterogeneityamongthestudieswasassessedbothgraphicallywithforestplotsandstatisticallywithI2statistics. RESULTS: Dutasteridewassuperiorto theplaceboinreducingurinarysymptomsmeasuredbyIPSS[weightedmeandifference(WMD) 2.50,95%condenceintervals(CIs) 2.91to 2.09],prostate volume (WMD 12.90,95%CIs 14.27to 11.53)andinincreasingQmax(WMD 1.30,95%CIs 0.99to1.61).However,riskratiosforanyadverseevents(1.27,95% CIs 0.99to1.62)andfordrug-relatedadverseevents(1.47,95%CIs 0.77to2.79) werenotsignicantlydifferentbetweenthosetreatedwithdutasterideandthose inthecontrolgroup.Whendutasteridewasadministeredwithtamsulosin,the reductioninIPSSscoresandtheincreaseinQmaxweregreaterthanifthemedica tionswereadministeredalone,althoughprostatevolumechangedlittle.Forboth BPH-relatedacuteurinaryretention(AUR)andBPH-relatedsurgery,nosignicant differenceinriskratiowasobservedbetweenthecombinationgroupandthedutasteridegroup. CONCLUSIONS: Dutasteridecanbeusedforthosewithmoderate tosevereBPHsymptomswhenbenetsfromefcacyareconsideredtobegreater thanthelossfromadverseevents. PUK3 DRUG-INDUCEDACUTERENALFAILUREUSINGTHEFDAADVERSEEVENT REPORTINGSYSTEMDATABASEAlhammadAM,AlHawajMA,AlsalmanAJ,AlhashemYN,HarpeSE,SmithWR, BrophyDF VirginiaCommonwealthUniversity,Richmond,VA,USAOBJECTIVES: Theincidenceofacuterenalfailure(ARF)isincreasing.Asubstantial proportionofthesecasesareduetodrug-inducednephrotoxicity.Datafrompostmarketingsurveillanceprograms,liketheFDAAdverseEventReportingSystem (AERS),canbeusefulinassessingtheriskofadversedrugevents.Ourobjectiveisto describereportswithARF(cases)andreportswithoutARF(non-cases)andoutcomesresultingfromARFusingtheAERSdatabase. METHODS: DatawereextractedfromtheAERSdatabasefrom20042009.Allreportsofcasesandnon-cases weredescribedandanalyzedintermofsuspecteddrugs,healthoutcomes,and othercharacteristics.Descriptivestatistics(frequenciesandproportions)andChisquaretestswereused. RESULTS: TheAERSdatausedinthisstudyrepresents 2,231,689reportsfrom184countrieswith27,190involvingreportsofARF(cases). Therewasapproximatelyatwo-foldincreaseinthepercentageofreportsofARF from2004to2009(n 3,594;13%ton 6,104;22%,respectively).Mostreportswere submittedbymanufacturers(cases 89%,non-cases 93%).Themajorityofreportswereforinfants(cases 75%,non-cases 57%).Incases,drugswerethe primarysuspect,thesecondarysuspectandconcomitantin18%,14%and66%of reports,respectively.ThevedrugswiththehighestreportedfrequenciesofARF wererofecoxib,valacyclovir,metformin,simvastatin,anddigoxin.Unfavorable outcomesweremorelikelytooccurincasesthannon-cases(death24%vs.12%,life threateningcondition21%vs.5%,initial/prolongedhospitalization76%vs.33%, disability6%vs.4%andrequiredintervention4%vs.2%; p -value 0.001,forall comparisons). CONCLUSIONS: ThesepreliminaryndingspresentanoverallpictureofreportswithARF.Thesendingscanbeinformativeforregulatoryauthoritiesandhealthcareprofessionals.Additionalanalyseswithreportingoddratios (ROR)areneededtosupporttheseinitialndings. PUK4 CUMULATIVEEXPOSURETONONSTEROIDALANTI-INFLAMMATORYDRUGS (NSAIDS)ANDTHEPROGRESSIONOFCHRONICKIDNEYDISEASE(CKD)YargerS1,NwokejiE1,TriceS2,ChaoS3,DevineJ2,PotykR2,GutkeG4,BonnemaA3 1DepartmentofDefense/GDIT,FortSamHouston,TX,USA,2DepartmentofDefense,FortSam Houston,TX,USA,3AirForceMedicalSupportAgency,SanAntonio,TX,USA,4Healthcare InformaticsDivision,SanAntonio,TX,USAOBJECTIVES: ToexaminetherelationshipbetweenNSAIDuseandtheprogression ofCKDforacohortofelderlypatientswithinthemilitaryhealthsystem. METHODS: Allpatientswithatleastoneserumcreatininemeasurementineachof thetimeperiodsJuly1December31,2006andJuly1December31,2008witha baselineestimatedglomerularltrationrate(eGFR)representingCKDstage2or3, whowere 67yearsofage,continuouslyeligibleforTRICARE,andreceivedcareat amilitarytreatmentfacilitywereincludedinaretrospectivedatabaseanalysis. LogisticregressionanalyseswereusedtoexploreassociationsbetweenNSAIDuse andrapidprogressionofCKDdenedasadecreaseineGFRof 15mL/min/1.73m2. Covariatesincludedage,gender,diabetes,hyperlipidemia,hypertension,and NSAIDexposurelevel. RESULTS: Atotalof34,295patients(medianage73.5,50.2% male)mettheinclusion/exclusioncriteria.Atotalof14,576patientshadsome NSAIDuse,while19,719hadnoNSAIDuse.RapidCKDprogressionoccurredfor 10.5%ofnouseNSAIDpatients(2,063of19,720),11.2%oflowtomediumNSAIDuse patients(1,465of13,125),and13.4%ofhighNSAIDusepatients(195of1,450). Logisticregressionresultsshowedthatstage2-3CKDpatientswithlowtomediumA74VALUEINHEALTH14(2011)A1–A214

PAGE 83

NSAIDusedidnothaveanincreasedrisk(stage2OR1.002,CI0.925-1.086;stage3 OR0.936,CI0.782-1.122)ofrapidCKDprogressioncomparedtothenoNSAIDexposuregroup.Stage2-3CKDpatientswithhighNSAIDusetrendedtoward,butdid notachieveastatisticallysignicantincreasedrisk(stage2OR1.185,CI0.994-1.413; stage3OR1.276,CI0.844-1.927)ofrapidCKDprogression. CONCLUSIONS: NSAID exposureoveratwoyearperiodwasnotassociatedwithanincreasedriskofrenal dysfunctioninourcohortofelderlypatientswitheitherstage2or3CKD. PUK5 TREATMENTPATTERNSINOVERACTIVEBLADDER:REAL-WORLDPATIENT POPULATIONEllsworthP1,BavendamT2,ZouKH2,MasteyV2,ChenCI2 1BrownUniversity,Providence,RI,USA,2Pzer,Inc.,NewYork,NY,USAOBJECTIVES: Toassesstheprescribing,switching,anddose-adjustmentpatterns ofoveractivebladder(OAB)drugsinareal-worldpatientpopulation. METHODS: A retrospectiveanalysiswasconductedusingtheMedstatMarketscandatabases, whichcontainpharmacyclaimsfor28millionindividualsacrosstheUS.Study patientsreceivedanOABdrugprescriptionbetweenApril1,2009andApril30,2010 withcontinuousenrollment6monthsbeforetheindexdate.Demographics,diagnosisrate,prescriptionlldates,patientswhoswitchedtoorfromagivenOAB drug,andpatientsadjustingdosagewereevaluated. RESULTS: Atotalof214,721 patientshadanOABprescription(meanage65y,76%women),23%werediagnosed withOAB;37%ofpatientsresidedinSouthern,35%inNorthernCentral,18%in Western,and9%inNortheasternregionsoftheUnitedStates.Manypatients switcheddrugtherapyduringthestudyperiod;thosereceivingdrugswithmultiple doseoptionsappearedtohaveoftenmadedoseadjustmentsratherthanswitching toanewagent (Table) .Forexample,amongpatientsprescribedsolifenacin,22%of switchesweretoincreasethedoseand39%weretodecreasethedose.Theproportionofpatientsswitchingtoanotheragentwashigheramongthosereceiving oxybutynin,tolterodineER,andsolifenacinthanwithfesoterodine,darifenacin, andtrospium.Among32,782(15%)patientswhoswitchedOABmedicationatleast once,themedian(range)durationoftimebetweenllingtheirinitialprescription andeitherdose-adjustingorswitchingtoanewagentwas139(0393)daysand theylledamedian(range)of4(241)prescriptionsbeforedose-adjustingor switching. CONCLUSIONS: Inthisreal-worldpopulation,patientstakeadvantage oftheabilitytoadjustbetweenmultipledosesofdrugsforwhichaexible-dose treatmentoptionisavailable.Manypatientsswitcheddrugtherapy,whichsuggestsanopportunitytoimprovemanagementofthesepatientsandtoidentifyan optimaltreatmentpathway. PUK6 LOW-MOLECULAR-WEIGHTHEPARININCHRONICHEMODIALYSISPATIENTS:A SYSTEMATICREVIEWANDMETA-ANALYSISBurkeN1,TarrideJE1,BowenJ1,vonKeyserlingkC2,LimW1,CrowtherM1,GoereeR1 1McMasterUniversity,Hamilton,ON,Canada,2ProgramsforAssessmentofTechnologyin Health(PATH)ResearchInstitue,Hamilton,ON,CanadaOBJECTIVES: Inpatientsundergoingchronichemodialysis,unfractionatedheparin (UFH)iswidelyusedforpreventionofextracorporealcircuitthrombosis.Lowmolecular-weightheparin(LMWH)mayhaveadvantagesintermsofbleedingrisk. Theobjectiveofthisstudywastoreviewtheeffectiveness,safetyandcost-effectivenessofLMWHversusUFHinchronichemodialysistoinformahospitalformularypolicydecision. METHODS: Asystematicreviewwasconductedofrandomized (RCTs)andnon-randomizedtrials(non-RCTs)publishedtoApril2010evaluating theeffectiveness,safetyorcost-effectivenessofLMWHsforuseinhemodialysis comparedwithUFH.Outcomesincludedextracorporealcircuitthrombosis,bleedingevents,vascularaccesscompressiontime,heparin-inducedthrombocytopenia,andcost-effectiveness.Wherepossible,datawaspooledviaarandom-effects model. RESULTS: Afterreviewing998citations,25studieswereincluded(1systematicreview,12RCTs,12non-RCTs).Meta-analysisofRCTsdemonstratednostatisticallysignicantdifferencesbetweenLMWHandUFHforbleeding(relativerisk (RR) 0.70,p 0.66)orcompressiontime(weightedmeandifference 2.49, p 0.31).PooledRRforcircuitthrombosiswas1.32(p 0.05)and0.67(p 0.69)in studiesassessingthenumberofdialysissessionandpatients,respectively.Pooled resultsforthenon-RCTswerenotsignicantlydifferentbetweenLMWHandUFH. Nostudiesevaluatingthecost-effectivenessofLMWHversusUFHwereidentied, however4costanalysesevaluatedtreatmentcosts.DespitehigherdrugacquisitioncostswithLMWH,overallcostswerelowerorcomparablewhenconsidering differencesinnursingtimeassociatedwithadministrationandmonitoring. CONCLUSIONS: LMWHdemonstratednoclearadvantageoverUFHintermsofeffectivenessorsafety,butmayresultinlowerorcomparableoverallhospitalcosts. Urinary/KidneyDisorders–CostStudies PUK7 ECONOMICEVALUATIONOFCINACALCETVERSUSSTANDARDTREATMENTIN PATIENTSWITHSECONDARYHYPERPARATHYROIDISMINMEXICOArochoR1,IanazzoS2,RiveraHurtadoR3,PaniaguaSierraJR4 1Amgen,Inc.,Barcelona,Spain,2AdResHealthEconomics&OutcomesResearch,Torino,Italy,3AmgenMexico,MexicoCity,Mexico,4MexicanInstituteforSocialSecurity(IMSS),MexicoCity, MexicoOBJECTIVES: Secondaryhyperparathyroidism(SHPT)isacommoncomplicationof chronickidneydisease(CKD).Cinacalcetisadrugthatcanregulatethelevelsof plasmaPTH,serumcalciumandphosphorusinpatientsnotcontrolledwithstandardtreatment(vitaminDsterolsandphosphatebinders).Weevaluatedtheuseof cinacalcetvs.standardtreatmentfromaPublicMexicanHealthcareperspective. METHODS: Acost-effectivenessandabudgetimpactanalyseswereperformed withasimulationmodeloverthewholepatient = slifetimeconsidering8-weeks cycles.Efcacydataofcinacalcetweretakenfrompublishedliterature.LocalmortalitydatawasadjustedforlevelsofCa,PandPTHandforRelativeRisk(RR)of deathindialysispatients.Eventsconsideredarecardiovascularevents,fractures andparathyroidectomy.Costsareevaluatedinlocalsettingandaccountfordrugs, dialysisproceduresandevents.Patientswereconsideredinthebudgetimpact (BIA)withapredicteduptakeofcinacalcetinpublicsectorof7%intherstyear. RESULTS: Thesimulatedpatientsinthecinacalcetgroupwithrespecttostandard treatmentgroupexperiencedamean(SD)increaseintheaveragetimeinrange (TiR)of1.91(4.04)patient-yearsonallparametersandalowerrateofhospitalization:RRwere0.43,0.82and0.96respectivelyforPTX,fractureandcardiovascular hospitalization.Consideringapriceforthepublicsectorof$3.29MXP/mg,the averageICERforallparametersresulted$210,060MXP/TiR.BIAwithcinacalcet priceforpublicsectorpredictedanextraexpenditureofapproximately$8MXP millionintherstyearconsideringdrugcostsanditrepresentedthe0.2%ofthe totalhealthcareexpenditureforCKD. CONCLUSIONS: Theuseofcinacalcetin comparisonwithstandardtreatmentisacost-effectivestrategyinpatientswith SHPTanditsuseimpliesaloweconomicimpactascomparedtothetotalhealthcareexpenditureassignedtoCKDinMexico. PUK8 KIT73STUDY:DOESTYPEOFDONORINFLUENCELONG-TERMCOSTSOFCARE? RESULTSFROMPOST-KIDNEYTRANSPLANTPATIENTSINTHEBRAZILIAN PUBLICHEALTHCARESYSTEM(SUS)Tedesco-SilvaHJ1,ManfroRC2,AsanoE3,NitaME3,BarbosaE3,CarvalhoJF3,SzaboSM4, LevyAR4 ,DonatoBM5,RahalE3,TheKIT73StudyGrouP3 1Funda‹oOswaldoRamos-HospitaldoRimeHipertens‹o,S‹oPaulo,SP,Brazil,2Hospitalde Cl’nicasdePortoAlegre,S‹oPaulo,SP,Brazil,3Bristol-MyersSquibbS/A,S‹oPaulo,SP,Brazil,4OxfordOutcomesLtd.,Vancouver,BC,Canada,5Bristol-MyersSquibb,Wallingford,CT,USAOBJECTIVES: Theobjectiveofthisstudywastoestimateresourceuseandassociatedcostsofcareamongrecipientsofkidneytransplantaccordingtodonortype, fromtheperspectiveoftheBrazilianPublicHealthSystem(SUS). METHODS: The KIT73studywasaretrospectiveobservationalstudyconductedat7transplant centers,including498adultsinglekidneyonlytransplantrecipientsfromJan/2004 toJan/2005.Recipientswereidentiedbydonortype(living(LD)ordeceased(DD)). TheDDgroupwasfurtherstratiedbystandardcriteriadonors(SCD)orexpanded criteriadonors(ECD).Five-yearcostswereadjustedto2004atanannualdiscount rateof5%andconvertedto2010USD.Acasemixgroupcostingapproachwasused todetermineaveragecostsperdayofhospitalization,stratiedbycause(surgical orclinicalcomplication)andtype(generalward,IntensiveCareUnit(ICU). RESULTS: All498recipientswereincludedintheanalyses.Forbothgroups(LDor DD),totalinpatientcostsofcarewerehigherthantotaloutpatientcosts(excluding medications)at1and5years.LDrecipientsincurredsignicantlylowercoststhan DDrecipients:mean(SD)costsperpatientwereUSD1,460(USD1,813)(LD)v0.s USD2,785(USD3,615)(DD)(p 0.0001)intherstyear;USD288(USD599)(LD)vs. USD568(USD1,540)(DD)(p 0.01)inthefthyear.ComparisonbetweenSCDand ECDgroups,however,showedthatthemean(SD)costsperpatientwereUSD2,692 (USD3,616)(SCD)vs.USD3,767(USD5,481)(ECD)(p 0.377);USD545(USD1,540)(SCD) vs.USD810(USD1,117)(ECD)(p 0.146)inthefthyear. CONCLUSIONS: Recipients ofkidneytransplantsfromlivingdonorshadlowercostsoverveyearsascomparedtorecipientsoftransplantsfromdeceaseddonors.Amongrecipientsfrom deceaseddonors,nostatisticaldifferencewasobservedinthecomparisonbetweenECDandSCDrecipients. PUK9 MEDICALCARESPENDINGASSOCIATEDWITHCHRONICKIDNEYDISEASEBY STAGEOFDISEASE:APOPULATION-BASEDSTUDYTaylorTN1,SalinitriFD2,SatterthwaiteDW1 1WayneStateUniversity,Detroit,MI,USA,2WayneStateUniversity,OakwoodHospitaland MedicalCenter,Detroit,MI,USAOBJECTIVES: Theobjectiveofthisresearchwastoestimateexpendituresattributabletochronickidneydisease(CKD)bydiseasestageandtoidentifypredictorsof variationinspending. METHODS: Aretrospectiveanalysisofde-identiedhealth insuranceclaims,laboratorydataandpharmacyclaimswasperformedusingthe MarketScanLabResultsdatabasefor2007.Wecalculatedestimatedglomerular ltrationrates(eGFR)usingthe4variablemodicationofdietandrenaldisease (MDRD)equationanddenedcasesasthosewithaneGFR 60mL/min/1.73m2. PresenceofcomorbidconditionswasbaseonICD-9codesreportedontheclaims. Propensityscoremethodswereusedtomatchcontrolstocases.Datawereanalyzedusingdescriptivestatisticsandgeneralizedlinearmodelswithagamma distributionandlog-linkcontrollingfordiseasestage,healthplantype,comorbid conditions,anddemographiccharacteristics. RESULTS: Atotalof2,436caseswere identiedwithaneGFRoflessthan60mL/min/1.73m2.Wewereabletomatch controlsto2,223(91.3%)oftheidentiedCKDcasesusingpropensityscores.Unadjustedexpenditureswere$10,353(sd23,262)forCKDcasescomparedto$9,310 (sd22,703)formatchedcontrols.Afteradjustingforhealthplantype,patientdemographiccharacteristics,andcomorbidconditions,adjustedexpenditureswere signicantlygreaterforpatientswithCKDcomparedtomatchedcontrolsdependingonCKDstage.Expendituresbystagerangedfrom8%higherforpatientswith stage3diseaseto62%higherforpatientswithstage5CKDcomparedtomatched controls.Thekeypredictorsoftotaltreatmentpaymentswerediseasestage,presenceofcomorbidconditions,age,andgender. CONCLUSIONS: PaymentsformedicalcareattributabletoCKDaresignicantlygreaterthanmedicalcarepayments forthosewhodonothaveCKDandvaryimportantlybydiseasestage.InterventionsthatpreventordelaytheprogressionofCKDhavethepotentialtosave considerableresources.A75VALUEINHEALTH14(2011)A1–A214

PAGE 84

PUK10 KIT73STUDY:RELATIONSHIPBETWEENCOSTSOFCAREAND12-MONTH GLOMERULARFILTRATIONRATESINPOST-KIDNEYTRANSPLANTPATIENTSIN THEBRAZILIANPUBLICHEALTHCARESYSTEM(SUS)David-NetoE1,EsmeraldoRM2,AsanoE3,NitaME3,NishikawaAM3,BarbosaE3, SzaboSM4 ,LevyAR4,CarvalhoJF3,DonatoBM5,RahalE3,TheKIT73StudyGrouP3 1HospitaldasCl’nicasdaFaculdadedeMedicinadaUSP,HCFMUSP,S‹oPaulo,SP,Brazil,2HospitalGeraldeFortaleza,Fortaleza,Ceara,Brazil,3Bristol-MyersSquibbS/A,S‹oPaulo,SP, Brazil,4OxfordOutcomesLtd.,Vancouver,BC,Canada,5Bristol-MyersSquibb,Wallingford,CT, USAOBJECTIVES: Theobjectiveofthisstudyistoanalyzetherelationshipbetween 12-monthrenalfunctionandcostsinpost-kidneytransplantpatientsfromthe BrazilianPublicHealthSystem(SUS)perspective. METHODS: Non-interventional, multicenter,retrospectivemedicalchartreviewofpatientsthatunderwentkidney transplantation.Threeyearsfollow-updataonresourceuseandassociatedcosts fromadults(age 18years-oldattimeoftransplantation),singlekidneyonlytransplantsfromJan/2004toJan/2005werecollectedfrom7transplantcentersinBrazilianhospitals.Datawerecensoredongraftloss.Estimatedglomerularltration rates(eGFR)at12-monthpost-transplantationwerecalculatedusingtheabbreviatedModicationofDietinRenalDiseaseequationandstratiedaccordingtothe NationalKidneyFoundationK/KidneyDiseaseOutcomesQualityInitiativerenal functioncategories.Costswereadjustedtoyear2004atanannualdiscountrateof 5%andconvertedto2010USD.Casemixgroupcostingapproachwasusedto determineaveragecostperdayofhospitalization,stratiedbycause(surgicalor clinicalcomplication)andtype(generalward,ICU). RESULTS: 498subjectswere eligiblefortheanalysis.Outpatientcareresourceuseandassociatedcosts(excludingcostswithimmunosuppressivetherapy)didnotsignicantlyvaryamongpatientsindifferenteGFRcategories,withaverageannualcostsrangingfrom664and 921USDintherstyear,194and245USDinthesecondand129and165USDinthe thirdyearpost-transplantation.However,atrendtowardsofincreasinginpatient costs(p 0.001)andincidenceofhospitalizationperyearwithdecliningofrenal functionwasobserved.After3yearspost-transplant,41%ofstage1ChronicKidney Disease(CKD)and91%ofstage4patientswerere-admitedtothehospitalatleast onceincurringanestimatedaveragecostof1,239and3,654USD,respectively. CONCLUSIONS: Renalfunction,measuredby12-monthpost-transplanteGFR,isan importantdeterminantofcostsofhospitalizationinkidneygraftrecipients. PUK11 UPDATEDECONOMICANALYSISOFFESOTERODINERELATIVETO TOLTERODINEANDSOLIFENACINFORTHETREATMENTOFOVERACTIVE BLADDER:THESWEDISHPERSPECTIVELeeR1,SnedecorSJ2,KvaszMG3,TrocioJ4,BorgmanB5 1WeillMedicalCollegeofCornellUniversity,NewYork,NY,USA,2PharmeritNorthAmerica, LLC,Bethesda,MD,USA,3Pzer,Paris,France,4Pzer,Inc.,NewYork,NY,USA,5PzerSweden, Sollentuna,SwedenOBJECTIVES: Toquantifythecost-effectivenessofoveractivebladder(OAB)treatmentwithfesoterodine(FESO)relativetoextendedreleasetolterodine(TER)and solifenancin(SOL)inSwedenbasedonrecently-availabledata. METHODS: A52weekdecision-treemodelwasdevelopedusingdatafromfour12-week,randomizedclinicaltrialsandpublishedliterature.Patientswhodidnotachieveresolution ofurgencyurinaryincontinence(UUI)atweek4weretitratedtohigherdosesinthe FESOandSOLmodelarms.TERisonlyavailableinonedoseandallpatientsthereforeremainedonTERatweek4regardlessofUUIresolution.ThepublisheddecreaseinUUIepisodeswasusedtoestimatetheproportionofUUIresolutionwith SOL.TrialdiscontinuationdatawerettoaWeibullsurvivalmodelandextrapolatedto52weeks.Changesinhealth-relatedqualityoflifewereassessedusing OveractiveBladderQuestionnairedatafromtwoofthetrialsandweretransformed intopreference-basedutilityvalues.Regressionanalysisdeterminedtheassociationbetweentrialutilitiesandmicturition/UUIepisodesinordertoestimatevalues forSOL.Medicalcostsincludedantimuscarinicdrugs,physicianvisits,laboratory tests,incontinencepadsandcostsofOAB-orincontinence-relatedcomorbidities. Probabilisticsensitivityanalysiswasusedtodeterminetheimpactofuncertainty inthemodelinputsonitsoutcomes. RESULTS: 19.5%,18.0%,and16.3%ofpatients receivingFESO,TER,andSOL,remainedontreatmentandwerecontinentafter52 weeks.Therespectivequality-adjustedlifeyearsassociatedwithFESO,TER,and SOL,were0.762,0.756and0.752;overallcostswerekr16,752,kr16,849,andkr17,047, indicatingthatFESOisalower-costandmoreeffectivetreatmentoptionthanTER andSOL.SensitivityanalysisconrmedthesuperiorityofFESOinover80%of modelsimulations. CONCLUSIONS: ResultssuggestFESOhaslowercostsand higherefcacyandthusdominatesbothTERandSOLintreatingOABwithUUIin Sweden. PUK12 COST-EFFECTIVENESSOFTOLTERODINEASTREATMENTFOROVERACTIVE BLADDER(OAB)INADULTMEXICANPATIENTSArreola-OrnelasH1,Rosado-BuzzoA2,Garcia-MollinedoM2,Camacho-CorderoL2, Muci–o-OrtegaE3 ,Mould-QuevedoJF4,Galindo-SuarezRM3 1Fundaci—nMexicanaparalaSaludAC,MexicoCity,Mexico,2Links&LinksS.A,deC.V.,Mexico City,Mexico,3Pzer,MexicoCity,Mexico,4Pzer,NewYork,NY,USAOBJECTIVES: Worldwideprevalenceofoveractivebladder(OAB)iscloseto10%in adults.Theurgentfeelingtogotothetoilet,frequencyofvoidsandurinaryincontinenceaffectslifestyleandproductivityofpatients.Theaimofthisstudywasto estimatethecost-effectivenessoftolterodineinthemanagingofOABfroman institutionalperspective. METHODS: Afour-stateMarkovmodelwasperformedto estimatehealthandeconomicconsequencesduringatimehorizonofoneyear (one-monthcycles).Effectivenessmeasuresweremeanpercentagereductionin: numberofvoidsperday,numberofurinaryurgencyeventsperdayandnumberof urinaryincontinenceeventsperday.Transitionprobabilitieswereobtainedfroma meta-analysisemployinginternationalpublishedliterature.Comparatorswere tolterodine(2mgbid)andtabletsofoxybutynin(5mgtid)asreferencetreatment. ResourceusewasobtainedfromtheInstitutoMexicanodelSeguroSocial(IMSS) databases(n 377cases).Costswereextractedfrominstitutional(IMSS)ofcial sources.Costsincluded:visitstogeneralpractitionersandurologists,hospitalization,drugs,medicalprocedures,laboratorytests,imagenologyandadverseevents management.Probabilisticsensitivityanalyseswereperformedemployingbootstrappingtechniques.Acceptabilitycurveswereconstructed. RESULTS: Tolterodineandoxybutyninannualcostsperpatientwere:US$1,654.55[95%CIUS$1,486.97US$1,822.13]andUS$1,537.38[95%CIUS$1,389.42-US$1,685.33],respectively (p 0.05).Tolterodineexhibitsbetteroutcomesthanreferenceinalleffectiveness measuresconsidered(p 0.05). Incrementalcost-effectivenessratiosforadditional percentagereductionof:numberofvoids,eventsofurinaryurgencyandeventsof urinaryincontinencewere:US$1,307.34[95%CIUS$1,268.73-US$1,345.96];US$1,091.39 [95%CIUS$1,059.15-US$1,123.62]andUS$2,915.31[95%CIUS$2,829.14-US$3,001.42], respectively.AcceptabilitycurveshowedthattheprobabilityoftolterodinebeingcosteffectiveisclosetooneatawillingnesstopayofUS$3,170(regardingpercentreduction ofnumberofvoids). CONCLUSIONS: Tolterodineismoreeffectivethanoxybutynin withsimilartreatmentcostsrepresentingamoreefcientalternativeinthemanagementofOABattheIMSS. PUK13 COST-EFFECTIVENESSANALYSISOFTHEEARLYCONVERSIONOFTACROLIMUS TOMTORINHIBITORSINPATIENTSWITHRENALTRANSPLANTATIONGamboaO1,MonteroC2,MesaL3,BenavidesC4,ReinoA5,CastilloJS6 1Fundaci—nEsensa,Bogot‡,Colombia,2ClinicaUniversitariaColombia,Bogot‡,Colombia,3FundacionValledeLili,Cali,Colombia,4Fundaci—nCardioInfantil,Bogot‡,Colombia,5Hospital SanVicentedePaul,Medell’n,Colombia,6InstitutodelaEvaluaciondelaCalidadyAtenci—nen salud(IECAS),Bogot‡,ColombiaOBJECTIVES: Toevaluatethecost-effectivenessoftheearlyconversionoftacrolimustomTORinhibitors(mTORi)-sirolimusoreverolimus-againstcontinuous treatmentwithtacrolimusinpatientswithrenaltransplantationinColombia. METHODS: AMarkovmodelsimulatingpatient'snaturalhistorywithrenaltransplantationwasconstructed.Themodelwasdesignedtopredicttheincidenceof rejectionepisodes,associateddeaths,adverseeventsandgraftloss.Anindividuallevelsimulationwasperformed,generating100,000patientsthatstartedinthe reject'sfreestage,usingmonthlycyclesupondeathorColombianlife-expectancy (76years).Themodelassessedseveralriskfactors(creatininelevels,rejectionepisodesanddiabetes)associatetograftloss.Thetransitionprobabilitieswereextractedfromasystematicreviewofpublishedliterature.Directmedicalcostswere gatheredfromofcialdatabases(SISMED's2010rstsemesterandtariffsmanuals). Thethirdpartypayer'sperspectivewasusedintheanalysis.Theeffectiveness measureusedwas:life-yearsgained(LYG) .A5% discountratewasappliedforthe costsandhealthoutcomes.Cost-effectivenessratio(CER)andincrementalcosteffectivenessratio(ICER)wereestimated.Lastly,deterministicandprobabilistic sensitivityanalyseswerealsoperformed. RESULTS: Comparedtotacrolimus, sirolimusshowedafavorabledifferenceof0.74LYG(10.26vs.9.52);withaCERof US$37,302andUS$39,731;respectively.Inaddition,ICERbetweenthetwomentionedalternativesresultedinUS$6,130.Foreverolimusnopublishedevidenceof efcacywasfoundfortheevaluatedscheme,thereforetheeffectivenessofeverolimuswasassumedsimilartosirolimusexhibitinganICERofUS$70.440against tacrolimus.Sensibilityanalysesmodifyingthislastassumptionshowedthemodel resultsarerobust. CONCLUSIONS: Thestudysuggeststhatsirolimusstrategyis cost-effectiveinColombiaforpatientswithrenaltransplantationusingthe3GDP percapitathresholdasrecommendedbytheWorldHealthOrganization. PUK14 AMARKOVMODELCOMPARINGSACRALNEUROMODULATIONAND BOTULINUMTOXIN-AFORMEDICAREPATIENTSWITHIDIOPATHIC OVERACTIVEBLADDERREFRACTORYTOCONSERVATIVECAREGanzM1,ClemensJQ2,AngerJ3,DenevichS1,ShahD1,CarlsonA4,WittekMR5, PashosC1 1UnitedBioSourceCorporation,Lexington,MA,USA,2UniversityofMichigan,AnnArbor,MI, USA,3Cedars-SinaiMedicalCenter,BeverlyHills,CA,USA,4DataIntelligenceConsultants,LLC, EdenPrairie,MN,USA,5Medtronic,Inc.,Minneapolis,MN,USAOBJECTIVES: Fewpublishedmodelscomparingsacralnerveneuromodulation (SNM)withintravesicalbotulinumtoxin-Ainjections(BTX)forthetreatmentof idiopathicoveractivebladder(OAB)includecomprehensivespecicationsofclinicalpathwaysandcosts.Wedevelopedamodelthataccuratelycapturestheclinical andeconomicrealitiesofSNMandBTXtherapies. METHODS: OurMarkovmodel, whichtakestheUnitedStatesMedicareperspective,comparestwocohortsofhypotheticalpatientswithidiopathicOABrefractorytoconservativecare:thoseinitiatingwithSNM(InterStim¨Therapy,Medtronic,Inc.Minneapolis,MN)andthose initiatingwithBTX(BOTOX¨,Allergan,Irvine,CA).Bothcohortsaresimilarand equallylikelytoreceiveeithertherapy.ThetimehorizonistheSNMneurostimulatorbatterylife(5.5years).Themodelaccountsfortreatmentsuccessprobabilities,adverseevents(AEs),andSNM-relatedadjustmentsbasedonclinicalliteratureandexpertclinicalinput.Quality-adjustedlife-years(QALYs)forpersistent OABandsuccessfultreatmentcomefromNationalOveractiveBLadderEvaluation (NOBLE)study.Patientsnotrespondingsuccessfullytoinitialtherapydiscontinue orswitchtherapies(patientstryeachtherapyonce).Costs(manufacturerpricing and2010Medicarepaymentschedules)andbenetsarediscounted3%perannum. RESULTS: Attwoyears,SNMwasmorecostly($20,240vs.$17,069perpatient)and moreeffective(1.60vs.1.56QALYsperpatient).SNMbecamelesscostlyat3years.A76VALUEINHEALTH14(2011)A1–A214

PAGE 85

At5.5years,initialSNMtherapywaslesscostly($23,504vs.$27,720)andmore effectivethaninitialBTX(4.08vs.4.04QALYs).Probabilisticsensitivityanalyses thatvariedSNMandBTXsuccessandrepeatBTXinjectionprobabilitiesandutilities,conrmedtheseresults.RepeatinjectionsanddifferencesinAEswereresponsibleformostofthechangingcostsovertime. CONCLUSIONS: Basedonamore clinicallycomprehensivedesignandsetofinputsthanpreviousmodels,treatment withSNMmaybemorecost-effectivethanBTX,especiallyoverlongerperiodsof time. PUK15 THECOST-EFFECTIVENESSOFLANTHANUMCARBONATEINTHETREATMENT OFHYPERPHOSPHATEMIAINDIALYSISPATIENTSFROMACANADIAN PERSPECTIVEVegterS1,TolleyK2,LevinA3,LokCE4,MortonAR5,SorokaSD6,KeithMS7,PostmaMJ1 1UniversityofGroningen,Groningen,TheNetherlands,2TolleyHealthEconomics,Buxton,UK,3StPaul'sHospital,Vancouver,BC,Canada,4UniversityofToronto,Toronto,ON,Canada,5Queen'sUniversity,KingstonGeneralHospital,Kingston,ON,Canada,6DalhousieUniversity, QueenElizabethIIHealthSciencesCentre,Halifax,NS,Canada,7ShirePharmaceuticals,Wayne, PA,USAOBJECTIVES: HyperphosphatemialeadstoincreasedhospitalizationsandmortalityinEnd-StageRenalDisease(ESRD).First-linetherapyinCanadaconsistsprimarilyofcalciumcarbonate(CC).Wedeterminedtheincrementalcost-effectiveness ratio(ICER)ofthenon-calciumphosphatebinderlanthanumcarbonate(LC)as second-linetherapy,fromaCanadianhealthcareperspective. METHODS: A Markovmodelwasdevelopedtodeterminethecost-effectivenessofsecond-line LCaftertherapyfailureonCC,comparedwithcontinuedCCtreatment;oralternativelywithsecond-lineSevelamer(SH).Patient-leveldata(n 380)fromaprospectiverandomizedtrialwereusedforLCandCCdrugefcacy.ForSHefcacy,an indirectcomparisonofeightclinicaltrialswasusedtocalculateadose-relativity betweenSHandLCof2.7:1.Costs,qualityoflife,mortalityandhospitalizationrates werebasedonCanadiandata.Univariateandprobabilisticsensitivityanalyses wereperformed. RESULTS: Modelling1,000dialysispatients,378(37.8%)didnot achievetargetserumphosphate(SP)levels( 1.78mmol/L)withrst-lineCCtherapyandwereeligibleforLC.Ofthese,168(44.4%)respondedtoLCtherapy,resultingin49lifeyearsand29Quality-adjustedlifeyears(QALYs)gained.TheICERof second-lineLCtreatmentcomparedwithcontinuedCCtreatmentwasCAN$13,200 ($4,600-$22,800)perQALYgained.Resultswererobusttoplausiblevariationsin modelparameters.One-yeardrugcostsperadditionalrespondertosecond-lineLC therapywere$2,600,comparedto$4,300forrst-lineLC.Themodelevaluating second-lineuseofLCvssecond-lineuseofSHestimatedthatLChadsimilarefcacybutwas16%lessexpensivethanSH. CONCLUSIONS: Second-linetreatment withLCiscost-effectivecomparedtocontinuedtherapywithCC;andislessexpensivecomparedtorst-lineLC.LChadlowercostscomparedwithSH,dueto lowerdoserequirementsforsimilarefcacy.TheseresultsreinforcecurrenttreatmentguidelinestotreatpatientsnotachievingtargetSPlevelsonCCwithsecondlineLC. PUK16 COST-EFFECTIVENESSOFVALGANCICLOVIR200DAYSPROPHYLAXISVERSUS 100DAYSPROPHYLAXISINKIDNEYTRANSPLANTPATIENTSATHIGH-RISK FORDEVELOPINGCYTOMEGALOVIRUSDISEASEVicenteC1,HumarA2,LemieuxC3,ZilbershteinR1,DouglasP4,FormicaL4,PiwkoC1 1PIVINAConsultingInc.,Thornhill,ON,Canada,2UniversityofAlberta,Edmonton,ON,Canada,3CentreHospitalierdel'UniversitŽdeMontrŽal,Montreal,QC,Canada,4Hoffmann-LaRoche Limited,Mississauga,ON,CanadaOBJECTIVES: Cytomegalovirus(CMV)diseasecanseverelyimpactpatientoutcomesandcostsassociatedwithkidneytransplant.TheIMPACTstudyhasshown thatkidneytransplantpatientssignicantlybenetfromextendingvalganciclovir prophylaxisfrom100to200days.Theobjectiveofthisstudywastodeterminethe cost-effectivenessofvalganciclovir200dayprophylaxiscomparedto100daysin kidneytransplantpatientsathigh-riskfordevelopingCMVdisease. METHODS: A Markovmodelwasdevelopedtocapturetimespentbypatientsinvarioushealth states,whichincluded:CMV,No-CMV,AcuteRejection,GraftFailure,Dialysisand Death.Resultswerereportedasincrementalcostperadditionalqualityadjusted life-years(QALY)gained,overa10-yearperiod.Transitionprobabilitiesfortherst yearwerederivedfromtheIMPACTstudy.Databeyondtherstyearwerederived fromthepublishedliteratureandbaselinemortalityratewasdeterminedfromthe CanadianOrganReplacementRegister.Thebasecaseanalysisfocusedondirect medicalcostsonlyfromtheperspectiveoftheMinistryofHealth(MoH).Asecond analysiswasconductedfromthesocietalperspective.Costdatawereobtained fromavarietyofsourcesandreportedas2010CanadianDollars.A5%discountrate wasappliedtobothcostsandpatientoutcomes.Multiplesensitivityanalyseswere undertakentotesttherobustnessofthemodel. RESULTS: FromtheMoHperspectivevalganciclovir200daysprophylaxisiscost-effectivewhencomparedto100 dayswithanincrementalcost-utilityratio(ICUR)of$34,818peradditionalQALY gained.Thecost-effectivenessisimprovedfromthesocietalperspective,withan ICURof$32,571peradditionalQALYgained.Resultswererobustoverawiderange ofsensitivityanalysestested. CONCLUSIONS: Valganciclovir200daysisacosteffectiveprophylaxisstrategyinkidneytransplantpatientsathigh-riskofdevelopingCMVwhencomparedtovalganciclovir100days. PUK17 UTILIZATIONANDECONOMICIMPACTOFIVIRONANDERYTHROPOIESIS STIMULATINGAGENTSINCHRONICKIDNEYDISEASEPATIENTS:A MULTI-HOSPITALSTUDYJoshiAD,HoldfordD,BrophyD,HarpeSE VirginiaCommonwealthUniversity,Richmond,VA,USAOBJECTIVES: TherateandextentofutilizationofIVironinanemicCKDpatients wasquantied.PredictorsofIVironandESAuseweredetermined.TheimpactofIV ironandESAusewasexaminedseparatelyfortotalhospitalcostsandlengthof stay(LOS). METHODS: ThisisaretrospectivecohortanalysiswithintheUHCdata warehouseintheperiodofJanuary1,2006,toDecember31,2008.Inclusioncriteria wereage 18yearswithaprimary/secondarydiagnosisofCKD.Theexposureof interestwasIVironandESAtherapy,andtheoutcomewasthedifferenceintotal hospitalcostsandlengthofstay.AbinomiallogisticregressionusingtheGEE methodologywasusedtoidentifypredictorsofIVironutilization.Propensity scoreswereusedtocontrolforconfounding.AGEEmodelusinggammadistributionandloglinkwasusedtodeterminetheadjustedhospitalcostandlengthof stayfortheIVironandESAandESAalonetherapygroups. RESULTS: 82,947patientsmetthestudycriteria.Ofthe82,947CKDpatientsonESAtherapy,only8% (n 6678)patientswereonIVironsupplementation.Age,race,primarypayer, admissionstatus,severityofillness,dialysisstatusandphysicianspecialtywere identiedasstrongpredictorsofIVironuse.ForpatientsusingbothIVironandESA (n 6678),meancostswere$34,756comparedto$31,404forESAusersalone (n 76,269).TheoverallmeanLOSforallpatientswas9.75days.ForthoseusingIV iron,theLOSwas10.71days,andforthoseonlyusingESA,theLOSwas9.66days. CONCLUSIONS: OurinvestigationshowedsignicantreductioninESAdoseswith theuseofIVironsupplementation,however,theoverallprevalenceofIViron usagewaslow.Intravenousironuserswereassociatedwithahighertotalhospital costandlongerlengthofstaythanESAusers. Urinary/KidneyDisorders–Patient-ReportedOutcomes&Preference-Based Studies PUK18 THEIMPACTOFAUTOSOMALDOMINANTPOLYCYSTICKIDNEYDISEASE (ADPKD)ONPATIENTS’HEALTHRELATEDQUALITYOFLIFE(HRQOL): DEVELOPMENTOFACONCEPTUALFRAMEWORKColeJC1,KrasaHB2,ChengR1,OberdhanD2,UrwongseJ3,SchaeferC3,CzerwiecF2 1CovanceMarketAccessServices,Inc.,SanDiego,CA,USA,2OtsukaPharmaceutical Development&Commercialization,Inc.,Rockville,MD,USA,3CovanceMarketAccessServices, Inc.,Gaithersburg,MD,USAOBJECTIVES: Theaimofourstudywastoconstructtheconceptualframeworkof theimpactofADPKDonpatientstosupporttheuseordevelopmentofaHRQoL instrumentinADPKD. METHODS: Basedonconceptsidentiedfromtheliterature andADPKDphysicians/researchersfromNorthAmerica,Europe,andJapan,patientswereaskedhowADPKDimpactstheirphysicalandsocialfunctioning,emotions,andurinarysymptomsaccordingtoapre-denedmoderatorguide.Twenty focusgroups(FGs)wereconductedacross11sites:threecitiesinNorthAmerica (n 42),sixcitiesinEurope(n 64),andtwocitiesinJapan(n 11).FGsweremoderatedbynativespeakersandoverseenonsitebyaUS-basedscientist.Asaturation tablewasdevelopedtosummarizeconceptsdiscussedintheFGs;saturationwas achievedoncenonewconceptswereidentied.Conceptswereidentiedbasedon themesmentionedby 2participantswithinaFG. RESULTS: Conceptsgenerated fromtheliteratureandADPKDphysicians/researchersdiscussionswereendorsed byADPKDpatients.Agreementofconceptsbetweengendersacrossallcountries wasobserved.Twenty-eightconceptswereidentiedandcategorizedinto:PhysicalImpact(impactonwork/housework,limitedfunctioningwithmild/moderate exertion,self-care,diminishsex/intimacy,pain/discomfortinextremities/core, painaffectingwork/housework,painoccurringwithactivity,modicationsinlifestyle),EmotionalImpact(fatigue,depression,anxiety,guiltofpassingittochildren, acceptance/self-education),UrinaryConcerns(urgency,frequency,nocturia). Otherconceptsidentiedbypatientsincludeeffectondiet,concernoverbody image,thirst,anddisruptionofsocial/leisureactivities.Fifty-sevenpercentof theseconceptswereidentiedintherstFG;100%saturationwasachievedinthe fthFG. CONCLUSIONS: Identiedconceptswereuniversallyapplicableandtheir strengthisevidentbytheachievementofcompletesaturationafter25%oftheFGs. Thecompletionofthisconceptualframeworkusinganiterativeprocessprovidesa strongbasistodevelopanADPKD-specicHRQoLinstrument. Urinary/KidneyDisorders–HealthCareUse&PolicyStudies PUK19 KIT73:PRELIMINARYREPORTSOFIMMUNOSUPRESSANTTHERAPYPATTERNS INACOHORTOFPOSTKIDNEYTRANSPLANTPATIENTSINBRAZILCarvalhoDDBM1,GarciaV2,Abbud-FilhoM3,NishikawaAM4,AsanoE4,NitaME4, SzaboSM5 ,LevyAR5,CarvalhoJF4,DonatoBM6,RahalE4,TheKIT73StudyGrouP4 1HospitalGeralBonsucesso,RiodeJaneiro,Brazil,2UniversidadeFederaldeCinciasdaSaœdede PortoAlegre,UFCSPA,PortoAlegre,Brazil,3ComplexoHospitalarSantaCasadePortoAlegre, S‹oJosedoRioPreto,SP,Brazil,4Bristol-MyersSquibbS/A,S‹oPaulo,SP,Brazil,5Oxford OutcomesLtd.,Vancouver,BC,Canada,6Bristol-MyersSquibb,Wallingford,CT,USAOBJECTIVES: TheaimofthisstudywastocharacterizeBrazilian-specicimmunosuppressant(IS)treatmentpatternsamongkidneytransplantpatients. METHODS: Non-interventional,multicenter,medicalchartreviewofpatientsundergoingkidneytransplant.Fiveyearfollow-updataonISmedicationusefromadult( 18 yearsoldattimeoftransplantation),singlekidneyonlytransplantsfromJan/2004 toJan/2005werecollectedfrom7transplantcenterinBrazilianhospitals.Patients werecensoredongraftlossorlossoffollow-up. RESULTS: Dataon498patients wereincluded;51.2%werefemale,the,averageagewas38.9 13.9years,and55.4% wasrecipientsofkidneysfromdeceaseddonors.Ofthe498patients,61.2%were treatedwithtacrolimus(TAC)-basedISregimensastheinitialcalcineurininhibitor (CNI),while38.8%weretreatedwithcyclosporine(CyC).CNIswitchingfromTACto CyCoccurredin4.6%,versus5.6%thatswitchedfromCyCtoTAC.TherewasaA77VALUEINHEALTH14(2011)A1–A214

PAGE 86

decreasingtrendinaverageCNIdosebyyear:attheendoftherstyear,themean doseofTACwas9.3 5.4mg/dayand282.1 118.3mg/dayforCYC,decreasingto 5.2 3.6and158.7 57.8mg/dayinthethirdyear,and4.8 7.3and144.2 58.4mg/ dayinthefthyear.Attheendoftherst(n 455)andsecondyear(n 408)posttransplant,themostcommonISregimenswereTAC/prednisone/mycophenolate mophetil(MMF)(19.1%inbothyears),TAC/prednisone/azathioprine(15.6%atrst yearand14.7%atsecondyear)andCyC/prednisone/azathioprine(13.2%atrst yearand14.0%atsecondyear).Bytheendofthefthyear(n 141),however,the mostcommonISregimenwasTAC/prednisone/mycophenolatesodium(MFS; 31.9%),followedbyCyC/prednisone/MFS(7.8%)andTAC/prednisone/MMF(7.1%). CONCLUSIONS: TacrolimuswastheCNIofchoiceforthemajorityof denovo kidney transplantpatientsin2004.ThemainIStherapyregimensduringtherstposttransplantyearswastacrolimus/prednisone/mycophenolatemophetil. PUK20 EARLYDETECTION,SCREENING,ANDMANAGEMENTOFCHRONICKIDNEY DISEASEAMONGACTIVELYEMPLOYED–ANINTEGRATEDPOPULATION HEALTHMANAGEMENTAPPROACHHarshmanR1,NaimA2,CarterJ1,RichersonG1,NairKV3 1SouthernCompany/GeorgiaPower,Atlanta,GA,USA,2CentocorOrthoBiotechServices,LLC, Horsham,PA,USA,3UniversityofColorado,Aurora,CO,USAOBJECTIVES: TheemployerburdenofCKDintermsoflostproductivity,shortand longtermdisabilityuse,andhightotalhealthcarecostshasbeenwell-documentedandwarrantsanemployer-sponsoredpopulationhealthmanagement programtoimprovethehealthandlivesoftheworkforce.GeorgiaPowerCompany (GPC)hasimplementedachroniccaremanagementprogramaimedatearlyidentication,diseaseawareness,andcounselingofemployeesthroughon-site screenings. METHODS: IndividualsareofferedvoluntaryparticipationintheCKD managementprogramwiththeirPCPsandnephrologistsdependingupontheirrisk andCKDstage.Healthoutcomesincluding,clinical,resourceutilization,andselfreportedhealthstatusandproductivityarecomparedpre-andpost-program implementation. RESULTS: Preliminaryresultsatthe6-monthmarkshowthat:a) 2,589employeeswerescreened,638(25%)metprogramcriteriaforparticipation and110(17.2%)agreedtoparticipateinthestudy;b)amongthecurrentenrollees, 17%havediabetesand51%havehypertension;c)meaneGFRratesare61.27,and HbA1Clevelsof7.7,andameanBMIof30.5indicatingapopulationathighriskfor developingCKD;d)participantsreportedmissingonaverage10.5hours/weekdue totheirCKD;ande)baselinetotalhealthcareexpenditureswere$19,776permemberperyearindicatingahighcostpopulationaswell. CONCLUSIONS: CKDisa high-costdiseaseforGPC.ResourcesinvestedincreatingnovelCKDmanagement programstoidentify,raiseawareness,andmanageCKDareaworthwhileinvestmentforemployers. PUK21 PATIENTCHARACTERISTICSASSOCIATEDWITHINITIATIONOFOVERACTIVE BLADDER(OAB)DISCUSSIONWITHAPHYSICIANBotrosS1,ChenCI2,BavendamT2,ZouKH2,GorenA3,GuptaS4 1NorthShoreMedicalGroup,Evanston,IL,USA,2Pzer,Inc.,NewYork,NY,USA,3Kantar Health,NewYork,NY,USA,4KantarHealth,Princeton,NJ,USAOBJECTIVES: ManypatientswithOABattemptself-managementandarereluctant toinitiateanOABdiscussionwithaphysician.WesoughttoidentifypatientcharacteristicsassociatedwithpatientinitiationofanOABdiscussionwithaphysician. METHODS: Of24,866respondentsofthe2009NationalHealth&WellnessSurvey, aninternet-basedquestionnaireonhealthcareattitudes,behaviors,andoutcomes, 2750recontactedrespondentsqualiedforandcompletedalongitudinalsurvey. Eligiblesubjects( 18y)hadanOABAwarenessTotalscoreof 14(men)or 16 (women)orusedanOABprescriptionmedication.Exclusioncriteriaincludedcurrentpregnancyorcatheteruse,hematuria,urinarytractinfectionsymptoms,benignprostatichyperplasia(BPH),useofBPHmedication,orprostatecancer.Analysisofproportionsandlogisticregressionanalysis(2-tailedP 0.05signicance level)identiedpatientcharacteristicsassociatedwithinitiationofOABdiscussionswithaphysician. RESULTS: 1325of2750(48%)OABpatientsreportedinitiatinganOABdiscussionwithaphysician.Theserespondents(meanage52y)were predominantlyfemale(65%)andwhite(71%).Signicantvariablesassociatedwith initiationofanOABdiscussionweredivorced/separated/widowedvssingle,more familiarwithOAB,alongerdurationofbladdercontrolsymptoms,lowerSF-12 PhysicalComponentSummaryscore,betterdailyactivityfunction,greaterpill burden,andregularphysiciancontact(Table).Patientage,income,employment status,andhealthinsurancecoveragewerenotsignicantlyassociatedwitha patient-physicianOABdiscussion. CONCLUSIONS: PatientsaremorelikelytoinitiateanOABdiscussionwiththeirphysicianiftheyhaveregularcontactwiththe physician,alongerdurationofsymptoms,arefamiliarwithOAB,andhaveworse physicalqualityoflife.Socioeconomicvariablesordruginsurancecoveragewere notsignicantlyassociated.Thesendingssuggestthatthephysicianrelationship isimportantforhelpseekingbehavior;andthereisaneedforearlypatienteducationonOABsymptomsandtreatments. Urinary/KidneyDisorders–ResearchonMethods PUK22 LEVELINGTHEPLAYINGFIELD:ACASESTUDYONTECHNICALPRECISIONIN COMPARATIVEEFFECTIVENESSRESEARCHFORCLINICALLYLOCALIZED PROSTATECANCER(PC)HughesKE1,DuffSB2,HusainN1,CooperbergMR3 1AvalereHealthLLC,Washington,DC,USA,2VeritasHealthEconomicsConsulting,Carlsbad, CA,USA,3UniversityofCalifornia,SanFrancisco,SanFrancisco,CA,USAOBJECTIVES: Determinetheimpactofcorrectingfordifferentdenitionsofprostate-specicantigen(PSA)recurrencewhenanalyzingrelativeeffectivenessof seventreatmentsforlocalizedPC;Emphasizetheimportanceofclinicalinputand real-worlddatainensuringmodelingaccuracy. METHODS: Weconductedtwo lifetimecost-utilityanalysescomparingmenundergoingopen,laparoscopic,or robot-assistedradicalprostatectomy,3-dimensionalconformalorintensity-modulatedradiationtherapy,brachytherapy,orcombinedexternal-beamradiation andbrachytherapy.WeconstructedaMarkovmodelquantifyinglifetimecostsand quality-adjustedlifeyears(QALYs)formenwithlocalizedPCinlow-,intermediate-,andhigh-riskstrata.Post-treatmentPSArecurrenceisdeneddifferentlyfor surgicalversusradiationtreatments.InStudy1,wemadeclinically-basedcorrectionsintimebetweenrecurrenceandmetastasis,thebasisfortime-to-diseaseprogression(biochemicalfailure[BCF])calculations,toaccountforthedifferent denitions.InStudy2,thesecorrectionswerenotmade. RESULTS: In"corrected" analyses,surgerytendedtoyieldmoreQALYsandlowercoststhanradiation,with minimaldifferencesamongsurgicalmodalities.Comparedtoabasecaseassuming a4-yearincrementbetweenthesurgeryandradiationmodalitiesinthemedian timefromBCFtometastasis,"uncorrected"analysesyieldsurgicalcostsrelatively overstatedby3to11%;QALYsunderstatedby1to11%;lifeexpectancyunderstated by 1%to8.5%;metastasisoverstatedby17to36%;andPCdeathoverstatedby17% to43%.Resultsaregenerallygreaterwithlargercorrectionsandoutcomesdiscountedtonetpresentvalue. CONCLUSIONS: "Withoutcorrection"resultsdonot affectconclusionsaboutrelativecostsforallpatientsorQALYsforlow-riskpatients.TheydoaffectQALYconclusionsforintermediate-andhigher-riskpatients, yieldingerroneousconclusionsontherelativesuperiorityofradiationversussurgicaltreatments.Withoutcorrection,thesameincorrectconclusionswouldbe reachedforsurvival,metastasis,andPCdeath.Literatureandreal-worlddatawere usedtovalidatethecorrections. PUK23 METHODOLOGICALCONSIDERATIONSFORCOST-EFFECTIVENESSANALYSIS OFONABOTULINUMTOXINAINPATIENTSWITHNEUROGENICDETRUSOR OVERACTIVITYGlobeD1,CarlsonJ2,PatelH3,ColaycoD1,HansenRN2,WatanabeJ2,SullivanSD2 1AllerganLLC,Irvine,CA,USA,2UniversityofWashington,Seattle,WA,USA,3Allergan,LLC, MarlowBuckinghamshire,UKOBJECTIVES: Selectionofanappropriatemodelingstructureisakeyconsideration ineconomicevaluations.Factorsthatinuencethechoiceofmodelingstructure includedataavailabilityandconsensuswithintheclinicalcommunityregarding clinicallymeaningfuldenitionsoftreatmentresponseanddenedandmeasureablehealthstates.Intheabsenceofanagreedconsensus,selectingamodelstructureischallenging.Weexploredtwomodelstructurestoassessthecost-effectivenessofonabotulinumtoxinAforthetreatmentofurinaryincontinence(UI)dueto neurogenicdetrusoroveractivity(NDO). METHODS: Themeritsandlimitationsofa modelbasedontreatmentresponseversusanabsolutemodelstructurewereconsidered.Intheresponsemodel,healthstatesweredenedbyprecisetrialoutcomes(percentreductioninUIepisodesfrombaseline).Intheabsolutemodel, healthstatesweredenedbycategoriesofUIepisodes/week.Intheabsenceof clinicallymeaningfulcutoffs,weplottedhealth-relatedqualityoflife(HRQoL) scoresversusUIepisodestoderivemeaningfulcutoffsforhealthstatesbasedon HRQoL. RESULTS: Intheresponsemodel,responsewasdenedasa 50%reductioninUIepisodes.Theprimarylimitationwiththisapproachistheheterogeneity withinnon-responder'patients(e.g.patientswitheithera0%ora49%reduction arebothnon-responders').NoclearhealthstatecutoffswereobservedwithrespecttoHRQoLandUIepisodesfortheabsolutemodel.Healthstatesweredened onpercentiledistributionwithinthetrialpopulation:dry,0-25%[1-14UIepisodes/ week], 25%-75%[15-32UIepisodes/week],and 75%[ 32UIepisodes/week].The primarylimitationstothisapproacharetheinabilitytocapturehealthimprovementwithinhealthstatesandthelackofclinicalrelevanceforhealthstatesbased onpercentiles. CONCLUSIONS: Intheabsenceofclinicalconsensus,modelstructureselectionshouldbeakeyconsiderationtocapturethetrueeconomicvalueof atherapy. PUK24 USEOFBIOMARKERSINPROPENSITYSCOREMATCHINGTOMITIGATE CHANNELINGBIASINARETROSPECTIVECOHORTOFESRDPATIENTSWilsonSM,RubinJL DaVitaClinicalResearch,Minneapolis,MN,USAOBJECTIVES: Retrospectiveanalysesareinherentlysubjecttobias.Techniques suchaspropensityscorematching(PSM)canfunctionaspseudo-randomization. ThePSM,however,isoftendonesolelyusingadministrativeclaimsdata.Wetested thehypothesisthatclaimsalonewouldprovideinadequatematchingwhencomparedtoclaimsplusrelevantbiomarkers. METHODS: Weuseddatabasesfroma largedialysisorganizationtoobtaintwocohortsofdialysispatientsprescribed differentdrugtherapieswithinthesameclass.Thecohortswererstmatchedon demographicsandcomorbiditiesonly.Thesamecohortswerethenrematched, addingbaselinebiomarkers(albumin,correctedcalcium,Kt/V,normalizedprotein catabolicrate(nPCR),parathyroidhormone(PTH),phosphorus)tothePSM.We usedgeneralizedmixedmodels(GMM)todetermineiftreatmentwasassociated withlaboutcomesovera16-weekperiod,runseparatelyonbothPSMcohorts. RESULTS: TherstPSMcohortsdidnotdifferonbaselinedemographicandcomorbidvariables,butsignicantdifferencesexistedinfour(nPCR,phosphorus,calcium,PTH)ofthesixbaselinebiomarkersexcludedfromthematch(allp 0.01). WhenbiomarkerswereincludedinthePSM,therewerenosignicantdifferences betweengroupsonanybaselinemeasures.GMManalysisshowedasignicantA78VALUEINHEALTH14(2011)A1–A214

PAGE 87

associationbetweendrugandlaboutcomesforthecohortmatchedondemographicandcomorbidinformationonly(p 0.001).Therewasnosignicantassociationinthecohortmatchedonbiomarkerspluscomorbidanddemographic variables(p 0.414). CONCLUSIONS: PSMusingadministrativeclaimsaloneleft considerablechannelingbiasinthesample,anderroneousconclusionswere drawnfromsubsequentGMManalysis.InclusionofbiomarkersinthePSMremovedthisbiasandyieldednon-signicantGMMresults.Administrativeclaims datawithoutbiomarkersmaynotbesufcientforconductingPSM. POSTERSESSIONII: SELECTEDHEALTHCARETREATMENTSTUDIES MedicalDevice/Diagnostics–ClinicalOutcomesStudies PMD1 IMPACTOFVAGUSNERVESTIMULATION(VNS)THERAPYONCLINICAL OUTCOMESANDCOSTSINMEDICAIDPATIENTSWITHDRUG-RESISTANT EPILEPSYDuhMS1,FaughtE2,GuerinA3,SardaSP1,SamuelsonT1,BunkerMT4,OlinBD4, JacksonSD4 ,HelmersSL2 1AnalysisGroup,Inc.,Boston,MA,USA,2EmoryUniversity,Atlanta,GA,USA,3AnalysisGroup, Ltee.,Montreal,QC,Canada,4Cyberonics,Inc.,Houston,TX,USAOBJECTIVES: ThisstudyevaluatedthebenetsofVNSTherapyonhealthresource utilization(HRU),epilepsy-relatedcomorbidities,andcostsovertimeindrug-resistantepilepsypatientsinaclinicalpracticesetting. METHODS: Anopen-cohort pre-postanalysiswasconductedusingMedicaiddatafromFL,NJ,IA,MO,andKS (January1997-June2009).Patientshad 1neurologistvisitwithepilepsydiagnosis (ICD-9345.xx,780.3,or780.39), 1procedureforVNSTherapyimplantation, 1 anti-epilepticdrug(AED), 6monthsofpre-andpost-VNSTherapycontinuous enrollment.The6-monthsprecedingVNSTherapyimplantationwasdesignatedas Pre-VNSTherapyperiod.Post-VNSTherapyperiodwasdesignatedfromthedateof VNSTherapyimplantationuntilremovalofdevice,death,orMedicaiddisenrollment.UnivariateandmultivariatePoissonregressions(incidencerateratios[IRR] and95%condenceintervals[CI])comparedHRUandepilepsy-relatedcomorbiditiesandGLMmodelsestimatedcostdifferencesbetweenthepre-andpost-VNS Therapyperiods.Costswereadjustedto$2009usingmedicalcarecomponentof theconsumerpriceindex. RESULTS: Ofthe1655patients,meanagewas29.4years and51.4%weremales.ComparedtoPre-VNSTherapyperiodtheincidenceof hospitalizationsandemergencyroom(ER)visitsdecreasedovertimeduringPostVNSTherapyperiod(Trend:0.9238,p .0001andTrend:0.9526,p .0001,respectively).Generalizedtonic-clonicseizureeventsshowedadecreasingtrendduring Post-VNSTherapyperiodthanPre-VNSTherapyperiod(adjustedIRR:1.23[CI: 1.03-1.46]inQ1to0.17[CI:0.11-0.27]inQ12).DuringPost-VNSTherapyperiod, averagetotalhealthcarecostsdecreasedfrom$42,540(Q1)to$14,316(Q12).After 1.5years,costsavingsinPost-VNSTherapyperiodstartedtooutweightotalcostsof VNSTherapydeviceandimplantationprocedures. CONCLUSIONS: Toconclude, VNSTherapyisassociatedwithdecreasedHRUandcommonepilepsy-relatedcomorbidities,suchashospitalizations,ERvisitsandgeneralizedtonic-clonicseizure events,resultinginnetcostsavingsforpublicpayersafterabout1.5years. PMD2 AMETA-ANALYSISOFBIOMARKERSANDDIAGNOSTICIMAGINGIN ALZHEIMER’SDISEASESullivanSD1,BloudekL2,SpackmanDE3,BlankenburgM4 1UniversityofWashington,Seattle,WA,USA,2Allergan,Irvine,CA,USA,3YorkUniversity, York,NONE,UK,4Bayer,Berlin,GermanyOBJECTIVES: MildAlzheimer'sDisease(AD)isoftendifculttodifferentiatefrommild cognitiveimpairment(MCI)ornon-ADdementias.Amultitudeofdiagnosticbiomarkersandadvancedimagingstrategieshavebeendevelopedtoaidinthediagnosisand managementofAD.Examplesofbiomarkersincludetotaltau,phosphorylatedtau (Ptau),andthe42aminoacidformofbeta-amyloid(A1-42)andimagingstrategies includeMRI,FDG-PET,SPECTandCT.WesoughttosystematicallyreviewandmetaanalyzethepublishedevidenceonkeytestcharacteristicsofmajordiagnosticstrategiesinAD. METHODS: Asystematicreviewwasundertakentolocateallstudiesof biomarkersordiagnosticimagingforADpublishedinEnglishfromJanuary1990to March2010.Meta-analysiswasperformedusingabivariatemixed-effectsbinaryregressionmodel.Wecalculatedsensitivity(SN),specicity(SP)andareaunderthe receiveroperatingcurves(AUROC),withcondenceandpredictioncontoursinorder toformulatebestestimatesofSNandSPofcommonlyuseddiagnostictestsforAD. RESULTS: Of1,840uniquestudiesidentied,119presentedprimarydatasufcientfor analysis.Foreachdiagnosticmethod,SNandSPwerecalculatedagainstnon-dementedcontrols,non-ADdementiaswithandwithoutMCI,ifavailable.Comparedto non-dementedcontrols,FDG-PETdemonstratedthehighestAUROC(0.96),with90% SN(95%CI84%to94%)and89%SP(95%CI81%to94%).FDG-PETalsowasmostaccurateindiscriminatingADfromdementedcontrols(includingMCI)withAUROC0.91, and92%SN(95%CI84%to96%)and78%SP(95%CI69%to85%).Fordiscriminationof ADfromnon-ADdementias(excludingMCI),CSFPtauandSPECTproducedidentical AUROC(0.86)with79%and86%SNand80%and81%SP,respectively. CONCLUSIONS: DiagnosticstrategiesforADshowwidevariationintestcharacteristicsandsome showpromiseforclinicalpractice. PMD3 PARTIALRESPONDERSINSCHIZOPHRENIAMillierA1,AzorinJM2,AngermeyerMC3,ToumiM4 1CreativCeutical,Paris,France,2SHUdePsychiatried'Adultes,MARSEILLES,France,3University ofLeipzig,Germany,4UniversityClaudeBernardLyon1,Lyon,FranceOBJECTIVES: Despitewellconductedtreatment,schizophrenicpatientsoftenremainsymptomatic.Althoughmultiplestudiesstudiedresistantpatients,nodata areavailableonpartialresponders(PR).Thepurposeofthisstudyistocompare clinical,qualityoflife(QoL),costofPRversusnonsymptomaticpatients(full responders FR),andtoassesstheevolutionofPR. METHODS: Ananalysiswas performedbasedonasampleofpatientsinFranceandUKfollowedprospectively over2years.Atbaseline,resistantaswellasacuteexacerbatedpatientswere excluded.Symptomaticpatientsweredenedaccordingto3clinicalcriteria:CGI score 4andPANSSscore 60(with3items 4).Evolutionwasassessedoverthe 2yearsperiod.Thefollowinginformationwascollectedusingstandardizedvalidatedinstruments:psychoticsymptoms,depression,sideeffects,compliance, functioning,specicandgenericQoL,resourceutilisation. RESULTS: 93(19%)patientswereidentiedasPRand284(59%)asFRoutof484patients.Thepopulation ofPRwaswelldiscriminatedatbaselineperiodwhencomparedtoFR.Clinical symptomsincludingdepressionandsideeffects,compliance,functioning,QoLare moresevereinPRthaninFR.CostofmanagementofPRappearssignicantly higherthanFR.PatientsidentiedasPRremainedPRoverthe2yearsperiod.The changeofprescriptionincludingswitch,dosageincreaseortherapyaugmentation issimilarbetweenPRandFR. CONCLUSIONS: ThePRisastablepopulationovertimewithhigherclinicalburdenandincreasedmanagementcostcomparedtoFR. Thelackofspecictreatmentpatternraisestheissueoftheneedforspecic diseasemanagementstrategyofPRandrelatedassessmentofsuchintervention. PMD4 PERCEPTIONOFIRONDEFICIENCYINCLINICALPRACTICES: MULTIDISCIPLINARYFRENCHSURVEY(SUPFERSURVEY)LasockiS1,LuporsiE2,JaminC3,DarneB4,MahiL5,MarteauP6 1CHUBichat-ClaudeBernard,Paris,France,2CentreAlexisVautrin,Vandoeuvre-Les-Nancy, France,3AFACS,Paris,France,4MonitoringForceSAS,MaisonsLaftte,France,5ViforPharma AG,NeuillysurSeine,France,6CHULariboisire,Paris,FranceOBJECTIVES: Toassesstheperceptionofirondeciencyanddescribethemanagementofirondeciencyinclinicalpractices. METHODS: Physicianswererandomly selectedfromaprofessionaldirectory.Theyreportedinaquestionnairetheestimatedfrequencyofirondeciencyandanemiaamongtheirpatients,thebiological examsperformedforirondeciencydiagnosisandtheconditionsofuseofintravenousiron.Thesurveyanalysiswasperformedon358questionnaires(return rate:12%)fromphysiciansofdifferentareasofexpertise(anesthesia,intensive care,surgery,n 67;gynecology,obstetrics,n 122;oncology,hematology,radiotherapy,n 39;hepato-gastroenterology,internalmedicine,geriatrics,rheumatology,n 130). RESULTS: Outof86%(309/358)ofthephysiciansreportedcasesofiron deciency.For63%ofthem(223/358),thefrequencyofanemiaamongtheirpatientswasmorethan10%.Surveyresponderswere25%(89/358)toreportthatthey systematicallyexploredirondeciencyand61%(217/358)onlyifanemiahadbeen previouslydiagnosed.Severeirondeciency,istreatedwithoralironin75%(269/ 358),39%(141/358)withintravenousironand20%(70/358)withtransfusion.For 70%ofthephysicianswhoprescribeintravenousiron(148/213)thelimitthreshold ofhemoglobinforprescriptionofintravenousironof8g/dL(median).Incontrast withhemoglobinlevel,serumferritinandtransferrinsaturation,wereinfrequently performed(15%[31/213]and5%[11/213],respectively).Theuseoferythropoiesisstimulatingagentswasreportedby44%(156/358)ofphysicianswithsystematic ironsupplementationforonly47%(74/156)ofthem. CONCLUSIONS: Oneofthe mainresultsofthissurveyistheapparentequationbetweenirondeciencyand anemiainmanyphysicians.Thus,mostphysiciansreportthatintravenousironis usedwhenhemoglobinisatamedianvalue 8g/dL,Therefore,irondeciencyis consideredonlywithinacontextofsevereanemiaformanyphysicians. PMD5 OUTCOMESTUDYOFDRUGELUTINGSTENT(DES)VERSUSBAREMETALSTENT (BMS)INHONGKONG:A6-MONTHPILOTSTUDYLeeV1,YanB1,ChanT1,WooA1,LeeKK2,YuCM1 1ChineseUniversityofHongKong,Shatin,HongKong,2MonashUniversityKualaLumpur SunwayCampus,SelangorDarulEhsan,MalaysiaOBJECTIVES: Thispilotstudywastoevaluateandcomparedrugelutingstent(DES) andbaremetalstent(BMS)in1)clinicaloutcome;2)humanisticoutcome;and3) economicoutcome,inpatientsundergoingpercutaneouscoronaryintervention (PCI). METHODS: AllpatientsundergoPCIinPrinceofWalesHospitalduringAugust1toOctober31,2009wererecruited.Clinicaloutcomewasmeasuredbythe occurrenceofmajoradversecardiacevents(MACE),whichisdenedascardiac death,non-fatalMIandtargetlesionrevascularization.AnEQ-5Dquestionnaire wasusedtomeasurethebaselinequalityoflifebeforethestentplacement,andsix monthspostPCI.Proceduralcostincludingtheinstruments,medicationsandhospitalization,aswellasallcardiacrelatedfollow-upfortherstsixmonthswas recorded.CosttoreduceoneMACE(ICER)andcosttogainonequality-adjusted life-year(QALY)wascalculatedtoassessthecost-effectivenessofDES. RESULTS: A totalof50patients(n 22andn 28forDESandBMSrespectively)wereenrolled intoourstudy.MACEin6monthwas5%inDES(n 1)versus7%(n 2)inBMS respectively(p 0.701).InDESgroup,theutilityscoreshowedsignicantimprovementinsixmonthsthanbaseline(0.92,IQR0.801.00vs.0.69,IQR0.220.77, p 0.001),whiletherewerenosignicantimprovementforBMSgroup(0.87,IQR 0.731.00vs.0.81,IQR0.661.00,p 0.337).ICERcalculatedwasHKD$17,112and costperQALYgainedwasHKD$142,600.six-monthtotalmedicalcostwassimilar, HKD$92,440 58,566forDES(median $74231)versus$95,223 64,301forBMS(median $76475)respectively,p 0.875,thehigherproceduralcostofDESgroupwas balancedbylowerfollow-upandhospitalizationcost,whencomparedwithBMS group. CONCLUSIONS: ThequalityoflifeofpatientunderwentplacementwithDESA79VALUEINHEALTH14(2011)A1–A214

PAGE 88

improvedsignicantlyandtheuseofDESwereshowntobecost-effective.The 6-monthtotalmedicalcostforDESandBMSweresimilar. PMD6 READMISSIONRATESANDCOSTSASSOCIATEDWITHFIBRINSEALANTUSE AMONGPATIENTSUNDERGOINGORTHOPEDICSURGERYYeX1,ShahM2,RupnowMF1,HammondJ1 1Ethicon,Inc.,Somerville,NJ,USA,2Xcenda,LLC.,PalmHarbor,FL,USAOBJECTIVES: Payersandhospitaladministratorsareincreasinglyconcernedabout readmissionratesinsurgicalpatients.Wesoughttoexaminethereadmission ratesandhospitalcostsassociatedwithEVICELbrinsealant(all-humanformulation),versusVITAGELbrinsealant(withbovinethrombin),ornoadjuncthemostatuseforpatientsundergoinginpatientjointreplacementsurgeries. METHODS: AretrospectiveanalysiswasconductedusingPremieradministrativedatafrom over500UShospitals.Hospitalizedpatients( 18years)whounderwentorthopedic surgeryandreceivedEVICEL,VITAGELornohemostatduringsurgerybetween January1,2009andNovember30,2009wereidentied.A1:1(EVICEL:VITAGEL)and 1:3(EVICEL:nohemostat)matchwasconductedusingsurgerytypeandpropensity scoresofreceivingEVICEL,basedonpatientandhospitalcharacteristicsviaa logisticregressionmodel.Theoutcomesincluded30-dayall-causereadmission ratesandtotalindexhospitalcosts.Differencesinreadmissionrateswereanalyzedusingconditionallogisticregression.Ageneralizedlinearmodelwithloglink/gammadistributionwasusedforanalyzingdifferencesintotalcosts. RESULTS: Atotalof316patientswereidentied(158percohort)fortheEVICEL versusVITAGELand1,808patientsforEVICEL(n 452)versusnohemostat (n 1,356)analysis.PatientsintheVITAGELcohortwere6.8timesmorelikelytobe readmittedtothehospitalcomparedtotheEVICELcohort(12.7%vs3.8%;OR 6.81, 95%CI1.62,28.66).Patientsinthenohemostatcohortwere1.6timesmorelikelyto bereadmittedcomparedtotheEVICELcohort.Totalindexhospitalcostwaslower fortheEVICELcohort($16,704)comparedtoVITAGELcohort($18,192p 0.001)on average.TheEVICELcohort($17,387)hadsimilartotalcostscomparedtonoadjuncthemostat($17,389)cohort. CONCLUSIONS: Readmissionpresentssignicant costsandhasbeenaddedtohospitalqualitymeasures.Inthisstudy,EVICELwas associatedwithlowerreadmissionratescomparedtoVITAGELornoadjuncthemostatuseininpatientjointreplacementsurgeries. PMD7 DIFFERENCESINOUTCOMESMEASURESOFDIABETESPATIENTSUSINGAN INSULINDEVICEANDACONVENTIONALHUMANINSULINVIAL/SYRINGEBaserO1,WangL1,YuceH2,XieL1,DysingerAH1 1STATinMEDResearch,AnnArbor,MI,USA,2NewYorkCityCollegeof Technology-CUNY/STATinMEDResearch,NewYork,NY,USAOBJECTIVES: Tocomparethemainoutcomesdifferencesincludingclinicalevents, healthcareutilizationandcostsofpatientsusinganinsulindevicefordiabetesversus patientsusingtheconventionalhumaninsulinvial/syringe. METHODS: Usingaretrospectiveanalysisofhealthinsuranceclaimsdatabetweentheyears2003and2008, weidentiedpatientswithadiagnosisofdiabetesandthendividedthemintoan insulindevicecohortandahumaninsulinvial/syringecohort,basedontheirprescriptionlls.Patients'demographics,healthcarevisitsandcostswerecomparedusing Chi-squaretestingandstandardizeddifferences.The12-monthfollow-upclinical eventrates,healthcarefacilityuseandcostsforthosepatientswerecompared.Risk adjustmentwasperformedusingthepropensityscorematchingmethodwiththe ProbChoicealgorithm. RESULTS: Atotalof12,400eligiblepatientswereidentiedas usinginsulinfordiabetes:1,236(9.97%)receivedtheinsulindeviceand11,164(90.03%) receivedtheinsulinvial/syringe.Comparedwithpatientswhoreceivedtheconventionalhumaninsulinvial/syringe,patientsintheinsulindevicegroupwerelikelytobe younger,liveintheMidwestoftheUnitedStates,andhavetypeIdiabetes.Although therewerenosignicantdifferencesinhypoglycemiceventsafterriskadjustment, patientsintheinsulindevicegrouphadsignicantlyfewercasesofcerebrovascular disease(4.14%vs.9.12%p 0.0055),congestiveheartfailure(7.18%vs.12.15% p 0.0267)andchronicobstructivepulmonarydisease(4.70%vs.10.50%p 0.0039),but morecasesofdyslipidemia(68.51%vs.54.42%p 0.0002).Althoughtheoutpatient costsforofcevisits($1888vs.$1895p 0.0257)werelowerforpatientsontheinsulin device,theirprescriptioncosts($5489vs.$4635p 0.0001)werehigher.Theoverall risk-adjustedhealthcarecostsdidnotdiffer($14,231vs.$18,096p 0.1160)between thetwogroups. CONCLUSIONS: Withoutsignicantadditiontothecosts,insulin administrationwiththedeviceisassociatedwithfewerclinicalevents. PMD8 POSITRONEMISSIONTOMOGRAPHYSCREENINGFORLUNGCANCER:A SYSTEMATICREVIEWChienCR,KaoCH,WangHN,LiangJA ChinaMedicalUniversityHospital,Taichung,TaiwanOBJECTIVES: Thereisnoestablishedeffectivelungcancerscreeningmodality. PositronEmissionTomography(PET)ishelpfulinlungcancerdiseaseextentevaluation.TheobjectiveofourstudyistoevaluatetheroleofPETinlungcancer screeningviasystematicreview. METHODS: Usingastrategysimilartoaprevious computedtomography(CT)lungcancerscreeningsystematicreview[Blacketal. Thorax2007;62:131138],wesearchedforprimarystudiesfocusingonPETscreeningforlungcancerusingthefollowingkeywords"(lungcancer)AND(positron emissiontomography)AND((screen)OR(screening))"inPubmed¨onNov30th, 2010.Tworeviewers(ChienC.R.&WangH.N)reviewedalltheidentiedstudies independentlytondoutstudiescompatiblewithourinclusion/exclusioncriteria. Furtherdiscussionwith3rdreviewer(KaoC.H.)wastakentoreachconclusion whentherewasanydisagreementamongthereviewers.Manualsearchingfor relevantstudieswasalsoperformedfromtheincludedstudies.Werestrictedour analysistonon-overlappedstudiespublishedsince2000andinEnglish. RESULTS: Amongtheidentiedstudies(n 2733),239studieswerepublishedbefore2000, 2440studieswereexcludedduetoirrelevanttitlesandkeywords,andanother34 studieswereexcludedafterreviewingtheabstracts.Fullpaperevaluationledto furtherexclusionof11studies,andmanualsearchledtoinclusionof2additional studies,leaving11studiesforanalysis.Nostudiesevaluatedtheefcacyofprimary PETscreeningspecicforlungcancer.EightstudiesfocusedonprimaryPET screeningforcancer,andthreestudiesreportedndinginlungcancerCTscreeningprogramswithselectivePET. CONCLUSIONS: TheroleofprimaryPETscreening forlungcancerremainsunknown.PEThasthepotentialtobeusedasascreening modalitynotspecicforlungcancerandasaselectivemodalityincombination withCTforlungcancerscreening.[1]Blacketal.Thorax2007;62:131-138 PMD9 ESTIMATIONOFQUALITY-ADJUSTEDLIFEEXPECTANCYANDLOSSOF QUALITY-ADJUSTEDLIFEEXPECTANCYINPATIENTSUNDERPROLONGED MECHANICALVENTILATION:APOPULATION-BASEDSTUDYDURING19982007INTAIWANHungMC1,WangJD2 1NationalTaiwanUniversity,Taipei,Taiwan,2NationalChengKungUniversityCollegeof Medicine,Tainan,TaiwanOBJECTIVES: Thequality-adjustedlifeexpectancy(QALE)andlossofquality-adjusted lifeexpectancy(lossofQALE)inpatientsunderprolongedmechanicalventilation (PMV)stratiedbydifferentunderlyingdiseasesweredetermined. METHODS: Asimplerandomsampleofall171,635patientswhowereperformedcontinualmechanical ventilationformorethan21daysduringthe1997-2007periodsinTaiwanleftus50,481 subjects.Afterstratifyingthepatientsaccordingtospecicdiagnoses,weperformed latentclassanalysis(LCA)tocategorizePMVpatientswithmultiplecomorbiditiesinto severalclusteredgroups.Thesurvivalfunctionswereestimatedforeachgroupwith Kaplan-Meiermethodandextrapolatedto300monthstoobtainthelifeexpectancies throughasemi-parametricmethod.TheresultswereadjustedwithautilitymeasurementofqualityoflifetoestimatetheQALE(quality-adjustedlifeexpectancies).Further,wecomparedtheage-,gender-matchedreferencepopulationstocalculatethe lossofQALE. RESULTS: TheQALEofPMVpatientswithchronicrenalfailurewere0.42 and0.19quality-adjustedlifeyears(QALY)forconsciousnessclearversusunclear states,respectively;thoseofpatientswithcancerwere0.48and0.22,respectively; thoseofpatientswithParkinson'sdiseasewere0.62and0.27,respectively;thoseof patientswithlivercirrhosiswere0.98and0.43respectively;thoseofpatientswith strokewere1.03and0.46respectively;thoseofpatientswithdegenerativeneurological diseaseswere1.47and0.64respectively;thoseofpatientswithinjuriesandpoisoning were1.81and0.78respectively.TheLCAclassiedcaseswithmultiplecomorbidities intoseveralcategories,ofwhichtherewasaconsistenttrendofdecreaseinQALEand lossofQALEaspeoplegrowold. CONCLUSIONS: Theresultscouldhopefullyreduce thegapbetweenpatients'familiesandhealthcareprovidersandassisttheclinicaland healthpolicydecisions. PMD10 SCREENINGTREATMENTANDCONTROLOFHYPERTENSIONINDIABETIC PATIENTSUSINGOUTPATIENTVISITDATASurbhiS,MLMackey St.John'sUniversity,Queens,NY,USAOBJECTIVES: Bloodpressurecontrolisagreatchallengeinthediabeticpatient populationsincethebloodpressuretargetislower, 130/80,ascomparedto 140/90ingeneralpopulation.Theobjectiveofthisstudywastoexaminetherate andtheassociationofpatientcharacteristics(demographic,accesstohealthcare andclinicalfactors)andpracticecharacteristicswithhypertensionscreening, treatmentandcontrolindiabeticpopulation. METHODS: NationalAmbulatory MedicalCareSurveyandtheNationalHospitalAmbulatoryMedicalCareSurvey 2006wereusedtoanalyzeoutpatientvisitsmadebyadults18yearsandolder diagnosedwithdiabetes.Descriptiveanalysiswasdonetondtherateandbinary logisticregressionwascarriedouttondthepredictors.Statisticalsignicance wassetatalphaof0.05. RESULTS: Hypertensionscreening,treatmentandcontrol ratewas66.9%,53.1%and28.4%indiabeticpatients.Theoddsofnotgetting screenedwerevisitsotherthanprimarycarephysician(OR7.52),withnodiagnostic tests(OR6.63),havingnocomorbidities(OR3.64),nonobese(OR1.72)andincreasing age(OR2.03,OR2.35,OR2.72).Oddsofnotbeingtreatedweresettingslocatedin southgeographicregion(OR1.29),providerotherthanprimarycarephysician (OR2.02),hospitalsetting(OR1.28),nodiagnostictests(OR1.97)andhavingnoco morbidities(OR1.558).Oddsofnothavingbloodpressurecontrolweregreaterfor blackrace(OR1.75),patientswithnopastvisits(OR1.79),obese(OR1.37)and havingnocomorbidities(OR1.40). CONCLUSIONS: Thestudyfoundthatalthough morethan50%ofthediabeticpatientswerescreenedandtreated,bloodpressure controlwasfoundinonlyonethirdofthepopulation.Boththepatientfactors; demographic,accesstohealthcare,clinicalfactorsandpracticecharacteristics wereresponsibleforpoorqualityofcare(hypertensionscreeningandtreatment) andpooroutcome(bloodpressurecontrol). MedicalDevice/Diagnostics–CostStudies PMD11 IDENTIFYINGPOTENTIALDRIVERSOFCOSTSAVINGSWITHINSULIN ADMINISTRATIONDEVICESINTYPE-2DIABETESINTHEUNITEDSTATESPollockRF1,CurtisB2,BoyeKS2,TimlinL3,ValentineWJ1 1OssianHealthEconomicsandCommunications,Basel,Switzerland,2EliLillyandCompany, Indianapolis,IN,USA,3LillyUK,Windlesham,Surrey,UKA80VALUEINHEALTH14(2011)A1–A214

PAGE 89

OBJECTIVES: Toinvestigatepotentialdriversofcostsandcostsavingswhenassessingthebudgetimpactofinsulinadministrationdevices(IADs)inpatientswith type2diabetes(T2D). METHODS: AMicrosoftExcel¨-basedmodelwasconstructed toassessthebudgetimpactofnewreusableIADsinpatientswithT2D.Themodel capturedcostsofinsulin,IADs,needles,self-monitoringofbloodglucose(test stripsandglucometers)andthedirectmedicalcostsofanumberofdiabetescomplications(minorandmajorhypoglycemia,myocardialinfarction,stroke,endstagerenaldisease,blindnessandamputation).Scenarioswereconstructedto assessthebudgetimpactofchangesinhypoglycemiarates,complianceandinsulinwastageassociatedwithanewIAD.Theanalyseswereperformedinaone millionmemberUShealthcareplanfromthepayerperspective.Futurecostswere discountedat3%perannumandsensitivityanalyseswereperformed. RESULTS: In amillionmemberplanwithadiabetesprevalenceof7.8%andannualdiabetes incidenceof0.52%,ahypotheticalIADcapableofreducinghypoglycemiaratesor insulinwastagewasprojectedtodecreasehealthcareexpenditureoveroneand ve-yeartimehorizons.Conversely,ahypotheticalIADcapableofimprovingcomplianceresultedinincreasedhealthcareexpenditure,drivenbyincreasedinsulin andneedlecosts.Adevicethatconferredallofthesebenetssimultaneouslywas foundtobecostsavinginthetargetpopulation,resultingina0.62%decreasein diabetes-relatedhealthcareexpenditureoverveyears. CONCLUSIONS: AnIAD capableofreducinghypoglycemiceventratesandinsulinwastagewhilstimprovingcompliancewouldlikelybecostsavingoverall. PMD12 GENDERDIFFERENCESINTOTALKNEEARTHROPLASTY(TKA)POSTOPERATIVE PAINMANAGEMENTINTHEUNITEDSTATESGraverRM1,DomyahnM1,ReadCE1,MenzieAM1,FeinglassSR2 1Zimmer,Minneapolis,MN,USA,2Zimmer,Warsaw,IN,USAOBJECTIVES: TheobjectiveofthisanalysiswastohighlightthedifferencesingenderpainmanagementafteratotalkneearthroplastyintheUnitedStates. METHODS: Weidentied103,600TKAproceduresfromtheThomsonReutersMarketScan¨CommercialandMedicareClaimdatabasesfromJanuary1,2003through June30,2009.Fortheseprocedures,weanalyzed873,237pharmaceuticalclaimsfor musclerelaxantsandanalgesics/antipyreticsforthe12-monthpostoperativeperiod.Themeandirectcostswerecalculatedforeachcalendaryearforgender cohortsandnotination-adjusted.Wilcoxon-Mann-Whitneytestswererunfor eachyeartodeterminestatisticalsignicanceamongthegendercohorts. RESULTS: Duringtheanalysisperiod,wefoundstatisticallysignicantdifferences inpharmaceuticalpainmanagementspendingbetweenmaleandfemalecohorts. Thefemalecohortspendingaveraged$464perpatientforthe12-monthpostoperativeperiodcomparedto$364forthemalecohort.Thisrepresentsa27%difference.Duringthissameperiod,theaveragenumberofpharmaceuticalpainmanagementclaimsperTKAdecreasedforallcohortsbuttheseaveragesexhibited convergencein2008asspendingbymalesincreasedslightly. CONCLUSIONS: Publishedstudieshavefailedtoagreeonthecorrelationbetweengenderandpostoperativepain.Arecentpostoperativepainstudyconcludedthat"genderwasnot foundtobeaconsistentpredictorastraditionallybelieved."However,ourretrospectiveanalysisofactualclaimdataprovidesastatisticallysignicantcorrelation betweenfemalegenderandincreasedconsumptionofpostoperativemusclerelaxantsandanalgesicsfortotalkneearthroplastyprocedures.Thesedatamayhelp surgeonsprovideappropriatepostoperativecarefortheirTKApatients.1Ip,Hui YunVivian,Et.al,PredictorsofPostoperativePainandAnalgesicConsumption:A QualitativeSystematicReview,Anesthesiology,September2009:11(3)657-677.MarketScan¨isatrademarkofThomsonReuters(Healthcare)Inc. PMD13 POSTINGUINALHERNIAREPAIRPAINMANAGEMENTCOSTS:ASTUDYUSING THEPREMIERPERSPECTIVE™DATABASE(PPD)MorseonM,HashemiL Covidien,Franklin,MA,USAOBJECTIVES: Severalstudieshavereportedontheabilitytoreducepost-operative painfollowingInguinalHerniarepairbyself-xatingmesh.Ourstudyexamines theoverallcostincreaseswhenpostoperativepainfollowingInguinalHerniaRepairissignicantenoughtobereportedbyanICD-9diagnosiscode. METHODS: ThePremierPerspectiveDatabase(PPD)wasusedtotrackhospitalcostsassociatedwithanoutpatientInguinalherniarepairbetweenJanuary,2008throughJune, 2010.ICD-9diagnosiscodeswereusedtofurtheridentifypatientwiththepresence ofacuteorchronicpostoperativepain.Forpostdischargepainmanagementstrategiesandcostswesponsoredmarketresearchwith30surgeonswhoperform InguinalHerniarepairsurgery.Responseswerecollectedtospecicpainmanagementstrategiesemployed,andaverageperpatientcostswhenpatientreported painwasmild,moderate,andsevere. RESULTS: Atotalof51,108patientsundergoingoutpatientInguinalHerniarepairwereidentiedinthePPDduringourstudy timeframe.AnICD-9diagnosticcoderepresentingacuteorchronicpainwaspresentin228discharges.Themeancostperdischargeforpatientswithdiagnosisof acute/chronicpainwas$3,309,comparedto$2,910forpatientswithoutthediagnosticofpain.Eightypercentofthephysicianssurveyedreportedtheywould prescribepaintherapywithacostoflessthan$100.00.Formoderatepainthe breakdownwas70%lessthan$100.00andforseverepain44%indicatedthecosts wouldbelessthan$100.00. CONCLUSIONS: Themanagementofpainfollowing InguinalHerniarepairincreasethepre-andpost-dischargecoststothehospital, insurersandpatients.Inaddition,costsappeartoincreaseinrelationtotheseverityofthepainreported.Inguinalherniarepairtechniqueswhichresultinreduced postoperativepainsuchasusingselfxatingmeshhavethepotentialtoreduce thecostsassociatedwithInguinalHerniarepair. PMD14 COMPARISONOFPERDISCHARGECOSTSOFTHEREPAIROF INCISIONAL/VENTRALHERNIASPERFORMEDWITHPERMACOL™AND STRATTICE™INTHEUNITEDSTATESMorseonM,HashemiL Covidien,Franklin,MA,USAOBJECTIVES: BiologicgraftimplantsareusedintherepairofcomplexIncisional/ Ventralherniaswheninthesurgeon'sjudgmentthecomplexityofthehernia and/orthepatient'sriskfactorsindicateabiologicmaterialprovidesanopportunityforbetteroutcomesthanrepairperformedwithsyntheticmeshmaterials.Our objectivewastoexaminetheaveragecostperdischargeofIncisional/VentralherniarepairwhenPermacolandStratticebiologicalmesheswereused. METHODS: ThePremierPerspectiveDatabase(PPD)wasusedtotrackhospital reportedcostsassociatedwithpatientstaysforIncisional/Ventralherniarepair utilizingPermacolandStratticeinthe27monthsfromApril,2008through June,2010.PPDisthelargesthospital-based,service-levelcomparativedatabasein theUSAprovidingdetailedresourceutilizationandcostdatacategorizedundera patients'principalandsecondaryprocedurecodes.ICD-9procedurecodeswere usedtoidentifypatientswithIncisional/Ventralherniarepair. RESULTS: Atotalof 36,105patientswithaprimaryprocedurecodeforIncisional/Ventralherniarepair wereidentiedbetweenApril,2008throughJune,2010.Biologicswereusedin4.2% ofthesepatients.Permacolwasused135patientsandStratticewasusedin287 patients.Therewerenosignicantdifferencesinpatient'sdemographicsbetween thetwogroups.MeanandmediancostperdischargewaslowerforPermacol casescomparedtoStrattice($19,221vs.$22,428formeanand$3,163vs.$5,102for median).Costof"CentralSupply"accountedforapproximately50%oftheoverall costinbothgroups,while"RoomandBoard"accountedforapproximately22%of theoverallcost. CONCLUSIONS: Themeancostofthebiologicimplantwasreportedas42%higherwhenStratticewasusedvswhenPermacolwasused ($6,596vs.$4,659)inIncisional/Ventralherniarepairs.Themeanoveralldischarge costswas17%higherwhenStratticewasusedvsPermacol($22,428vs. $19,221). PMD15 ECONOMICEVALUATIONOFSELF-MONITORINGOFBLOODGLUCOSEREGIMES PLUSCONVENTIONALPHARMACOLOGICTREATMENTFORTYPE-2DIABETIC PATIENTSINMEXICO:ESTIMATIONBYDISCRETEEVENTSIMULATIONZanelaOO1,CabraHA1,Mu–ozDF2 1Johnson&JohnsonMedical,MexicoCity,D.F.,Mexico,2InstitutoTecnol—gicoAut—nomode MŽxico,MexicoCity,D.F.,MexicoOBJECTIVES: Estimatetheyearlycostsattributabletoandresultingfromtheprovisionofdifferentself-monitoringofbloodglucose(SMBG)regimes vs. noSMBGin thetreatmentofpatientswithtype-2diabetes(T2D)fromtheMexicanpublic healthsystemperspective. METHODS: Theindividualexperienceofapatientwith T2DwassimulatedusingadiscreteeventsimulationbuiltinArena.Patients werecreatedwithunique,randomlyassignedclinical,epidemiologicanddemographicbaselinecharacteristics,cloned3times,andsenttoeachoftheSMBG regimesconsidered(0,1,2and3timesdaily).Clinicalguidelines(ADA,ALAD)were usedtodeterminediabetic-andcomplication-specicpharmacologictreatment, resourceutilizationandtreatmentalgorithms&goals.Glycosilatedhemoglobin (HbA1c)wasthemaindriverofdiseaseprogression,determininginitialstate,drug titulation/combinationsandinsulindosage.Treatmenttherapiesincludelifestyle modicationsalone,oralantidiabetics(OADs)andinsulinuse.Complicationand acuteeventdevelopmentforeachSMBGregimewasassessedthroughrelative risks,takenfromlocalpublishedstudies.AllOADsandinsulinswereassumedto beequallyeffectiveinreducingHbA1c.T2D-relatedcomplicationsandmortality wereconsidered.Allclinicalandcostdatawereobtainedfrompublishedliterature. Simulationwasrunwith16,000patientsfor20yearsusinga4.5%annualdiscount rate.Per-patientcostsforselectedyearsareshowninination-adjusted2010MXP. RESULTS: 1,2and3timesdailySMBGregimesresultedinlessercoststhanno SMBGafteryears1,3and4,respectively.Year5accumulatedcostsfortheformer were$114,099,$108,152and$110,898,and$120,723fornoSMBG. CONCLUSIONS: SMBGadoptionyieldsbettermid-andlong-runclinicalandeconomicoutcomes thannoSMBG.Potentialsavingsareduetofewercomplicationsandslowerdisease progression.HealthcareinstitutionsshouldadoptSMBGtocontrolHbA1clevels andreducethesocialandeconomicburdenT2Dimposes. PMD16 COMPARISONOFCOSTANDCLINICALOUTCOMESOFOPENVERSUS LAPAROSCOPICTHORACICPROCEDURESINTHEUNITEDSTATESHashemiL1,MinshallME2 1Covidien,Franklin,MA,USA,2Covidien,Manseld,MA,USAOBJECTIVES: Laparoscopicthoracicprocedureshaveoftenbeenassociatedwith shorterhospitalizations.However,theeconomicimpactoflaparoscopicthoracic procedurescomparedtoopenthoracicprocedureshavenotbeenfullyassessed.In thisstudy,weexaminedthedifferencesincostandhospitallengthofstayforopen versuslaparoscopicthoracicproceduresintheUnitedStates. METHODS: ThePremierPerspectiveDatabase(PPD)wasusedtoestimatetheincidenceandcostsof laparoscopicversusopenthoracicproceduresintheUnitedStates.PPDisthelargesthospital-baseddatabaseintheUSAprovidingdetailedresourceutilizationand costdata.PatientswithprincipalprocedurecodesforopenorlaparoscopicproceduresbetweenJanuary1,2007throughDecember31,2009wereselected.CombinationsofICD-9diagnosiscodesandCPTprocedurecodeswereusedtoidentify surgicalsiteinfections,hemorrhageandbloodtransfusions. RESULTS: Atotalof 22,640patientswithaprimaryprocedurecodeforthoracicprocedureswereiden-A81VALUEINHEALTH14(2011)A1–A214

PAGE 90

tiedbetweenJanuary1,2007andDecember31,2009.Thesurgicalapproachwas laparoscopicin57.3%patients,andopenin42.7%patients.Meancostperdischarge wassignicantlyhigherinopenthoracicprocedurescomparedtolaparoscopic procedures,$24,995vs.$19,238,respectively(p 0.001).Patientsundergoinglaparoscopicthoracicprocedureshadasignicantlylowerrateofsurgicalsiteinfections comparedtopatientswhounderwentopenprocedures(4.8%vs.5.8%,respectively, p 0.001).Therewasasignicantlyhigherrateofbloodtransfusionswithpatients undergoingopensurgerycomparedtopatientsundergoinglaparoscopicprocedures(13.2%vs.6.3%,respectively,p 0.001). CONCLUSIONS: Laparoscopicthoracicprocedureswereassociatedwithshorterhospitallengthsofstay,lowerrateof surgicalsiteinfections,hemorrhage,bloodtransfusionandmortalityrates.The meancostsforlaparoscopicproceduresweresignicantlylowerthanmeancosts foropenprocedures.Theseobservationshighlightthepotentialcostadvantagesof providingthoracicproceduresthroughlaparoscopictechniquesasamethodto potentiallysaveincreasinglyscarcehealthcarefundsforhospitals. PMD17 ANALYSISOFCOSTDRIVERSINSTRUCTUREDSMBGINPOORLYCONTROLLED, NON-INSULINTREATEDTYPE-2DIABETES:RESULTSFROMTHESTEPSTUDYMyersJ1,BerndtK2,WegmannN1,ReesC1,MastO2,WagnerR1 1RocheDiagnosticsOperations,Inc.,Indianapolis,IN,USA,2RocheDiagnosticsGmbH, Mannheim,GermanyOBJECTIVES: Analyzethedifferentialinselecteddirectcostsofacollaborative structuredbloodglucosetestinginterventioninnon-insulintreatedpatientswith type2diabetesmellitus(T2DM)whencomparedtoenhancedusualcare(active controlgroup(ACG)). METHODS: DatawasderivedfromtheStructuredTesting Program(STeP )-a1 year,prospective,cluster-randomized,multicenterstudythat examinedtheutilityofacollaborativeinterventionusingstructuredself-monitoringofbloodglucose(SMBG)in483poorly-controlled(HbA1c 7.5%)T2DMsubjects comparedtotheACG.Thestructuredtestinggroup(STG)usedtheACCU-CHEK¨ 360¢aView3-dayproletoolthatfacilitatescollectionandinterpretationof7-point glucoseproles.FromaUSpayerperspective,directcostsofdiabetesmedications, labHbA1ctests,physicianvisits,andbloodglucosetestingstripsassociatedwith STGwerecomparedwithACGusingstudentt-testatasignicancelevelof5%. RESULTS: Intheintent-to-treatpopulation,STGshowedasignicantlygreater HbA1creductionover12monthsthantheACG(-1.2%vs. 0.9%; -0.3%;p 0.04). Duringthestudy,STGincurred $180.95meanPPPY(PayperPatientYear)total costfordiabetesmedications,but-$173.73meanPPPYforSMBGteststrips,-$5.20 meanPPPYforlabHbA1ctests,and-$2.15meanPPPYforphysicianvisitscompared toACG.TherewasnosignicantdifferenceindirectcostsbetweenSTGandACG (p 0.9898). CONCLUSIONS: Useofacollaborativestructuredtestingintervention improvedHbA1cinSTGwithoutincreasingdirectcost.TheincreasedSTGmedicationcostwasoffsetbyadecreaseduseofbloodglucoseteststrips.Aspreviously reported,STGsubjectsperformedsignicantlyfewertests/daythanACGsubjects (mean 0.9vs.1.2,p 0.0003)overtheyear.Structuredtesting,fro ma1yearUS payerperspective,isaneffectiveandoverallcost-neutraltoolformanagementof non-insulintreatedpatientswithtype2diabetes. PMD18 THECOST-EFFECTIVENESSOFINITIATINGSENSOR-AUGMENTEDPUMP THERAPYVERSUSMULTIPLEDAILYINJECTIONSOFINSULININADULTSWITH TYPE1DIABETES:EVALUATINGATECHNOLOGYINEVOLUTIONKambleS1,PerryBM2,SharoffJ2,SchulmanKA1,ReedSD1 1DukeClinicalResearchInstitute,Durham,NC,USA,2MedtronicDiabetes,Northridge,CA,USAOBJECTIVES: Sensor-augmentedpumptherapy(SAPT)demonstratedasignicant reductioninHbA1clevelsintype1diabetespatientscomparedtomultipledaily injectionsofinsulin(MDI)inarecentrandomizedtrial.Weanalyzedthedataon medicalresourceusecollectedwithinthetrialandevaluatedthelong-termcosteffectivenessofSAPTversusMDIinadultswithtype1diabetesfromthehealthcare systemperspective. METHODS: Wecombinedestimatesderivedfromthetrialdata andthemedicalliteraturetopopulatethepreviously-validatedCOREDiabetes Model,whichincludesaseriesofMarkovconstructsthatsimulatetheprogression ofdiabetes-relatedcomplications.Duringthe52-weektrialperiod,SAPTpatients wereprovidedinsulinpumpsand3-daysensors.Electronicrecordsindicatedthat sensorswerewornonlybypatients65%ofthetime.Theincrementalreductionin meanglycatedhemoglobinwas0.6percentagepointsinpatientsrandomizedto SAPTrelativetoMDI. RESULTS: Amongthe329adults(19-70years),meanagewas 41yearsandmeandurationofdiabeteswas20years.Totaltreatmentcostsoverthe 52-weekfollow-upperiodwereestimatedat$10,760forSAPTpatientsand$5,072 forMDIpatients(2010US$).Discounted(3%peryear)lifetimeestimatesofdirect medicalcostsandQALYswere$253,493and10.794forSAPTpatientsand$167,170 and10.418forMDIpatients.ThecorrespondingICERwas$229,675perQALY(95% CI:139,071to720,865).Sensitivityanalysesrevealedthatevolvingtechnologies couldimprovethecost-effectivenessofSAPT.Witha6-daysensor,theICERdecreasedto$168,104perQALY.Upondevelopmentofa6-daysensorthatrequires onlyoneteststripperreplacementforcalibration,theICERwoulddropto $72,417/QALY. CONCLUSIONS: Ourbase-casendingsrevealedthatthecurrent SAPTtechnologyisnoteconomicallyattractive.However,withtechnologicaladvancescurrentlyinprocess,thecost-effectivenessoftheSAPTcouldsignicantly improve. PMD19 COST-EFFECTIVENESSOFEPIDERMALGROWTHFACTORRECEPTORGENE MUTATIONTESTINGINTHESELECTIONOFFIRST-LINETHERAPYFOR PATIENTSWITHADVANCEDNON-SMALLCELLLUNGCANCERINONTARIOChenW1,EllisP2,LevinL3,KrahnM4 1UniversityofToronto,Toronto,ON,Canada,2JuravinskiCancerCentre,Hamilton,ON,Canada,3OntarioMinistryofHealthandLong-TermCare,Toronto,ON,Canada,4TorontoHealth EconomicsandTechnologyAssessment(THETA)Collaborative,Toronto,ON,CanadaOBJECTIVES: Toassessthecost-effectivenessofepidermalgrowthfactorreceptor (EGFR)genemutationtestingforguidingtheapplicationofgetinibasrst-line therapyinpatientswithadvancednon-smallcelllungcancer(NSCLC)livingin Ontario. METHODS: AdecisionanalyticmodelwasdevelopedtocompareEGFR genemutationtestingstrategyversusnotestingstrategyinpatientswithadvancedNSCLC.Underthetestingstrategy,patientstestedpositiveformutation wouldreceivegetinibasrst-linetherapy.Undernotestingstrategy,patients wouldreceiveconventionalchemotherapyasrst-linetherapy.Probabilityvariableswereestimatedthroughliteraturereview.Utilityvariableswereestimated fromamultivariatelinearregressionanalysistakingintoaccountoftheclinical responsesandside-effectsassociatedwithtreatmentforNSCLC.Costvariables werebasedontwoOntariocoststudiesforNSCLC.Bothbenetsandcostswere discountedat5%perannum. RESULTS: Comparedtonotestingstrategy,theincrementalcost-effectivenessratioforEGFRgenemutationtestingwas$46,021per lifeyearor$81,071perqualityadjustedlifeyear(QALY).Thecost-effectivenessof EGFRgenemutationtestingwassensitivetothecostandefcacyofgetinib.The budgetimpactanalysisprojectedthatEGFRgenemutationtestingwouldcostOntariohealthcaresystem$4.6M,$7.0M,$7.9M,$8.1M,and$8.1Mmoreayearinthe nextveyears. CONCLUSIONS: EGFRgenemutationtestingwouldnotbecosteffectiveinpatientswithadvancedNSCLCinOntariountilwillingness-to-paywas above$81,000perQALY.Theefcacyandcostofgetinibsignicantlyaffectedthe cost-effectivenessofEGFRgenemutationtesting. PMD20 COST-EFFECTIVENESSANALYSISOFANEWINDEXFORPROSTATECANCER DETECTIONNicholMB1,WuJ1,AnJJ1,HuangJT2,DenhamD2,FrencherSK3,JacobsenSJ4 1UniversityofSouthernCalifornia,LosAngeles,CA,USA,2BeckmanCoulter,Inc.,Brea,CA,USA,3UCLA,RANDInstitute,LosAngeles,CA,USA,4KaiserPermanenteSouthernCalifornia, Pasadena,CA,USAOBJECTIVES: Anewprostatecancerdetectionindexwasdevelopedasacombinationofserumprostate-specicantigen(PSA),freePSA,andaPSAprecursorform [-2]proPSAtocalculatetheprobabilityofprostatecancerandusedasadiagnostic aidformenage50yearsorolderwithPSA4-10ng/mLandnonsuspiciousdigital rectalexam.TheindexhasdemonstratedimprovedspecicityfordetectingprostatecanceroverPSAtestaloneinclinicaltrials.Thecurrentstudyevaluatedthe cost-effectivenessofearlyprostatecancerdetectionwiththeindexaddingtoPSA comparedwithPSAtestalonefromtheU.S.societalperspective. METHODS: A Markovmodelwithprobabilisticsensitivityanalysiswasconstructedtoestimate thecostsandhealthstateutilitiesofprostatecancerdetectionandconsequent treatmentfortheannualprostatecancerscreeningofthemalefromage50through 75years.Thetransitionprobabilities,healthstateutilities,andprostatecancer treatmentcostswerederivedfromthepublishedliterature.Thediagnosticperformanceoftheindexwasobtainedfromamulti-centersimulationstudyofthe index.Diagnosticrelatedcostswereobtainedfromthe2009MedicareFeeSchedule.Expectedcostsandeffectswerediscountedat3%. RESULTS: Over25annual screeningcycles,thestrategyoftheindexaddingtoPSAdominatedthePSAtest aloneforprostatecancerdetection.Itwasestimatedtosave$234,withanexpected gainof0.02qualityadjustedlifeyears(QALYs).Theprobabilityoftheindextest beingcost-effectiveisapproximately85%attherangeof$50,000/QALYto$200,000/ QALYwillingnesstopay.Modelresultsaremostinuencedbyscreeningstarting age,discountrate,andbiopsyutilizationrates. CONCLUSIONS: Theindexasanaid addingtoPSAtestmaybeanimportantstrategyforprostatecancerdetectionas comparedtousingPSAalonetesting. PMD21 ISCYBERKNIFEACOST-EFFECTIVEOPTIONFORTREATINGPROSTATE CANCER?ParthanA1,PruttivarasinN2,TaylorD2,DaviesD3,YangG4,PawarV2,WeinsteinM5 1i3Innovus,SanFrancisco,CA,USA,2i3Innovus,Medford,MA,USA,3AccurayWorldwide Headquarters,Sunnyvale,CA,USA,4TheLewinGroup,FallsChurch,VA,USA,5HarvardSchool ofPublicHealth,Boston,MA,USAOBJECTIVES: Toassessthecost-effectivenessofCyberKnife(CK)comparedtosurgeryandradiationtherapiesforthetreatmentoflocalizedprostatecancer(PC) fromsocietalandthirdpartypayerperspectives. METHODS: AMarkovmodelwas usedtocomparetreatmentwithCK,intensitymodulatedradiationtherapy(IMRT), protontherapy(PT),andsurgery,in60-year-oldpatients.Themodelreectedboth procedure-relatedmortalityandthecomparativerisksoflong-termtoxicityamong survivors,denedasadverseevents grade2onRadiationTherapyOncology Groupscaleoccurringatleast12monthsfollowingtreatment:genitourinary(GU); gastrointestinal(GI);and/orsexualdysfunction(SD).Intheabsenceofevidenceon comparativeeffectivenessweassumedthatlong-termdiseasecontrolandmortalitywouldnotdifferacrosstreatments.Toxicityprobabilitieswerederivedusing meta-analyticaltechniques.UtilitiesforadverseeventswerederivedfromapublishedsurveyofPCpatientsusingstandardgambletechnique.Model-projected expectedlifetimecostsandqualityadjustedlifeyears(QALYs)foreachtreatment wereusedtocalculatetheincrementalcost-effectivenessofCKversuscomparators.Thesocietalperspectiveincludedproductivitycostsowingtotimespentin treatment.Extensivesensitivityanalyseswereconducted. RESULTS: Fromapayer'sperspective,surgerywasleastexpensivefollowedbyCK,IMRTandPT.However,CKpatientshadhigherexpectedQALYs(9.54)thanotheroptions(9.08-9.49). IncrementalcostperQALYgainedforCKversussurgerywas$13,100/QALY.ComparedtoIMRTandPT,CKwaslesscostlywithhigherQALYs(dominant).FromaA82VALUEINHEALTH14(2011)A1–A214

PAGE 91

societalperspective,CKcosts$8,400/QALYcomparedtosurgery;CKremained dominantversusIMRTandPT.ResultsweremostsensitivetocostofCKand surgery,andutilityweightsforGUandSD. CONCLUSIONS: CyberKnifewasfound tobecost-effectiveversussurgery,andresultedincostsavingsandimproved quality-adjustedsurvivalcomparedtoradiationoptionsforthetreatmentoflocalizedPC. PMD22 COST-EFFECTIVENESSANALYSISOFTHREEWOUNDDRESSINGSFORTHE TREATMENTOFPRESSUREULCERSFROMTHEPUBLICHOSPITALPERSPECTIVETolentinoACM,TakemotoML,FernandesRA,TakemotoMMS,SantosPML ANOVA-KnowledgeTranslation,RiodeJaneiro,BrazilOBJECTIVES: EachyearinBrazil,42.3%ofallinpatientspresentskinulcersduring thehospitalizationperiod.Currentclinicalpracticeguidelineshavenotestablished agoldstandardprotocolofcareforpressureulcers.Thus,thisstudyaimedto developcost-effectivenessanalysiscomparingthreedifferentdressings:hydrocolloid,silversulfadiazine1%cream(SSD)andsalinegauze(SG),undertheperspectiveofBrazilianpublichospitals. METHODS: Themeantimetohealing(MTH)for eachprotocolwasobtainedfromsystematicreviews.DatafromtheBrazilianHospitalInformationSystemfromJanuary1sttoDecember31st2009wasusedto denetheannualnumberofhospitaladmissionsduetopressureulcers(only non-surgicalrecordswithL89ICD-10codewereincluded).Themodelassumed thatSGisthecurrentpracticeinBrazilianpublichospitalsandpatientsaredischargedatthetimetheirwoundheals.ThedifferenceinMHTwasappliedtothe averagelengthofstay(LOS)reportedinthedatabase.Resourceusewasestimated throughexpertpanelandunitcostswereobtainedfromBrazilianofcialpricelists. RESULTS: 934hospitalizationswereidentiedwithmeanLOSof13.72days.HydrocolloidandSSDwouldreducetheMHTin4.72and2.72days,thusreducing patients'LOS.Thecostperchangewasestimatedas15.52BRL,43.20BRLand 15.72BRLandthecostperprotocolperpatient(dailyroomchargesanddressing changes)was902.50BRL,559.61BRLand907.99BRL,forSG,hydrocolloidandSSD, respectively.Adoptinghydrocolloidaswoundmanagementprotocolwouldsave 342.89BRLperpatientand-BRL320,259forthe2009cohort.SSDprojectedsavings was 5.49BRLperpatientand-BRL5,125.79fortheentirecohort. CONCLUSIONS: HydrocolloiddressinghasshownhigherefcacywhencomparedtoSGorSSD dressings,withfewercosts.Theclinicalandeconomicincrementalresultsbetweendifferentdressingsreinforcetheneedofevidence-baseddecisionmaking andrationalresourceallocation. PMD23 COST-EFFECTIVENESSOFSPECIFICUSEOFTHELANCETBDQUIKHEELIN SCREENINGPROGRAMOFNEONATALCONGENITALHYPOTHYROIDISMIN MEXICOVallejosPar‡sA1,ContrerasI2,OjedaMŽndezJ1,YamOntiverosC2 1UniversidadNacionalAut—nomadeMŽxico,MŽxicoCity,Mexico,2InstitutoMexicanodelSeguro Social,MŽxicoCity,MexicoOBJECTIVES: Congenitalhypothyroidism(CH)isaseriousconditionsandexpensiveillness.Thepurposeofthisstudywastodevelopaneconomicmodelinorder toevaluatethecost-effectivenessratiosbetweentwoheellancetdevicesinterms successoftheprocedureinneonatesundergoingthenewbornscreeningtestatthe SocialSecurityMexicanInstitute(IMSS)fromthehealthcarepayer'sperspective. METHODS: Acost-effectivenessanalysiswasdevelopedusingadecision-tree model.Themodelsimulatescostsandeffectivenessoutcomesina15yearsperiod. Thecomparatorswere:HeellanceBDSafety-Flowlancet¨andtheBDQuikHeel lancet¨.Theeffectivenessmeasurewasthenumberofcasesofseverecongenital hypothyroidismavoidedattheendofthefollow-upperiod.Effectivenessdataand transitionprobabilitiesweretakenfrominternationalpublishedliterature.The estimationofresourceusewasperformedemployinglocalexpertopinionsurveys. Costandeffectivenesswerediscounted5%annually.Themodelwascalibrated accordingtointernationalpharmacoeconomicsguidelines.One-wayandprobabilisticsensitivityanalyseswereperformedusingMonteCarloSimulationsecondorderapproach. RESULTS: Regardingeffectiveness,thelancetContactActivated wouldprevent75%ofpatientswithCHdetectedinadvancedstagesandwiththe useofthelancetQuickheelthepercentagewouldincreaseto98%,preventing23% morecasesofsevereCH.Themeancostpernewbornscreenedincaseofusingthe lancettypewouldContactActivated$12.2andwiththelancetQuikheeltype $11.71.ThepotentialsavingsfromtheuseofthelancetQuikheelwouldbearound $779,545.75intheholdprogram. CONCLUSIONS: Withthesedatawecanconclude thattheuseofthelancetQuikheeltypeisthebestalternative,asitoffersbetter resultsatlowercost.InMexicotheuseofthelancetQuikheelbringssignicant clinicalandeconomicbenetsthatmakeoftheneonatalscreeningprograma betterone. PMD24 GLYCOSYLATEDHEMOGLOBINORFASTINGGLUCOSETESTINGFOR SCREENINGDIABETESINCOLOMBIA:ACOST-EFFECTIVENESSANALYSISVecino-OrtizAI1,GarrisonL2,Alfonso-CristanchoR3 1InstituteforHealthMetricsandEvaluation-UniversityofWashington,Seattle,WA,USA,2UniversityofWashingtonDepartmentofPharmacy,Seattle,WA,USA,3Universityof Washington,Seattle,WA,USAOBJECTIVES: Diabetesisagrowingcauseofdeathandmorbidityinmiddle-income countriescarryingalargenancialburdenforhealthsystems.Inrecentyears, diabetesscreeningpracticesinColombiahaveshiftedfromusingthefastingglucosetest(FGT)towardsrelyingonglycosylatedhemoglobin(HbA1c)asphysicians consideritamorereliabletooltoearlydetectdiabetes.Thisanalysisassessesthe cost-effectivenessofalternativescreeningstrategiesingeneralpopulationinColombiafromahealthsystemperspective. METHODS: Adecision-analyticmodel wasdesignedtocomparefourstrategies:(1)FGT conrmatoryFGT(standard practice);(2)FGT conrmatoryHbA1c;(3)OnlyHbA1cand(4)noscreening.Longtermhealthoutcomes(lifeyears)andcost-outcomesweremodeledviaa20-year Markovmodelwiththreediabetesstates:non-complicated,complicateddiabetes anddeath.ParametervalueswerebasedondatafromtheNationalHealthSurvey andfromtheMinistryofSocialProtectionongeneralpopulation.ThemodelprojectionswerecomparabletothosepublishedelsewhereusingCOREandColombian literatureondiabetescosts.One-waysensitivityanalysisontheHbA1ccostwas alsoperformed. RESULTS: Nosignicantdifferencesoncostsperlife-yearsaved werefoundwhencomparingFGT conrmatoryFGT($1,047)andFGT conrmatoryHbA1c($1,069)againstno-screening.However,theuseofonlyHbA1cwas associatedwithahighercostperlife-yearsaved($2,455.93)whencomparing againstno-screening.Thisresultwasmainlydrivenbynotusingaconrmatory testratherthanbythetestitself. CONCLUSIONS: ThisanalysisfoundtheperformanceofFGT conrmatoryFGTandFGT conrmatoryHbA1ctobesimilarand highlycost-effectiveingeneralpopulationolderthan45whenincomparisontono screening.ThisndinghasimplicationsinbothclinicalandhealthcarepolicymakersinColombiaandinotherLatinAmericancountrieswithsimilardiabetesprevalenceandtreatmentoptions. PMD25 THECOST-UTILITYANDVALUEOFINFORMATIONOFTRANSCATHETER AORTICVALVEIMPLANTATIONCOMPAREDTOSTANDARDMANAGEMENT ANDSURGICALAORTICVALVEREPLACEMENTINPATIENTSWITHSEVERE SYMPTOMATICAORTICVALVESTENOSISDobleBM,PerampaladasK,vonKeyserlingkC,CampbellK,BlackhouseG,GoereeR, XieF McMasterUniversity,Hamilton,ON,CanadaOBJECTIVES: Theprimaryanalysisofthisstudywastoestimatethecost-effectivenessofTAVIcomparedtostandardmanagement(SM)ininoperablepatientswith severe,symptomaticaorticvalvestenosis(SSAVS).Thesecondaryanalysiswasto preliminarilyexploreboththecost-effectivenessandvalueofinformationofTAVI comparedtosurgicalaorticvalvereplacement(SAVR)inoperablepatientsathigh surgicalriskwithSSAVS. METHODS: AcombineddecisiontreeandMarkovmodel wasdevelopedtocomparethecosts,life-years(LYs)andquality-adjustedlifeyears(QALYs)ofTAVI(transfemoral(TF)andtransapical(TA)approaches)toSM andSAVRovera5-yeartimehorizon.Thisevaluationwasconductedfromathird partypayer'sperspective. RESULTS: Intheprimaryanalysis,comparingTFandSM resultedinanincrementalcost-effectivenessratio(ICER)of$126,874/LYand $222,378/QALY.ComparingTAandSMresultedinanICERof$262,672/LYand $1,454,241/QALY.Inthesecondaryanalysis,TFandSAVRwerecompared,resulting inanICERof$39,676/LYand$81,758/QALY.ComparingTAandSAVRresultedinan ICERof$183,454/LY.TAwasdominatedbySAVRwhencomparingQALYs.Thetotal expectedvalueofperfectinformation(EVPI)wasatamaximumwithavalueof $6,928ataWTPthresholdof$80,000/QALY.Theexpectedvalueofpartialperfect information(EVPPI)washighestforthe30-dayclinicaleventrateswhencompared totheEVPPIvaluesforthe1-year,2-yearand 3-yearclinicaleventrates. CONCLUSIONS: ThiseconomicevaluationsuggestedthatTAVImightnotbea cost-effectiveoptionforinoperablepatientsincomparisontoSM.Thesecondary analysissuggestedthatTAVImightbeacosteffectiveoptionforoperablepatients comparedtoSAVR.Toreducetheuncertaintyinourestimateitmightbeworthwhiletoobtainempiricalevidencerelatedtoclinicaleventratesoccurring30-days postoperation. PMD26 ECONOMICEVALUATIONSFORSCREENINGANDTREATMENTSOFDIABETIC RETINOPATHYANDDIABETICMACULAREDEMA:ASYSTEMATICREVIEWLinV1,YehWS2,KowalskiJ2 1UniversityofWashington,Seattle,WA,USA,2Allergan,Irvine,CA,USAOBJECTIVES: Newtechnologiesinscreeningandtreatmentofdiabeticretinopathy (DR)anddiabeticmacularedema(DME)haveemergedrecently.Thegoalofthis systemicliteraturereviewistoidentifyandcomparecriticalgapsinthepublished economicevaluationliteraturesforscreeningsandtreatmentsstrategiesof DR/DME. METHODS: AsystematicliteraturesearchwasconductedtoidentifyliteratureinEnglishfrom1980-2010,usingPubMed,Embase,NHSNICEEED,andTuft CEARegistry.Keytermsincluded"diabeticretinopathy","diabeticmacular edema",andrelevanttermsforeconomicevaluations.Allstudiesrelatedtoeconomicanalysisanddecisionmodelingwereincluded.Studiesthatfocusedonlyon costorutilitywereexcluded. RESULTS: 52articleswereidentiedand33were excludedbasedonthepre-speciedexclusioncriteria.Oftheremaining19studies, nine(47%)focusedonscreeningmethods,six(32%)ondiabeticcare,andfour(21%) onDR/DMEtreatment.Amongtheincludedstudies,12studies(63%)werecosteffectiveanalyses,andsight-yearssavedwasthemostfrequentlyusedendpoint. Sevenstudies(37%)werecost-utilityanalyses.Abroadrangeofdecision-analytic frameworksandhealthstatedescriptionswereobserved.Someonlyassignedutilitytoblindness,nottodiabeticpatientswhomayhaveimpairedqualityoflifedue tovisionloss.Inaddition,althoughDR/DMEisabilateralcondition,moststudies didnotclearlyexplainhowcost,utility,anddiseaseprogressionweremodeledfor thesecondeye. CONCLUSIONS: Thisliteraturereviewidentiedarangeofdecision analyticframeworks,healthstatesdenitions,andutilitysourcesemployedin economicevaluationsofscreeningandtreatmentofDR/DME.Futurestudiesassessingcost-effectivenessofnewtechnologiesshouldtransparentlyaddressthese areas.A83VALUEINHEALTH14(2011)A1–A214

PAGE 92

PMD27 COST-MINIMIZATIONANALYSISOFBIOPSY-BASEDRISKSTRATIFICATION TOOLSININTERMEDIATEANDHIGHRISKPROSTATECANCERPATIENTS BASEDONRESULTSFROMPHYSICIANCASESTUDIESZubekVB,KhanF,KarvirH AureonBiosciences,Inc.,Yonkers,NY,USAOBJECTIVES: ProstateCancer(PCa)diseasemanagementischallengingduetothe uncertaintyinriskatdiagnosis.PatientsclassiedbytheAmericanUrological Association(AUA)asintermediate(IR)orhighrisk(HR)maybeguidedtotherapies thatareexpensiveandunnecessarilyimpactqualityoflife.ProstatePx isabiopsy-basedassaythatprovidesimprovedprognosticinformationandaimstofacilitatemoreappropriatediseasemanagement.Weperformedacost-minimization analysisofPCamanagementaccordingtostandardcareandofmanagementpost Px ,basedoncasestudiesreviewedbyurologists.Thisstudyisanextensionof previousIRworkwhichemployedtheoreticaltreatmentdistributionsratherthan leveragingactualphysicianrecommendationsbasedoncasestudies. METHODS: 23urologistsnationwidereviewedcasestudiesofPCapatientsandrecommended treatmentswithandwithoutthePx results.233AUAIR(83%)andHR(17%)cases wereanalyzed.Adecisionanalysismodel,accountingforveprimarytreatments (radicalprostatectomy,radiationtherapy,primaryhormonaltherapy,brachytherapyandactivesurveillance)plussecondarytreatments,wasusedtocomparestandardcarewithmanagementpostPx reclassication.Bothprimarytreatment distributionsandprobabilitiesofPx reclassicationwerebasedonthephysician casestudies.Thediseaseprogressionmodelswereinformedwithcostsandprobabilitiesfromtheliterature.TheanalysisisfromMedicare'sperspectivewitha10 yeartimehorizon. RESULTS: Theexpectedcostperpatientofstandardmanagementwas$35,886and$33,997forPx (includingthe$3,200listprice),resultingin anexpectedcostsavingsof$1,889.One-waysensitivityanalysisof40variables conrmedthatmodeledmanagementpostPx savescostsforallbutthreevalues ofprogressionprobabilitiespostradicalprostatectomy. CONCLUSIONS: PersonalizedriskassessmenttoolssuchasPx improvePCariskstraticationofclinically challengingintermediateandhighriskpatients,impactingdiseasemanagement andpotentiallysavingcostsovercurrentstandards. PMD29 PATIENTTIMEANDINDIRECTCOSTSASSOCIATEDWITH SENSOR-AUGMENTEDINSULINPUMPTHERAPYINTYPE1DIABETESKambleS1,WeinfurtKP2,PerryBM3,SchulmanKA1,ReedSD1 1DukeClinicalResearchInstitute,Durham,NC,USA,2DukeUniversityMedicalCenter,Durham, NC,USA,3MedtronicDiabetes,Northridge,CA,USAOBJECTIVES: Inarecentrandomizedtrial,sensor-augmentedpumptherapy (SAPT)demonstratedimprovedglucosecontrolinType1diabetespatientscomparedtomultipledailyinjectionsofinsulin(MDI).ItisimportanttoevaluatepatienttimedemandstoachieveclinicalbenetsofSAPT.Usingdatafromthistrial, wecomparedreportedestimatesofweeklytime,changesintimeestimatesacross the52-weekstudyperiod,andindirectcostsassociatedwithdiabetes-relatedcare betweentreatments. METHODS: Duringthe52-weektrial,patients(age7-70years, n 483)providedweeklytimeestimatesfordiabetes-relatedcare.Givenadistinct differenceinthedistributionsofpatienttimebetweentreatmentsduringthepump initiationperiod,weusednonparametricbootstrappingtoestimatetheaverage weeklytimedifferenceanditsassociateduncertaintyforWeeks1-7.ForWeeks 8-52,weappliedalinearmixedmodeltoevaluatetheincrementaltimespentby patientsonSAPT(versusMDI)andchangesinweeklytimeforbothtreatments.We comparedindirectcostsusingnonparametricbootstrapping. RESULTS: Atbaseline,MDIpatientsspentanaverageof4.0hoursondiabetes-relatedcare.During thepumpinitiationperiod,patientsonSAPTversusMDIspent1.9hoursmoreper week(95%CI:1.2,2.6).Followingthisperiod,SAPTpatientsspent4.4hoursperweek andMDIpatientsspent3.4hoursperweekondiabetes-relatedcare(difference:1.0, 95%CI:0.4,1.7).Onaverage,timespentbypatientsinbothtreatmentsdecreasedby 1.2minutesperweek(95%CI: 1.7,-0.7),acrossWeeks8-52.Totalindirectcostsper patientaveraged$4,600withSAPTand$3,523withMDI(difference:$1,077,95%CI: $491,$1,638). CONCLUSIONS: PatientsonSAPTversusMDIspentapproximately twohoursmoreperweekondiabetescareduringpumpinitiationandonehour moreperweekthereafter,resultinginhigherindirectcosts.Thisadditionaleffort wasassociatedwithsignicantimprovementinglycatedhemoglobinlevelswith SAPT. MedicalDevice/Diagnostics–Patient-ReportedOutcomes&Preference-Based Studies PMD30 IMPACTOFACCESSTOAUSUALSOURCEOFCAREONCOMPLIANCETO BREASTANDCERVICALCANCERSCREENINGDatarMV1,WuWK2 1UniversityofMississippi,Oxford,MS,USA,2St.John'sUniversity,Jamaica,NY,USAOBJECTIVES: Trendsincompliancetocancerscreeningamongwomenhavebeen decreasingduringthelasttenyears.Inaddition, HealthyPeople2010 hasnoticeda declineintheprevalenceofausualsourceofcareintheAmericanwomen.The objectiveofthisstudywastoassesstheimpactofaccesstoausualsourceofcare oncompliancetomammographyandPAPsmearscreening. METHODS: The2007 MedicalExpenditurePanelSurvey(MEPS)datawasused.Screeningguidelines givesbytheAmericanCancerSocietywereusedasameasureofcompliance. MultivariatelogisticregressionswereperformedtodetermineassociationsbetweenaccesstoausualsourceofcareandcompliancetomammographyandPAP smearwhilecontrollingforappropriatedemographiccharacteristics.Samplewas weightedandanalysiswasperformedusingSPSS(PASW)17.0. RESULTS: 80%ofthe USnon-institutionalizedwomenpopulationhadausualsourceofcarein2007. 59.1%womenaged41-70yearsshowedcompliancetomammographyguidelines. Womenwhohadausualsourceofcare,hadahigherlikelihoodofmammography compliance(OR 2.540;CI:2.5352.545).80.2%womenaged22-70yearswerecomplianttoPAPsmear.Womenwhohadausualsourceofcareshowedahigher likelihoodofPAPsmearcompliance(OR 2.041;CI:2.0382.044). CONCLUSIONS: TheresultsindicatethatprevalenceofmammographyandPAPsmearisdecreasingandisstillbelowtherequired HealthyPeople2010 objectivesof70%and 90%respectively.AccesstoausualsourceofcareisessentialforincreasedcompliancewithmammographyandPAPsmearscreening.Thisstudyindicatesthat amendmentsmadeattheentrylevelofthehealthcaresystemmighthelpachieve betterhealthoutcomesinthefuture.Publicpoliciesshouldbedevelopedtoencouragetheconceptofausualsourceofcaresoastoachieveincreaseduseof preventivecare. PMD31 IMPACTOFGASTRICELECTRICSTIMULATIONONHEALTHSTATEUTILITIESOF DIABETICGASTROPARESISPATIENTS:RESULTSFROMAPROSPECTIVE CLINICALTRIALSolimanAM1,CarlsonA2,WittekMR3 1CollegeofPharmacy,UniversityofMinnesota,Minneapolis,MN,USA,2DataIntelligence Consultants,LLC,EdenPrairie,MN,USA,3Medtronic,Inc.,Minneapolis,MN,USAOBJECTIVES: Diabeticgastroparesis(DG)posesasignicanteconomicandclinical burdenduetopoorlyunderstoodphysiologicalmechanismsandunresponsivenesstomanymedicaltreatments.Gastricelectricstimulation(GES)isatreatment modalitythataimstoreducesymptomsofnauseaandvomitingassociatedwith gastroparesisinmedicallyrefractorypatients.Theaimsofthisanalysiswere:1)to compareHRQOLinDGpatientstopatientswithcommonchronicdiseases;and2) todeterminetheimpactofGESonhealthstateutilities. METHODS: HRQOLdata wascollectedduringaprospectiverandomizedcontrolledtrialwhereagastric electricstimulatorwasimplantedinpatientswithrefractoryDG(n 55).Following implant,thedevicewasturnedONfor1.5monthsafterwhichpatientswererandomizedtogroupshavingconsecutivecrossoverperiodswiththedeviceturned ONorOFF.Perprotocol,aftercrossoverthedevicewasturnedONfortheremainingperiodofthetrial.SF-36resultsatbaselineandat12monthswereusedfor analysis.Domainscoreswerecomparedtopublishedpopulationnormsandaveragescoresforcommonchronicconditions.Scoreswerealsousedtocompute healthstateutilitiesusingNicholsandcolleagues'mappingalgorithm. RESULTS: MeanSF-36domainscoresforenrolledpatientswerelowerthanscorespublished forpopulationnorms,andforpersonswithdiabetesandotherchronicconditions. Meanhealthstateutilityvaluesforpatientspriortodeviceimplantationwas0.54 (0.50-0.59);implantingGESresultedinanimprovementinhealthstateutilitiesto 0.65(0.60-0.71),whichwasstatisticallysignicant(p 0.001). CONCLUSIONS: PatientswithrefractoryDGhadsignicantlyworsehealthstateutilitiescomparedto patientswithseriousmedicalconditions.UsingGESforaperiodofapproximately 9monthsresultedinsignicantimprovementsinHRQOLforpatientswhoseDG wasrefractorytocurrentlyavailablemedications. PMD34 PREFERENCESFORGENETICTESTINGTOASSESSCOLON-CANCERRISKSMohamedAF1,JohnsonFR1,KnightSJ2 1RTIHealthSolutions,ResearchTrianglePark,NC,USA,2UniversityofCalifornia,SanFrancisco, SanFrancisco,CA,USAOBJECTIVES: Toelicitpreferencesforgenetictestingandestimatetherelative importanceandvalueoffeaturesofgenetictestingforcoloncancer. METHODS: UnitedStatesresidentsaged50yearsandoldercompletedaweb-enabled,choiceformatconjointsurvey.Thesurveypresentedsubjectswith9test-choicequestions.Eachquestionrequiredevaluatingapairofhypotheticalblood-testproles denedbychanceofdevelopingcoloncancer,chanceofreceivingafalse-negative result,personwhoreceivesthegenetic-testresults,andpersonalcostversusa no-testoption.Test-choicequestionswerebasedonapredeterminedexperimentaldesignwithknownstatisticalproperties.Thesurveyalsoelicitedpreferences forcolorectal-cancerriskreductionconditionaloninformationthatsubjectshad Lynchsyndromegeneassociatedwithahighriskforcoloncancer.Subjectswere askedtochoosebetweenagivennumberoflifetimecolonoscopiesoracolectomy atspeciedcoststoreducetheriskofdyingfromcolorectalcancer.Randomparameterslogitwasusedtoestimatemodelparameters,predictedtestingprobabilitiesforspecictests,andthemoney-equivalentvalueofgeneticinformation. RESULTS: Atotalof451subjectscompletedthesurvey.Giventherangeoflevelsof eachattributeinthestudy,themostimportantattributewascost,followedbywho elseseesthetestresults,thegenetic-testingpreference,andchanceofafalsenegativetestresult.Forthe"best"testcomparedwithnotestwherethebesttest has0%chanceofafalse-negativetestresult,resultsaredisclosedtoaphysician, andcostis$500;thepredictedprobabilityofchoosingthetestis97%(CI:95%-99%). Settingthetestattributestothemeaneffect,theoverallmonetaryvalueforgenetic testingwas$622. CONCLUSIONS: Costandthepotentialforinsurancediscriminationbasedongeneticsarethemostimportantconsiderationsintestingdecisions andaremoreimportantthantestaccuracyandreductionofcancerrisk.A84VALUEINHEALTH14(2011)A1–A214

PAGE 93

MedicalDevice/Diagnostics–HealthCareUse&PolicyStudies PMD35 SHORTTERMOUTCOMESASSOCIATEDWITHACHANGEINREIMBURSEMENT FORDEEPBRAINSTIMULATIONINPARKINSON’SDISEASEHaleyJ,RasuR UniversityofMissouri-KansasCity,KansasCity,MO,USAOBJECTIVES: DeepBrainStimulation(DBS)isaParkinson'sdisease(PD)treatment thatsurgicallyimplantsvibratingprobesinthebraintoreducesymptoms.This studyexaminedwhetherpatientcharacteristics,hospitalcharacteristics,oroutcomesdifferbeforeandafteraCenterforMedicaidandMedicareServicesNational CoverageDecision(NCD)forDBSinPD. METHODS: TheHealthcareCostandUtilizationProject(HCUP)NationalInpatientSamplefortheyears2002-2004wasused forthisanalysis.HCUPissponsoredbytheAgencyforHealthcareResearchand Quality.Thisobservational,cross-sectionalstudyexamined12-monthsbeforeand afterthe04/01/2003NCD.InpatientstayswereidentiedusingthePDICD-9code andtheDBSCPTcode.Multivariateanalysestestedwhetherdischargestatusor lengthofstay(LOS)differedaftertheNCD. RESULTS: Thisstudyevaluated460 inpatientstaysforPDDBS.Thispopulationhadameanageof63.19(range,32-87); was66.09%male;and84.13%ofsurgerieswereinlargehospitals.Inthepostperiod,patientsweremorelikelytoliveintheSouth(p 0.0001);lesslikelytousea largehospital(p 0.0001);morelikelytousehospitalsdoingfewerDBSprocedures (p 0.001);andmorelikelytousephysicianswithfewerDBSprocedures(p 0001). Afteradjustingforhospitalandpatientcharacteristics,multivariateanalysespredictedthatcomparedtothepre-period,dischargetoashort/longtermcarefacility ratherthanhomewas3.671timesmorelikelytooccurinthepost-period[OR:3.671, CI(1.178-1.441)];andpredictedthatlongerLOSwouldoccurinthepost-period(Coefcient 0.3022,p 0.0003). CONCLUSIONS: Thisanalysissuggeststhatcompared tobeforetheMedicarebenetexpansion,thepatientpopulationaftertheexpansionexperiencedworseoutcomes.Furtherresearchisneededtoexaminewhether thesedifferencesremainovertimetoinformPolicymakersabouthowandwhy outcomesmaydifferbeforeandafterabenetexpansion. PMD36 FAILUREOFINTERNATIONALORGANIZATIONFORSTANDARDIZATION GUIDELINESTOREPORTCLINICALLYRELEVANTINFORMATIONABOUT POINT-OF-CAREINRDEVICEPERFORMANCEShermockKM1,LavalleeDM2,StreiffMB3 1TheJohnsHopkinsMedicalInstitutions,Baltimore,MD,USA,2CenterforMedicalTechnology andPolicy,Baltimore,MD,USA,3TheJohnsHopkinsUniversitySchoolofMedicine,Baltimore, MD,USAOBJECTIVES: ArevisedISOguidelinewasrecentlypublishedthatrecommended methodstobeusedduringthedeviceapprovalprocessandtoassessthequalityof point-of-careINRmeasures.Pointofcaredevicesshouldleadclinicianstothe properclinicaldecisionthesamedecisiontheywouldmakeusingtheclinicallyacceptedstandardlaboratorymeasure.Ourmethodwasdevelopedexplicitlyto correspondwiththeclinicaldecision-makingprocess.Weassessedtherelative performanceoftheISOmethodsandamethodthatwedevelopedandvalidated. METHODS: Inaprevioustrial,patientsprovidedtwoINRmeasuresatthesame clinicvisit:onebypoint-of-caredeviceandtheotherbyvenoussampleanalyzedat aclinicallaboratory.TheresultantclinicaldecisionsfromtheINRmeasureswere directlymeasuredintheprevioustrialandwereusedasastandardforthisanalysis.WecomparedtheabilityoftheISOandourmethodstopredictwhenINR measuresleadtoidenticalordifferentclinicaldecisions. RESULTS: TheShermock methodwassignicantlybetter(82%ofpredictionscorrect)thantherevisedISO method(61%correct,p 0.0001)atpredictingwhentwoINRmeasuresleadtoidenticalordifferentclinicaldecisions.Only41%ofdecisionspredictedtodisagreeby therevisedISOmethodactuallydisagreed(Shermock'smethod 64%);whileonly 51%ofdecisionsthatactuallyagreedwere correctlypredicted(Shermock'smethod 83%). CONCLUSIONS: TheShermockmethodissuperiorinprovidinginformation thatcliniciansandpatientscaremostaboutregardingthequalityofINRmeasures usedtoguideclinicaldecisions.TheFDAshouldincorporateShermock'smethod intothedeviceapprovalprocess.LocalclinicallaboratoriesshouldusetheShermockmethodinqualityassuranceassessments.Othermethodsshouldnotbe usedunlesstheyprovideinformationthathasclinicalmeaningandisvaluedby end-users. PMD37 ANATIONALSTUDYOFOUT-OF-POCKETEXPENDITURESFORMAMMOGRAPHY SCREENINGLeMastersT,SambamoorthiU WestVirginiaUniversity,Morgantown,WV,USAOBJECTIVES: Toidentifyvariationsinscreeningmammographyexpenditures,primarilyout-of-pocketandtotalexpendituresofwomenintheU.S.4064yearsof age,andfactorsassociatedwithvariations. METHODS: Retrospectiveanalysisof datacollectedfromthe2007and2008MedicalExpenditurePanelSurvey.Sample included2,020women4064yearsofagewhoreceivedonemammograminthe year2007or2008.Ordinaryleastsquaresregressiontodescriberelationshipsbetweenout-of-pocketexpenditures,out-of-pocketexpenditureasapercentageof totalmammogramexpenditure,andtotalyearlyout-of-pocketexpenditureforall healthcareservices,andindependentvariablessuchasinsurancestatusandtype, income,regionoftheU.S.,andtypeoffacilitywhereamammogramwasreceived. RESULTS: Theaverageout-of-pocketexpenditureforamammogramin2007or 2008was$33,representing14.1%ofthetotalmammogramexpenditure($266). Evenaftercontrollingdemographicandhealthfactors,insurancestatusandtype, income,regionoftheU.S.,andtypeoffacilityremainedsignicantpredictorsof mammographyexpenditures.Womenwithoutinsurancepaidagreaterpercentage(31.0%)comparedtothosewithprivateinsurance(13.7%);womenwhoreceived theirmammogramatanoutpatientfacility(18.0%)andresidedintheSouthern regionoftheU.S.(16.5%)alsopaidagreaterpercentage. CONCLUSIONS: Large variationsinout-of-pocketexpenditureswereobservedamongwomenwithand withoutinsurance,betweeninsurancetypes,geographicregionsoftheU.S.and typesoffacilitieswheremammogramswerereceived.Highernancialburdenof mammographyscreeningamongsubgroupsofwomenmayactasabarrierto futuremammographyscreening. PMD38 METAANALYSISOFTHEIMPACTOFHEALTHINFORMATIONTECHNOLOGYON MEDICATIONSAFETYINHOSPITALS:ANINTERRUPTEDTIMESERIESSTUDYChangJ1,PatelI1,KimB2,KwonI1,BalkrishnanR1 1UniversityofMichigan,AnnArbor,MI,USA,2UniversityofTexasHealthScienceCenterat Houston,Houston,TX,USAOBJECTIVES: Severalstudieshaveexaminedtheeffectsofinterventionrelatedto healthinformationtechnologyliketheElectronicPrescribing(eRx)systemonhospitalmedicationsafety.However,thesestudieshadrelativelysmallsamplesizes withhighlyvariableresults.Thestudyobjectiveistoperformmeta-analysessystematicallyreviewandquantifytheeffectsofinterventionrelatedtoElectronic Prescribingsystemonhospitalmedicationsafety. METHODS: MEDLINE,EMBASE, andtwootherdatabasesweresearchedfrom1998to2010forstudiesassessingthe usageofElectronicPrescribinginhospitals.Among29papersobtainedusingdata extraction,atotalof9studiesdiscussingpreandpost(Timeseriesstudy)utilizationofElectronicPrescribingwerechosentostudytheirimpactonthemedication safetyofthepatients.Outofthe9studies,6studiesinvolvedinpatientsettingsand theremaining3studiesinvolvedoutpatientsettings.InordertoquantifytheeffectsofElectronicPrescribingsystemonhospitalmedicationsafety,weusedthe MetaanalysisbySTATA¨SEversion10. RESULTS: Overall,medicationsafetywas signicantlyimprovedininterventiongroupinvolvingElectronicPrescribingsystem(OR 1.97,p 0.002)comparedwithcontrolgrouputilizinghandwritingsystem.Specically,medicationsafetyininpatientsettingswassignicantlyimprovedinElectronicPrescribingsysteminterventiongroup(OR 2.08,p 0.01). CONCLUSIONS: TheusageofElectronicPrescribingallowsentryandretrievalof patientclinicalinformationatthepointofcare,therebyenhancingdeliveryof qualityhealthcareservicesandloweringdeliverycostsforthoseservices. PMD39 MINI-HTATRENDSFORMEDICALDEVICESINTHENORDICSGrundstromJP,FribergS,MedinE HERONEvidenceDevelopmentLLC,Stockholm,SwedenOBJECTIVES: Amini-HealthTechnologyAssessment(mini-HTA)isadenedtoolin theformofachecklistdesignedtoquicklyadviseevidence-baseddecision-making forfundingwhenmedicaldevicesareintroducedoradapted.Thepurposeofthis studywastodescribetheexistingstructuresandtrendsformini-HTAsinthe Nordiccountries. METHODS: Astructuredweb-basedsearchwasperformedto identifyandcompileinformationfromnational,regionalandlocalauthoritiesand agencies.Themini-HTAsystemswereassessedbycomparingthedifferentstructuresacrosstheNordiccountries. RESULTS: Denmarkhasthemoststructured process,whichhasbeendevelopedandrenedsincetheearly1990sandistoday compulsoryinsomesettings.Themini-HTAformatisanationalstandardised checklistof3-5pagesbasedpartlyofalimitedsystematicreviewandconcerning theprerequisitesforandconsequencesofthemedicaldevicecoveringfourperspectives:device,patient,organisationandeconomics.Non-compliancewith thesefourperspectivesmayleadtorejectionofthemedicaldeviceforfunding. FourregionalHTAunitsinSwedenhaveimplemented,andotherregionsaredeveloping,standardisedprocessesformini-HTAswithastructureresemblingthe Danishformat.TheFinnishManagedUptakeofMedicalMethods(MUFF)group afrmsthebenetsofimplementingastructuredmini-HTAprocess.UntilaFinnishformatisinplaceithasbeenproposedtousetheDanishformatforevaluating medicaldevices.InNorway,experiencesfromreviewsofothermini-HTAsystems arebeingusedtoestablishanationalprocessforassessingnewmedicaldevices. CONCLUSIONS: CountriesintheNordicshaveseenadvantagesofusingmini-HTAs whentimeandresourcesareinsufcienttoexecuteacomprehensiveHTA.MiniHTAshaveimportantcriteriaandhavefacilitatedamorestandardisedandquick uptakeprocessofmedicaldevicesintheNordicsandhavegiventheindustrynew marketaccesspossibilities. MedicalDevice/Diagnostics–ResearchonMethods PMD40 METHODOLOGICALCONSIDERATIONSINMODELINGTHEECONOMICVALUE OFDIAGNOSTICACCURACYFoleyK1,LimS2,SchulmanKL3 1ThomsonReuters,Cambridge,MA,USA,2JeffersonUniversity,Philadelphia,PA,USA,3OutcomesResearchSolutions,Inc.,Bolton,MA,USABACKGROUND: Clinicalevaluationsofdiagnostictestsfocusprimarilyonthesensitivityandspecicityofatestrelativetoagoldstandard.Whenassessingeconomicvalue,however,payerswanttoknowofthosetested,whatpercentare accuratelydiagnosed?Thus,economicanalysesoftenrequirethemodeltobegin withthepopulationtestedandapplythepositiveandnegativepredictivevalues (PPV/NPV)ofthetesttodeterminethenumberoffalse-positives(FP)andfalsenegatives(FN).AlthoughPPVandNPVarefunctionsofthesensitivityandspecicityofatest,theyarealsofunctionsoftheunderlyingdiseaseprevalence.A85VALUEINHEALTH14(2011)A1–A214

PAGE 94

OBJECTIVES: Toillustratetherelationshipsamongsensitivity/specicity,disease prevalenceandPPV/NPV,anddemonstratetheimpactonmodelresultswhen diseaseprevalenceisnotconsidered. METHODS: Usingtheliteratureonthesensitivityandspecicityofsingle-photonemissioncomputedtomography(SPECT) andpositronemissiontomography(PET)fordetectingcoronaryarterydiseaseas anexample,wederivethePPV/NPVtakingprevalenceintoaccount.ThenwederivetherateofFNandFPfromthereportedmeansensitivity/specicityaswellas fromthevaryingPPV/NPV. RESULTS: MeansensitivityandspecicityforPETwere 0.9and0.83,withassociatedFNandFPof0.1and0.17.WhenusingPPV/NPVto deriveFNandFP,FNrangedfrom0.05to0.22whenprevalencevariedfrom0.3to 0.7,whileFPrangedfrom0.31to0.07.SensitivityandspecicityforSPECTwere reportedas0.85and0.72,withassociatedFNandFPof0.15and0.28.UsingPPV/NPV andaprevalencerangefrom0.3to0.7,FNvariedfrom0.08to0.33andFPvaried from0.43to0.12. CONCLUSIONS: Modelsresultscanbesignicantlybiasedif prevalenceisnottakenintoaccountwhenderivingFNandFPforeconomicmodels ofdiagnosticaccuracy. PMD41 SELECTINGASECONDARYDATASOURCEFORALOW-VOLUMEPROCEDUREIN ASPECIALTYPOPULATION:ASTUDYUSINGINPATIENTPEDIATRIC COLONOSCOPYAnastassopoulosKP,KnightTG,BaikR CovanceMarketAccessServices,Inc.,Gaithersburg,MD,USAOBJECTIVES: Theobjectiveofthisstudywastocompareestimatesfromsurveyand claimsdatasourcesforalow-volumeprocedurewithinaspecialtypopulation (specically,theannualnumberofinpatientcolonoscopiesperformedinthe UnitedStates(US)pediatricpopulationwithprivateinsurance)tobetterunderstandtheconsiderationswhenchoosingadatasource. METHODS: Aretrospective analysisofUShealthinsuranceclaimsandnationalsurveydatawasperformed usingThomsonReutersMarketScan¨CommercialClaimsandEncountersDa tabase(MarketScan¨),theNationalHospitalDischargeSurvey(NHDS),andthe NationalHealthcareCostandUtilizationProjectKids'InpatientDatabase(KID). Estimatesamongtheprivately-insuredwereobtainedoverallandbyage(0to17 years),usingICD-9-CMprocedurecode45.23forcalendaryear2007inMarketScan¨andNHDSand2006inKID(2007datanotavailable). RESULTS: Theoverall, annualestimateofinpatient,privately-insuredpediatriccolonoscopieswas similarbetweenNHDSandMarketScan¨;however,NHDSestimateswerenot statisticallyreliableanddiffereddramaticallyfromtheprioryear(overallestimatewas5.6timeshigherthan2006).TheoverallestimatefromMarketScan¨wasapproximately4.3timeshigherthanKIDwithestimatesbyagealsohigher; 2.8timesforage 1to4.2timesforagerange15-17.TheKIDdatabasereported thatapproximately50%ofcolonoscopieswereperformedinpediatrichospitals. CONCLUSIONS: TheestimatesweremarkedlydifferentbetweenMarketScan¨and KID,andwerenotstatisticallyreliableinNHDS.Lowrepresentationofpediatric hospitalsthatperformalargenumberofcolonoscopiesinthispopulationpossibly ledtoinconsistentorunderestimatedprojectionsinNHDSandKID,respectively. Researchersshouldbeinformedaboutthefrequencyoftheprocedureofinterest andrepresentativenessofthepopulationinthedatawhenselectingadatasource. Furtherresearchiswarrantedwithotherlow-volumeproceduresperformedin specialtypopulationstosubstantiatethesendings. Surgery–ClinicalOutcomesStudies PSU1 FREQUENCIES,COSTSANDCOMPLICATIONSOFCATHETERABLATIONSFOR PEDIATRICTACHYCARDIA:RESULTSFROMANATIONALPEDIATRIC INPATIENTDATABASE(YEARS2000-2006)DesaiVC1,KeltonC1,CzosekR2,HeatonPC1 1UniversityofCincinnati,Cincinnati,OH,USA,2CincinnatiChildren'sHospitalMedicalCenter, Cincinnati,OH,USAOBJECTIVES: Overthelasttwodecades,catheterablation(CA)hasrevolutionized thetreatmentofpediatrictachycardiabyprovidingarelativelysafealternativeto open-heartsurgeryorlifelongpharmacotherapy.Despiteitshighsuccessrate, however,CAhassomeriskofmajorcomplications,suchascompleteorseconddegreeatrioventricularblock,andahigherriskofminorcomplications,suchas hematoma.NostudytodatehasestimatednationalfrequenciesandcostsforCA anditscomplicationsinchildren.Theobjectiveswereto1)determinethefrequencyofCA;2)determinetheextentofmajorandminorcomplicationsassociated withCA;3)estimatetheaveragecostandlengthofstay(LOS)forchildrenundergoingCA;and4)predictthelikelihoodofmajorcomplicationsbasedonpatientand hospitalcharacteristics. METHODS: DatawereobtainedfromtheKids'Inpatient Database(KID)fortheyears2000,2003,and2006.DischargeswereselectedifCA (ICD-9code37.34)waslistedasaprimaryorsecondaryprocedure.Costswere computedusingtheKIDcost-to-chargeratios.Logisticregressionwasusedtopredicttheoddsofmajorcomplications. RESULTS: In2000,therewere1977pediatric CAs;in2003,therewere2049;and,in2006,2254.CAsinvolvingchildren0-11 monthsoldincreased2.8foldbetween2000and2006.Thepercentageofvisits involvingCAcomplicationsrangedfrom5.56%in2000to7.72%in2006.In2006,the meancostandmeanLOSforvisitswithCAwere$19,425( $1,022)and2.39( 0.2) days,respectively.TheoddsofmajorcomplicationsfromCAforatrialtachycardia weretwice(p 0.01)thatofsupraventriculartachycardia. CONCLUSIONS: In2006, thenationalinpatientcostofCAanditscomplicationstotaled$43,783,950.Since majorcomplicationsaresignicantlymorelikelyinyoungchildren(upto6years old),cautionisrequiredasmoresuchproceduresareperformedinthisagegroup. PSU2 USINGELECTRONICMEDICALRECORDSTOIDENTIFYPOTENTIALPREDICTORS FORBARIATRICSURGERYSUCCESSBenoitS1,HunterTD2,SteinbuchM3,FegelmanE4,FrancisD5,SeeleyR1 1UniversityofCincinnati,Cincinnati,OH,USA,2S2StatisticalSolutions,Inc.,Cincinnati,OH, USA,3Johnson&Johnson,NewBrunswick,NJ,USA,4JewishHospital,Cincinnati,OH,USA,5Johnson&Johnson,Cincinnati,OH,USAOBJECTIVES: Bariatricsurgeryforthetreatmentofobesityanditsunderlyingcomorbidconditionsisawell-toleratedandsafeprocedure.Whileasurgicalapproachisbecomingtherecommendedmethodoftreatmentformorbidobesity,not allpatientsexperiencesignicantimprovement.Theobjectivesofthisstudywere toutilizetheGECentricityelectronicmedicalrecords(EMR)databasetoperforma retrospectivecohortanalysisonweightlossovertimeinpatientswitharecordof havinglaparoscopicgastricbandingorbypasssurgerytoidentifypredictorsof weightlossfollowingbariatricsurgery. METHODS: Analyzablepatientrecords weredrawnfromtheGECentricitydatabasefromJanuary1990throughMarch 2010.Tobeeligibleforinclusion,patientshadarecordedCPTProcedureCodeof 43644/43645or43770; 18yearsofage;andavalidrecordedBMIwithin6months priortoprocedureandatleastonetimepointpostsurgery.UnivariateandmultivariableanalyseswereperformedusingSASversion9.2. RESULTS: Therewere somenotablepre-surgerydifferencesbetweenbypassandbandcohorts,including age,weight,BMI,selectedcomorbidconditionsanduseofantidepressants. Patients withpre-opBMIvaluesoflessthan30didnotexperiencesustainedweightloss,on average,regardlessofsurgerytype.Patientswithpre-surgeryBMIvaluesof30toless than35faredbetterwithbanding(20 %BMIloss)thanwithbypass(5%BMIgain)at2 yearspostsurgery.Finally,patientswithpre-surgeryBMIvaluesof35andabovehad greatersuccess,onthewhole,withbypasssurgery. CONCLUSIONS: ThepercentBMI lossovertimebyfourcohortsofpre-surgeryBMIshowatrendthatmaybeuseful inpredictingtreatmentsuccessasdenedbyweightloss.ThisanalysisofEMRs demonstratespotentialclinicalbenetswhenevaluatinglaparoscopicgastricbypassandbandinginarealworldsetting. PSU3 LONG-TERMCLINICALEFFECTIVENESSANDSTABILITYOFLASIKAND SURFACEABLATIONINKOREA:ASYSTEMATICREVIEWAPPROACHLeeNR1,ChoiJ1,JangEJ1,JungSY1,LeeEJ1,JooCK2 1NationalEvidence-basedHealthcareCollaboratingAgency(NECA),Seoul,SouthKorea,2The CatholicUniversity,Seoul,SouthKoreaOBJECTIVES: Thisstudyisaimedtoprovideevidenceforclinicaleffectivenessand stabilityafterLaserinsitukeratomilieusis(LASIK)andsurfaceablationsurgery (LASEK:Laser-AssistedSubepithelialKeratectomy,PRK:photorefractivekeratoectomy)formyopiainKorea. METHODS: Wesearchedelectronicdatabaseincluding Ovid-Medline,EMBASE,theCochraneLibraryandKoreandomesticdatabasesuch asKmbase,KoreaMed,NDSL,Kisti,KISS,andKJO(JournalofTheKoreanOphthalmologicalSociety).Twoindependentreviewersextracteddataandassessedthe qualityusingMINORS(MethodologicalindexforNon-RandomizedStudies).The changesofUCVA(uncorrectedvisualacuity)andmanifestrefractivefrompre-surgerywereestimatedusingtherandom-effectmodel.Cochrane-Q-statisticandI2statisticwereusedtoassessheterogeneity.Subgroupanalyseswereperformedby lengthoffollow-up(from24monthto84month)andmyopiadegreeunder-6D(diopter)andover-6D. RESULTS: Fifteenobservationalstudieswereincludedformeta analysisandallstudypatientswereKorean.AfterLASIKsurgery,comparetobefore surgery,UCVAwasimproved0.5decimalinhighmyopiagroup(95%CI:0.20-0.70). Accordingtosubgroupanalysisinhighmyopiagroup,thegroupof 6D -10D UCVAwasimproved0.63decimal,thegroupofover 10Dwasimproved0.27 decimal.Also,aftersurfaceablationsurgery,UCVAchangeinhighmyopia group(over 6D)was0.83decimalandinmildmyopiagroup(under 6D)was0.61 decimal.InadditiontoUCVAchangewasdecreasedoverfollowuptime.However, aftersurfaceablationsurgerymanifestrefractivechangewas7.45diopterinhigh myopiaand4.00diopterinmildmyopiagroup.Also,overfollowuptime,refractive changewasdecreased. CONCLUSIONS: Thisstudyprovidestheevidencethat UCVAwasdecreasedaccordingtoovertimeandinhighmyopiagroup.Also,Refractivechangewasfallendownincompliancewithovertime,butincreasedin highmyopiagroup.Ourstudyistherstonetoevaluateclinicaleffectivenessand stabilityofLASIKandsurfaceablationsurgeryinKoreathroughsystematicreview. PSU4 SURGEON’SVOLUME-OUTCOMESRELATIONSHIPFORLOBECTOMIESAND WEDGERESECTIONSFORCANCERUSINGVIDEO-ASSISTEDTHORACOSCOPIC TECHNIQUESDavidG1,GunnarssonC2,MooreM3,HowingtonJ4,MillerD5,MaddausM6, McKennaJr.R7 ,MeyersB8,SwansonS9 1UniversityofPennsylvania,Philadelphia,PA,USA,2S2StatisticalSolutions,Inc.,Cincinnati, OH,USA,3EthiconEndo-Surgery,Cincinnati,OH,USA,4NorthShoreUniversityHealthSystem, Evanston,IL,USA,5TheEmoryClinic,Inc.,Atlanta,GA,USA,6UniversityofMinnesota, Minneapolis,MN,USA,7Cedars-SinaiMedicalCenter,LosAngeles,CA,USA,8Washington UniversityinSt.Louis,St.Louis,MO,USA,9BrighamandWomen'sHospitalandtheDana FarberCancerInstitute,OldGreenwich,CT,USAOBJECTIVES: Thisstudyquantiesthebenetsofsurgeon'svolumeonoutcomes inlungsurgery:lobectomiesandwedgeresections.Thegoalofthisanalysisisto analyzetheeffectoftechnique-specicexperience(VATS)oncost,utilizationand adverseevents. METHODS: ThisstudyutilizesthePremierhospitaldatabasewhich containsclinicalandutilizationinformationonpatientsreceivingcareinover600 UShospitals.Eligiblepatientswerethoseofanyageundergoinglobectomyor wedgeresectionusingVATSforcancertreatment.Volumemeasuresuseaddi-A86VALUEINHEALTH14(2011)A1–A214

PAGE 95

tionaldataforpatientsundergoinglobectomyorwedgeresectionusingopenthoracotomy.Ourmeasureofvolumerepresentstheaggregateexperiencelevelofthe surgeonpersixmonthwindow.Multivariablelogisticregressionanalyseswere estimatedforthebinaryoutcome-adverseevents.OrdinaryLeastSquares(OLS) regressionwasusedforcontinuousoutcomes:hospitalcosts,surgerytime,length ofstay,andnumberofadverseevents.Inaddition,thefollowingexplanatoryvariableswereincluded:patientdemographics,diagnosis,comorbidconditions,patientseverityindexandhospitalcharacteristics. RESULTS: Of7137patientsinthe databasewithelective,inpatientresectionsforlungcancer,atotalof2698patients underwentlobectomy(n 716)orwedgeresection(n 1982)usingVATS.Morethan 70%oftheseprocedureswereperformedbythoracicsurgeons(n 1896).Apositive volume-outcomerelationshipappearedasfollows:Therelationshipisstrongerfor costandutilizationoutcomesversusadverseevents,forthoracicsurgeonsversus othersurgeons,andforVATSlobectomyproceduresversusVATSwedgeresection procedures.Finally,wendthatwhiletherewasareductionincostandresource utilizationassociatedwithgreaterexperiencewithVATS,theseoutcomeswerenot stronglylinkedwithgreaterexperiencewithopenprocedures. CONCLUSIONS: ThereisareductionincostandresourceutilizationassociatedwithgreaterexperiencewithVATS.Greaterexperiencewithopenproceduresdoesnotcorrelatewith betterVATSoutcomes. PSU5 LONG-TERMCLINICALSAFETYOFLASIKANDSURFACEABLATIONINSOUTH KOREATHROUGHLINKAGEOFRETROSPECTIVECOHORTANDNATIONWIDE CLAIMSDATABASELeeEJ1,ChoiJE1,JungSY1,JangEJ1,LeeNR1,JooCK2 1NationalEvidence-basedHealthcareCollaboratingAgency(NECA),Seoul,SouthKorea,2DepartmentofOphthalmologyandVisualScience,CatholicUniversityCollegeofMedicine,Seoul, SouthKoreaOBJECTIVES: Toassessthelong-termsafetyofLaserinsitukeratomilieusis(LASIK) andsurfaceablationsurgeryformyopiainSouthKorea. METHODS: Aretrospective cohortforpatientswhohadaLASIKorsurfaceablationsurgeryin6hospitals during2002-2005wereconstructedbasedonchartreviewsandlinkedtoHealth InsuranceReviewandAssessment(HIRA)claimsdatabaseduring2005-2009.A totalof2,422patients(4,688eyes)werelinkedusinganonymizedidentierafter excludingpatientswithahistoryofeyedisease,diabetes,hyperopia,ordifferent surgerymethodsforbotheyes.Thefrequencyofeyediseasesuchascataract, glaucoma,retinaldetachment,cornealopacity,andkeratoconusduring3-8years aftersurgeryandthepre-surgerycharacteristicsincludingage,UCVA(uncorrected visualacuity),manifestrefractive(MR)ofpatientswitheyediseasewereanalyzed. RESULTS: Therewere1,637LASIKpatientsand785surfaceablationpatients.The eyediseaseafterLASIKwerecataract(0.2%)andcornealopacity(0.9%)during3-4 yearsandcataract(0.1%),glaucoma(0.2%),retinaldetachment(0.4%)andcorneal opacity(0.7%)during4-5years.Theretinaldetachment(0.3%)andcornealopacity (1.3%)wereobservedduring7-8yearswhilenokeratoconuswasidentiedduring 3-8yearfollow-up.Thepre-surgeryMR(mean SD)were 6.98 1.41Diopter(D), 7.46 2.87D, 6.13 2.67D, 6.23 2.62Dinpatientswithcataract,glaucoma, retinaldetachment,andcornealopacity,respectively.Theincidenceofeyedisease aftersurfaceablationwassimilartoLASIK. CONCLUSIONS: Thelinkageofclinical baselinedatabaseandnationwideclaimsdatabasecanmakeupforlosstofollow-upinretrospectivecohortstudy.Althoughitisdifculttoexplainthecausality ofsurgery,wecouldascertaintherewasnoseriouseyediseaseaftersurgery. Surgery–CostStudies PSU6 ANALYSISOFFACTORSINFLUENCINGINPATIENTMORTALITYANDCOSTS AMONGPEDIATRICHEARTTRANSPLANTATIONRECIPIENTSOlveyEL1,NolanPE1,SkrepnekG2 1UniversityofArizonaCollegeofPharmacy,Tucson,AZ,USA,2UniversityofArizona,Tucson, AZ,USAOBJECTIVES: Toassesstherelationshipofpatient,payer,andhospitalcharacteristicswithcostandmortalityinpediatrichearttransplantrecipients. METHODS: Datafromthe1997,2000,2003,and2006AgencyforHealthcareResearchand Quality(AHRQ)HealthcareCostandUtilizationProject(HCUP)Kids'InpatientDatabase(KID)wereutilized.Pediatricpatientsaged0-17yearsreceivinghearttransplantswereincluded.Logisticregressionandgeneralizedlinearmodelswereused fortheanalyses.Inpatientmortalityandination-adjustedcostsweretheoutcomesofinterest.Patientdemographicsandclinicalcharacteristicsweretheprimaryindependentvariablesofinterest,withpayerandhospitalcharacteristics usedascontrolvariables.Weightingprocedureswereutilizedtoyieldnationally representativeresults. RESULTS: Overall,1154patientsreceivedhearttransplants duringthestudytimeframe.Theaveragewas6.5( 6.1)years.Meaninpatient hospitalcostswere$206,895( 150,504).Inpatientmortalitywas7.2%.Children 1 yearweresignicantlylesslikelytodieduringhospitalizationthanthose 1year (p 0.05).Hemorrhage(OR 4.11,p 0.005)andrenalfailure(OR 6.45,p 0.001) werecomplicationsassociatedwithmortality.Mortalitywasalsosignicantlyassociatedwithhighernumberdiagnosesandhigherincomeandinverselyrelated withWestregionandmorerecenttransplants.Mortalitycasesincurred1.68times highercoststhanthoseroutinelydischarged(p 0.001).Sepsis(exp(b) 1.18, p 0.04),respiratorydisease(exp(b) 1.08,p 0.021),hemorrhage(exp(b) 1.22, p 0.001)andrenalfailure(exp(b) 1.15,p 0.003) werecomplicationsassociated withhighercosts.Highercostswerealsosignicantlyassociatedwithmediumhospitalbedsize,age 1year,admissionfromanotherfacility,longerlengthsofstay,West region,morerecenttransplantsandhigherpatientincome,withlowercostsfortransplantsoccurringinachildren'sunitofageneralhospital. CONCLUSIONS: Multiple patientandhospitalcharacteristicsappeartosignicantlypredictinpatientmortalityandcostsinpediatrichearttransplantrecipients. PSU7 ADISCRETEEVENTSIMULATIONMODELTOESTIMATETHELONGTERM OUTCOMESOFBARIATRICSURGERYINMEXICOZanelaOO1,CabraHA1,AnayaP2,RodriguezS1,MelendezG2 1Johnson&JohnsonMedical,MexicoCity,D.F.,Mexico,2Fundaci—nMexicanaparalaSaludAC, MexicoCity,MexicoOBJECTIVES: Estimatethereturnofinvestment(ROI)onbariatricsurgery vs .conventional,non-surgicalapproachastreatmentformorbidobesityfromtheMexicanpublichealthsystemperspectiveinthelongterm. METHODS: Theindividual experienceofamorbidly-obesepatientwasassessedusingadiscreteeventsimulationmodelbuiltinArena.Patientswerecreatedwithunique,randomlyassignedclinicalandepidemiologiccharacteristics,clonedandsenttoeitherbariatricsurgery(BS)orconventionaltreatmentpharmacologictreatmentofassociated comorbiditiesandlifestylemodications(controlarm).Evaluatedcomorbidities weretype-2diabetes,hypertensionandhypercholesterolemia.Preoperativeprevalencesandup-toyear2recoveryratesweretakenfrompublishedmeta-analyses. 2-and10-yearprevalenceswerederivedfromincidenceandrecoveryratesshown inSOSstudy;in-betweenprevalenceswereinterpolatedassumingexponential growth,thusallowingclinicalstateworseninginbotharms.Additionalassumptionsincludeinfrastructurerestrictions,noperioperativecomplicationsandshort termmortality.5%ofpatientsincontrolgroupwereallowedtohavesurgeryafter year5.Consideredcostsincludedthebariatricprocedureandcomorbidity-specic pharmacologictreatment,takenfrompublichealthinstitution'sDRGs.Simulation wasrunwith150patientsfor10yearsand10iterationsusinga4.5%annualdiscountrate.Resultsareshowninyearsand2010ination-adjustedMXP.95%condenceintervalswereestimated. RESULTS: 10-yearaccumulatedcostforaBS patientwas$125,902($125,041-$126,763),and$259,413($258,098-$260,728)fora controlpatient.ROIonBSwasachievedonyear6.94(6.88-7.00).Costdifferencesare duetothereducedresourceutilizationaftersurgeryresultingfromresolutionof comorbidities. CONCLUSIONS: InvestmentinBSoffsetsitscostandisrecouped withinareasonabletime,thusallowinginstitutionstoreducetheburdenimposed bymorbidobesity.Long-rundataforotherassociatedcomorbiditiesisneeded,as theirinclusionintheanalysiscoulddecreaseROI. PSU8 PATTERNSOFACUTECLOPIDOGRELUSEANDINPATIENTCOSTSINACUTE CORONARYSYNDROMEPATIENTSUNDERGOINGCORONARYARTERYBYPASS GRAFTINGJohnstonSS1,BellK2,GdovinBergesonJ2,McMorrowD3,GrahamJ2 1ThomsonReuters,Washington,DC,USA,2Bristol-MyersSquibb,Plainsboro,NJ,USA,3Thomson Reuters,Cambridge,MA,USAOBJECTIVES: Treatmentguidelinesstipulatethatclopidogrelshouldbeinterrupted 5dayspriortoelectivecoronaryarterybypassgrafting(CABG)toreducetherisk ofbleeding.However,ifurgentCABGisindicated,experiencedsurgeonsmayperformCABGin 5days.Thisstudyisthersttodescribepatternsofacuteclopidogreluseandinpatientcostsamongacutecoronarysyndrome(ACS)patients undergoingCABG,whichwillhelpdecisionmakersunderstandtherelativebenets/costsofantiplatelettherapiesinreal-worldpractices. METHODS: Thestudy usedtheMarketScan¨Commercial,Medicare,andHospitalDrugDatabases,comprisingadministrativehealthcaredataforover63millionindividuals.ACSepisodes,denedashospitalizationsforACS(primaryICD-9-CMdiagnosis410.xx, 411.1x),occurringbetween1/1/2005-6/30/2009wereidentiedfrompatientsaged 18years.Outcomesincludedcostsandlengthsofstay(LOS)ofACSepisodeswith CABGand,amongclopidogrel-treatedpatients,numberofdaysbetweentheday CABGwasperformedandthelastclopidogreldoseadministered.Analyseswere descriptive. RESULTS: 160,168ACSepisodeswereidentied;meanpatientage 63.5years.CABGepisodescomprised9.3%(14,896/160,168)ofACSepisodes.The meanLOSwas9.8[SD6.8]daysperCABGepisode.Meaninpatientcostswere $71,140[SD$68,012]perCABGepisode.Amongclopidogrel-treatedpatientswith inpatientdrugdatawhounderwentCABG(n 8,101),themeandaysbetweenthe dayCABGwasperformedandthelastdoseofclopidogreladministeredwas3.3[SD 2.6]daysandthemajority(62.1%)underwentsurgerywithin2-3daysaftertheirlast clopidogreldose.Themeanincrementalincreaseininpatientcostsassociatedwith 1extraLOSdaywas$1,950. CONCLUSIONS: Datasuggestthatsurgeonscommonly performCABGwithin 5daysafterclopidogreladministration.However,among patientsforwhomurgentCABGisnotindicated,withholdingCABGmayonly minimallyaffectinpatientcostsandmustbeconsideredinthebroadercontextof patientmanagement. PSU9 ECONOMICEVALUATIONOFTHECURRENTTRENDTOWARDSMORE UNCEMENTEDFIXATIONINPRIMARYHIPARTHROPLASTIESANDTHE POTENTIALIMPACTOFCHANGESOFTHEDEVELOPMENTINENGLANDAND WALESKontekakisA,GaiserS HeraeusMedicalGmbH,Wehrheim,GermanyOBJECTIVES: Thepurposeofthestudywastoquantifythepotentialeffectofthe currenttrendtowardsmoreuncementedprimarytotalhiparthroplasties(THAs)in termsofhospitalcostsandnumberofrevisionsovera10yearperiod(2011to2021) inEnglandandWales.Inadditionthepotentialimpactofchangesofthisdevelopmentwasevaluated. METHODS: RegistrydatafromtheNationalJointRegistryfor EnglandandWalesfrom2004to2009wasusedtopredictthenumbersofTHAsandA87VALUEINHEALTH14(2011)A1–A214

PAGE 96

thetrendsforcementedanduncementedxationtechniqueuntil2021bylinear regression.AdynamicMarkovmodelwithacyclelengthofoneyearandtime frameof10yearswasdevelopedtocalculatethenumberofrevisionsandcosts fromahospitalresource-useperspective.Patientsenterthemodelafterprimary THAandcanprogresstosepticorasepticrevisionanddeath.Thetransmission probabilitieswerederivedfromScandinavianregistrystudies.Hospitalcostswere takenfrompublishedliteratureaboutresourceutilizationinEnglandandwere discountedby3%annually. RESULTS: Anannualincreaseof3,960THAs(R-square, 0,87)wasobservedduringtheperiod.TheshareofuncementedTHAsincreased from20,2%in2004to42,2%in2008withanannualincreaseof3,707uncemented THAs(R-square,0,99).TheproportionofcementedTHAsdecreasedsteadily.A continuationofthistrendwouldleadtohospitalcostsof£7,246millionand16193 revisionsovera10yearperiod.Areversalofthistrendwouldleadtocostsof£7,176 millionand14980revisions,resultinginsavingsof£69,8millionandtheavoidance 1212revisions. CONCLUSIONS: Theresultssuggestthatthecurrenttrendtowards uncementedxationofTHAsisunjustied.Ashifttowardscementedxation techniqueshouldbeconsidered. PSU10 COST-EFFECTIVENESSOFPRIMARYPCIWITHOUTONSITESURGICALBACKUPRamosP UniversityofLouisville,Louisville,KY,USAOBJECTIVES: TheAmericanCollegeofCardiologistsandtheAmericanHeartAssociationstronglyrecommendthatahospitalperformingPCImustalsohavecoronaryarterybypassgraftcapabilities.Followingtheserecommendations,thestate ofKentuckyhaslimitedthenumberofhospitalsallowedtoperformPCIand therebylimitingaccesstothislife-savingprocedure.Recently,thestateofKentuckyevaluatedifhospitalswithoutsuchcapabilitiesshouldbeallowedtoperform primaryPCI.TheresultingdataallowedtheestablishmentofthemedicalsoundnessofallowingsuchhospitalstoperformprimaryPCI.Thecurrentstudyaimsto evaluatethenancialfeasibilityofallowingthesehospitalstodoemergencyPCIin additiontohospitalswithonsiteopen-heartsurgerycapabilities. METHODS: Estimateshavebeenderivedfromasystematicliteraturereviewofnationalstudies basedonPCIregistriesaswellasourearlierstudyKENTUCKYPILOTPROJECTFOR PRIMARYPCIWITHOUTONSITECABG .CostsestimateswerederivedfromtheNationalInpatientSample.Indeterminingcosts,theobservationswereextractedby lteringusingICD-9codesusingSAS.Adeterministicmodel,implementedinMicrosoftExcel,wasdevelopedsothatmoreuncertaintywouldnotbeintroduced. Theevaluationestimatedtheincrementalcosteffectivenessratio(ICER)ofallowingregionalhospitalstoperformprimaryPCIfromapayer'sperspective.Uncertaintyaboutthemodelparameterswasinvestigatedthroughsensitivityanalysis. RESULTS: Thestudyfoundthattherewerenostatisticallysignicantdifferencesin outcomesbetweenhospitalswithandwithoutCABGcapabilities.Theonlycharacteristic,whichwassignicantlydifferentbetweenthesetwogroups,wastotal charges.Thealternativetoallow RegionalHospitalsaswell toperformprimaryPCI dominatedtheotheralternativeof OnlyAllowingHospitalswithOnsiteCABG toperformPCI. CONCLUSIONS: States,suchasKentucky,thatstrictlyfollowtheAHA/ ACCrecommendationsaboutPCIshouldconsiderthecost-effectivenessevidence whenmakingpublicpolicy. PSU11 REDUCTIONINHEALTHCAREUTILIZATIONANDEXPENDITURESINPATIENTS TREATEDWITHCATHETERABLATIONFORATRIALFIBRILLATION(AF)INTHE UNITEDSTATESLadapoJA1,DavidG2,GunnarssonC3,HaoS4,WhiteS5,MarchJ6,ZangM5,ReynoldsM1 1BethIsraelDeaconessMedicalCenter,Boston,MA,USA,2UniversityofPennsylvania, Philadelphia,PA,USA,3S2StatisticalSolutions,Inc.,Cincinnati,OH,USA,4SutterPacicMedical Foundation,SanFrancisco,CA,USA,5BioSenseWebster,DiamondBar,CA,USA,6BioSense Webster,Boston,MA,USAOBJECTIVES: Toestimatetheimpactofradiofrequencyablation(RFA)onshortand long-termhealthcareutilizationandexpendituresingeneralandMedicarepopulationswithAF. METHODS: WeanalyzeddatafromtheUSMarketScan¨CommercialClaimsandEncountersDatabaseandMedicareSupplementalDatabasefrom ThomsonReuters.Thedatabasesarecomprisedofde-identiedpatient-levelrecordsfromemployer-sponsoredandpublichealthinsuranceplans.Weuseda codingalgorithmtoidentify3,194patientswithanRFAprocedurefortreatmentof AF,whohadcontinuousenrollmentinthedatabase6monthspriortotheirrst ablation,andaminimumofoneyearfollow-uppostablation.Multivariableregressionmodelsforutilizationandexpenditureswerebuiltforallpatients,andsubanalyseswereperformedforpatients65-yearsandover.Resultsarereportedfor vesamples,basedondurationofavailablefollow-up:12months,18months,2 years,2.5years,and3yearspostablation. RESULTS: Comparedtothesixmonths priortoRFA,thereweresignicantreductionsinthenumberofoutpatientappointments,hospitalinpatientdays,andemergencyroomvisitsinthetotalstudypopulationaswellasthesubsetofpatients65andolder.Therewasastatistically signicant(p .01)decreaseintotalexpenditures(insurerandpatientout-ofpocketexpenditures)acrossallvetimeperiods,withaveragesavingsofapproximately$5,000overeachsixmonthperiodafterablation($800permonth).For patients65andover,theaveragesavingswereapproximately$700permonth. Drugutilizationalsosignicantlydeclined(p .05),withaveragesavingsin medicationexpendituresrangingfrom$317to$578per6monthperiod.The reductionindrugutilizationwasdurableandsignicantinpatients65andover. CONCLUSIONS: RFAforAFreducesreal-worldhealthcareutilizationandexpendituresupto3yearspostablation.Thisreductionwasconsistent,signicant,andhas implicationsforbothgeneralandMedicarepopulations. Surgery–Patient-ReportedOutcomes&Preference-BasedStudies PSU12 IMMUNOSUPPRESSANTTHERAPYADHERENCEANDREJECTIONOUTCOMESIN LIVERANDKIDNEYTRANSPLANTPATIENTSGorevskiE,MinkaraA,BainB,MogilishettyG,MartinBooneJ UniversityofCincinnati,Cincinnati,OH,USAOBJECTIVES: Non-adherencetomedicationregimensafterorgantransplantation isamajorriskfactorforacuterejectionandgraftloss,however,thereislimited dataquantifyingtheassociatedrisk.Thisstudyexaminestheassociationbetween adherencetoimmunosuppressanttherapyandrejectioninliverandkidneyorgan transplantrecipients. METHODS: Thisisarandomized,crosssectionalanalysis. AdherencetoimmunosuppressantswasassessedbyadministeringtheImmunosuppresiveTherapyAdherenceScale(ITAS)topatientsatthetimeoftheirclinic appointment.Rejectionwasdenedasbiopsyprovenrejection.Thenumberof biopsies,numberofrejections,rejectionseverityandlengthofhospitalstaydueto rejectionwerecollected.Immunosuppressantregimens,levels,numberofdose changes,andnumberoftablets/capsulesperdaywereobtained.Logisticregressionmodelwasusedtodeterminetheassociationbetweenrejectionandadherence,genderandnumberoftablets/capsulesconsumedperday. RESULTS: Atotal of67patientswereincludedinthisstudy,fromwhich21(31%)weremaleswithan averageageof51years,SD 11.6.Therewere49(73.1%)kidneyrecipients,14 (20.9%)liverrecipientsand4(5.9%)combination.Atleastonerejectionwasobservedin26(38.8%)ofthepatients,fromwhich18(69.2%),6(23.1%)and2(7.7%) werekidney,liverandcombination,respectively.Atotalof35(52.3%)ofthepatientswerefoundtobenon-adherentwiththeirmedications,fromwhich18 (51.4%)hadatleastonerejection.Patientsweretakinganaverageof18tablets/ capsulesperday.Thelogisticregressionshowedthatgenderandtablets/capsules consumedperdayhadnoimpactonrejection,however,non-adherenceissignificantlyassociatedwithrejection(OR 0.264,95%CIis0.089-0.784).Nonadherent patientare74%morelikelytohavearejectioncomparedtoadherentpatients. CONCLUSIONS: Non-adherencetoimmunosuppressivemedicationresultsinincreasedrejectionratesinkidneyandlivertransplantrecipients. PSU13 MORBIDOBESITYANDHEALTHRELATEDQUALITYOFLIFETayyemR1,AliA2,AtkinsonJ3,MartinCR4 1WhittingtonHospital,London,UK,2TheAyrHospital,Ayr,UK,3UniversityoftheWestof Scotland,Paisley,UK,4UniversityoftheWestofScotland,Ayr,UKOBJECTIVES: Signicanthealthrisksimpactonindividualswithmorbidobesity includingdeleteriouseffectsonhealthrelatedqualityoflife(HRQOL).Amonocentricprospectivestudywasconductedtoevaluatetheimpactofbariatricsurgery andmorbidobesityonHRQOL. METHODS: Datawerecollectedfrom236participants:83patientswhowereunderconsiderationforbariatricsurgery,68patients whoalreadyhavehadabariatricprocedureand85volunteers.Participantswere askedtocompleteBariatricandObesitySpecicSurvey(BOSS),ShortFormHealth survey(SF-36),HospitalAnxietyandDepression(HADS)scale,Moorehead-Ardelt QualityoflifeQuestionnaire(M-AQoLQII),andademographicdatasheet. RESULTS: ObesepeopledisplayedlowerlevelsofHRQOLcomparedtonon-obese peoplewithanalmostlinearnegativerelationshipbetweenbodymassindex(BMI) andHRQOL.ObesefemaleswithBMIbetween30and40weremoresusceptibleto theadverseeffectsofobesitycomparedtomales,particularlyinthedomainsof "Incapacity"and"SexualHealth".TherewasnostatisticallysignicantHRQOL difference,however,betweenmenandwomeninthenonobese(BMI 30)and morbidobesity(BMI 40)."AppearanceandHealth"and"SexualHealth"domains werecomparablebetweenpatientswhohadsurgeryandthecontrolgroup. CONCLUSIONS: Femalesaremoresusceptibletotheadverseeffectsofweight comparedtomalesuptocertainweightlevelafterwhichgenderequityinadversity becomesthenorm.Thedomainsof"AppearanceandHealth"and"SexualHealth" arethequickestdomainsthatwillrecoverfromthenegativeimpactofobesity followingbariatricsurgery. PSU14 HEALTH-RELATEDQUALITYOFLIFEFORPATIENTSONKIDNEYTRANSPLANT WAITINGLIST-WHATFACTORSREALLYMATTER?OngSC1,ChowWL1,vanderErfS1,JoshiVD1,TeePS2,LuYM2,KeeTYS2 1SingaporeHealthServices,Singapore,Singapore,2SingaporeGeneralHospital,SingaporeOBJECTIVES: Studiesonhealth-relatedqualityoflife(HRQoL)ofpatientsonkidney transplantwaitinglistremainedscarcedespitethefactthatitcanpredictfuture morbidityandmortality.Theaimsofthe studywerethereforetomeasuretheHRQoL ofpatientsonkidneytransplantwaitinglistcomparedwiththatofthegeneralpopulation,toidentifynon-diseasefactorswhichcouldimpactontheHRQoLscoresandto evaluatethereliabilityoftheSF-36inthisgroupofpatients. METHODS: Thiswasa cross-sectionalstudyofkidneytransplantwaitinglistpatientsmanagedatatertiaryrenalunitinSingaporeusingtheSF-36.ASF-36normativecalculatorgeneratedHRQoLscoresforamatchedcohortoftheSingaporegeneralpopulation.ReliabilityofSF-36wasassessedbyCronbach'salphacoefcients. RESULTS: There were265respondents.OurstudyshowsthatHRQoLscoresforpatientsonkidney transplantwaitinglistwerelowerthanthepopulationnormsacrossallsubscales andwereclinicallysignicantforGeneralHealth,RolePhysical,BodilyPain,Social FunctioningandMentalComponentSummaryscores.Non-diseasefactorslike beingChinese,married,employed,undergoinghaemodialysisandnotbeingofferedalivingkidneydonationpredictedbetterHRQoLscoresafteradjustingfor possibleconfounders.Age,gender,educationallevel,householdincome,historyof kidneytransplant,durationontransplantwaitinglistandyearsondialysisdidnot signicantlyinuenceSF-36acrossallsubscalesscores.Sevenof8subscalesinA88VALUEINHEALTH14(2011)A1–A214

PAGE 97

SF-36hadacceptablereliabilityofCronbachalphavaluesof 0.7. CONCLUSIONS: PatientsonkidneytransplantwaitinglisthadworseHRQoLthanthegeneralpopulation.MoreresearchcouldbedoneintoreasonsforpoorerHRQoLamongthe at-riskpatients.SF-36wasareliabletoolinassessingtheHRQoLofpatientson kidneytransplantwaitinglistinSingapore. PSU15 DEVELOPMENTOFTHEBARIATRICANDOBESITYSPECIFICSURVEY(BOSS)TayyemR1,AliA2,AtkinsonJ3,MartinCR4 1WhittingtonHospital,London,UK,2TheAyrHospital,Ayr,UK,3UniversityoftheWestof Scotland,Paisley,UK,4UniversityoftheWestofScotland,Ayr,UKOBJECTIVES: Thereisalackofapsychometricallyrobustbariatric-specichealth relatedqualityoflife(HRQOL)tool.Amono-centricprospectivestudywasconductedtodevelopandvalidateanewbariatricspecic81-itemselfreportHRQOL instrumentcalledtheBariatricandObesitySurgerySurvey(BOSS). METHODS: Datawerecollectedfrom236participants:83patientswhowereunderconsiderationforbariatricsurgery,68patientswhoalreadyhavehadabariatricprocedure and85volunteers.ParticipantswerealsorequiredtocompletetheShortForm Healthsurvey(SF-36),HospitalAnxietyandDepression(HADS)scale,MooreheadArdeltQualityoflifeQuestionnaire(M-AQoLQII),andademographicdatasheet. Twoweeksfollowingthecompletionofthese5questionnaires,participantswere askedtocompleteBOSSoncemorealongwithafeedbacksheet. RESULTS: Exploratoryfactoranalysisrevealedamultidimensionalinstrumentconsistingof42 itemsdistributedover6domainsthataddressvariousHRQOLaspectsanddimensionspertinenttobariatricsurgery,andrelevanttomorbidlyobesepatients.FurtherpsychometricanalysisshowedthatBOSShasadequateinternalconsistency reliability(Cronbach 0.970),testre-testreliability(ICC 0.926),constructvalidity,criterionvalidity,facevalidityandacceptability. CONCLUSIONS: BOSSisa validandreliablemultidimensionalinstrumentthatprovidesaclinicallyuseful andrelevantmeasuretoassessHRQoLinpatientsundergoingbariatricsurgery. Surgery–HeathCareUse&PolicyStudies PSU16 LAPAROSCOPICSUPRACERVICALHYSTERECTOMYVERSUS LAPAROSCOPIC-ASSISTEDVAGINALHYSTERECTOMY:POST-SURGICAL OUTCOMESANDCOSTSWatersH1,SongX2,PanK2,SubramanianD1,SedgleyR3,RaffGJ4 1Ethicon,Inc.,Somerville,NJ,USA,2ThomsonReuters,Cambridge,MA,USA,3ThomsonReuters, Washington,DC,USA,4IndianaUniversitySchoolofMedicine,Indianapolis,IN,USAOBJECTIVES: Tocomparetheincidenceofpost-operativecomplications,healthcareutilization,andcostsinlaparoscopicsupracervicalhysterectomy(LSH)versus laparoscopic-assistedvaginalhysterectomy(LAVH)patients. METHODS: Women 18yearswithLSHorLAVHfrom1/1/2007-9/30/2008wereidentiedinthe ThomsonReutersMarketScan¨CommercialClaimsandEncounterDatabase.Patientswererequiredtohave6monthsofcontinuousmedicalandprescription coveragepriorandsubsequenttothehysterectomydate.Patientswereexcludedif theyhadadiagnosisofcancer,hadindexdateprocedurecodesforbothLSHand LAVH,orifthelengthofstayassociatedwiththeindexprocedureexceeded20 days.Post-operativeoutcomesandgynecologic-related(GYN-related)utilization andcostsweremeasured30and180dayspost-surgery.Multivariateanalysiswas conductedtocomparepost-surgicaloutcomesandcostsbetweenpatientswith LSHandLAVHcontrollingfordemographicandclinicalcharacteristics. RESULTS: Atotalof6,198LSHpatientsand14,181LAVHpatientsmetthestudycriteria. ComparedwiththeLAVHcohort,LSHpatientsweremorelikelytohavedysfunctionaluterinebleeding(32.6%vs.27.9%)andleiomyomas(38.0%vs.26.3%)astheir primarydiagnosisandlesslikelytohaveendometriosis(9.3%vs.10.4%)andprolapse(1.5%vs.8.0%),p 0.01inallcases;hadsignicantlyloweroverallinfection rates(6.2%vs.7.4%,p 0.002);andhad signicantlylowertotalGYN-relatedcosts ($252vs.$385,p 0.001,30dayspost-surgery;$350vs.$569,p 0.001,180dayspostsurgery).Aftermultivariateadjustmentforpatientcharacteristics,totalcostswere estimatedtobe$108and$174lowerforLSHpatientsthanforLAVHpatientswithin30 and180daysoffollowup,respectively(p 0.001).LSHpatientshadsignicantlylower hazardsofdevelopinginfection(hazardratio[HR] 0.830),hematologiccomplication (HR 0.667),andanalgesicuse(HR 0.812). CONCLUSIONS: LSHpatientsdemonstratedfewerpost-operativeinfectionsandlowerGYN-relatedcostscomparedto LAVHpatients. PSU17 EXCESSPAYMENTSFROMMEDICAREFORINPATIENTSURGERYBaserO STATinMEDResearch,AnnArbor,MI,USAOBJECTIVES: ToexaminethevariationinoutlierpaymentsacrossU.S.hospitals andtheextenttowhichthevariationisexplainedbypatientandhospitalfactors, includingqualityofcare. METHODS: UsingtheNationalMedicareClaimsdatabase for2002,weexaminedoutlierpaymentsinpatientsundergoingcoronaryartery bypassgrafting(CABG)(n 165,226),lowerextremitybypasssurgery(n 43,886)and colectomy(n 101,345).Wethencategorizedhospitalsperformingtheseproceduresaccordingtotheiroutlierpaymentrates.Usingmultiplelogisticregression, weexploredtherelationshipsbetweenhospitaloutlierpaymentrates,patientcase mixandhospitalquality,asreectedbyrisk-adjustedmortalityrates. RESULTS: Forallthreeprocedures,theproportionofpatientsassociatedwithoutlierpaymentsvariedfrom10%(colectomy)to14%(CABG).Averageoutlierpaymentswere considerable,rangingfrom$18,000toover$24,000perpatient.Themostconsistent riskfactorsforoutlierpaymentsincludedraceandadmissionacuity.Higherhospitalandsurgeonvolumesandteachingstatuswereassociatedwithlowerratesof outlierpayments.Therewasanegativecorrelationbetweenrisk-adjustedmortalityratesandoutlierpayments.Forallthreeprocedures,theproportionofoutlier paymentsvariedwidelyacrosshospitalsfromlessthan5%togreaterthan20%. Measurablepatientandhospitalfactorsexplainedasmallproportionofvariation acrosshospitals. CONCLUSIONS: Outlierpaymentsareanimportantcomponentof medicalcostsforinpatientsurgery.Althoughexplainedinpartbyquality,the reasonsforawidevariationinoutlierpaymentsacrosshospitalsremaintobe claried. PSU18 IMPACTOFHEALTHINSURANCEONRECEIVINGBREASTCONSERVING SURGERYWITHRADIATIONINFLORIDAAliAA,XiaoH FloridaA&MUniversity,Tallhassee,FL,USAOBJECTIVES: 1.Examinetheimpactofinsuranceontreatmentoflocalized breastcancerusingBreastConservingSurgery(BCS)withradiation.2.Identify factorsthatcontributetowomen'sreceivingbreastconservingsurgerywith radiation. METHODS: Breastcancercasesdiagnosedduring1997-2002wereobtainedfromtheFloridaCancerDataSystem.Womenaged40andabovewithlocalizedbreastcancerwereincluded.Demographic,insurance,andtreatmentinformationwereextractedandlinkedwith2000Censusdatatogettract-level informationoneducationandpovertylevel.Multi-levellogisticregressionanalysis wasconductedtodeterminefactorsthathavecontributedtoBCSwithradiation treatment. RESULTS: 41,508womenwerediagnosedwithlocalizedbreastcancerin Floridaduring1997-2002.ThestudyfoundthatBCSwithoutradiationandmastectomywerethetwomajortreatmentsforlocalizedbreastcancer.Theaverageageof thewomenwas66yearswith8.5%ofthemreceivingBCSwithradiation.Women withthefollowingcharacteristicsweremorelikelytoreceiveBCSwithradiation thantheircounterparts:havingprivateorMedicareinsurance,beingmarried,livinginneighborhoodwithhigherpercentageofhighschooleducation,andbeing recentlydiagnosed. CONCLUSIONS: AlthoughBCSwithradiationisrecommended totreatwomenwithlocalizedbreastcancerbyclinicalpracticeguidelinesandthe useofthistreatmenthassignicantlyincreasedovertime,therearestilldifferencesinreceivingthetreatmentamongwomenwithdifferenthealthinsurance andmaritalstatus.Possiblereasonsforthedifferencesrequirefurtherresearch. PSU19 NATIONWIDEUTILIZATIONPATTERNOFWHOLEBRAINRADIOTHERAPYAND STEREOTACTICRADIOSURGERYFORBRAINMETASTASISJungSY,HwangJS,ChoiJE,ShinSY,KimJH,HyunMK,KimYE,BaeJM NationalEvidence-basedHealthcareCollaboratingAgency(NECA),Seoul,SouthKoreaOBJECTIVES: Toassessthepatternofradiotherapyandradiosurgeryfornewly diagnosedandrecurredbrainmetastasis. METHODS: UsingtheKoreanHealth InsuranceReview&AssessmentService(HIRA)claimsdatabase,patientswhoaged 20ormore,diagnosedasbrainmetastasis(ICD10 C79.3)duringJanuary1st,2006 andJune30th,2008,treatedwholebrainradiotherapy(WBRT)orstereotacticradiosurgery(SRS),andwithouthistoryofbrainmetastasisdiagnosisortreatment within6monthspriortoindextreatmentwereidentied.Withapermissiblegapof 30days,eachtreatmentepisodesweredened.Theepisodeswerecategorizedas thosefornewlydiagnosedandrecurrentbrainmetastasispatients.Characteristics ofpatients,frequencyofWBRT/SRSasinitialandrecurrencetreatment,andthe meannumberofre-treatmentwereanalyzed. RESULTS: Atotalof7,449newly treatedpatientsand2,008recurrentpatientssatisedtheselectioncriteria.Among newpatients,4,797(64.4%)treatedWBRT,1,439(19.3%)treatedSRS,and1,213 (16.3%)treatedWBRTwithSRS.Mostfrequentprimarytumorwaslungcancer (57.2%).Recurrencerateofthenewpatientswas27.0%.Meannumberofre-treatmentofWBRTorSRSwas1.38.AmongpatientswhotreatedWBRTandSRSas initialtreatments,69.8%and50.2%treatedthesametreatmentasrsttreatment forrecurrence,respectively. CONCLUSIONS: Whileclinicalguidelinesforbrain metastasisrecommendsWBRTwithSRSforbrainmetastasispatientswithfavorablefunctionalstatus,theproportionofWBRTwithSRSwasrelativelylow.Further researchtoexplorethereasonofthegapbetweenevidenceandrealpracticeis needed. Surgery–ResearchonMethods PSU20 ISTHEREINDICATIONBIASOFRETROSPECTIVEOBSERVATIONALSTUDY?ChoiJE1,JangEJ1,JungSY1,LeeEJ1,LeeNR1,JooCK2 1NationalEvidence-basedHealthcareCollaboratingAgency(NECA),Seoul,SouthKorea,2Catholic UniversitySchoolofMedicine,Seoul,SouthKoreaOBJECTIVES: Observationalstudyisdifculttointerpretduetoanumberofmethodologicalissues;cohorteffects,thelearningcurveofthehealthcareproviderand confoundingbyindication:ThislargescaleretrospectivestudyisaimedtoascertainthatthereareindicationbiasofLASIKsurgerybetweenlocalhospitaland tertiaryhospitalinSouthKorea. METHODS: Aretrospectivecohortforconsecutive 3,401eyesreceivedLASIKin6multicenterincludingtertiaryhospitalandlocal hospitalfrom2002to2005wereregistered.Sociodemographic,operationcharacteristics,preoperativebaselinemanifestrefractions(MR)andadditionalfollowup dataat3month,6monthand12monthaftersurgerywerecollected.Weused independentt-testtocomparepreoperativedifferencebetweentertiaryhospital andlocalhospitalandgeneralizedlinearmixedmodelwithrandominterceptto testcharacteristicofhospitalandfollowuptimemightinuencetoMRafter surgery. RESULTS: ThebaselineMR(mean SD)oftertiaryhospitalwas 5.24 2.07 andthatoflocalhospitalwas 4.27 1.71(p 0.001).ThemeanMRat3month,6 monthand12monthoftertiaryhospitalwere 0.46 0.70, 0.55 0.69andA89VALUEINHEALTH14(2011)A1–A214

PAGE 98

0.63 0.67increasinglyaccordingtothefollowupduration.ThemeanMRat3 month,6monthand12monthoflocalhospitalwere 0.48 0.62, 0.43 0.52and 0.47 0.47.AgeneralizedlinearmixedmodelwitharandominterceptforMR changerevealedthatthehighermanifestrefractionwassignicantlyinuencedby thelongerfollowuptime(p 0.001)butthehospitalcharacteristicwasnotinuencetoMR(p 0.319).Nosignicantinteractionbetweenfollowupdurationand levelofmedicalinstitutionwasdetected. CONCLUSIONS: Althoughobservational studymightbeconfoundedbyindicationbias,noevidenceofsuchbiaswasdetected.Itmayreectrealworldandassessmentofindicationbiasstillprovideus withbenetsonclinicallyrelevantinterpretationaboutcurrentexperienceofmedicalinterventions. PSU21 THEREUSEOFSINGLE-USESURGICALDEVICES:EVALUATIONOFTHE ECONOMICEVIDENCEEmmermannA1,SchroeerP1,HamerslagL2,KuselJ2 1EthiconEndo-Surgery(Europe)GmbH,Norderstedt,Germany,2CostelloMedicalConsultingLtd, Cambridge,UKOBJECTIVES: Thereprocessingandreuseofsingle-usesurgicalinstrumentsisa relativelycommonpracticeinhospitalsinNorthAmericaandEurope.Guidelines forthismattervarywidelybetweencountries,whichhasledtoareviewofthe issuebytheEuropeanCommission.Thepotentialrisksofthispracticedueto decreasedfunctionalityandsafetyofdevicesareassumedtobeoutweighedby economicadvantages.Theaimofthisstudywastoassessthebreadthandquality ofevidencefortheeconomicoutcomesofmultipleuseofdisposablesurgical devices. METHODS: Tobuilduponaprevioussystematicreviewthatsearched literatureonthistopicuptoJuly2007,additionalliteraturesearcheswereconductedacrossmultipleelectronicdatabasesuptoJune2010. RESULTS: Twopublishedstudiesandoneunpublishedstudywereidentiedinadditiontothenine studiesidentiedbytheprevioussystematicreviewbyJacobsetal.Themedian savingaffordedbyreusecomparedtosingleuseofdisposablesurgicaldeviceswas 40%,butoutcomesrangedwidelybetween 15%and92%.Studyqualitywasgenerallypoor.Onestudywasarandomizedcontrolledtrial,whereastheotherswere observationalstudies.Fewstudiesincludedastatisticalanalysisandonlytwo studiesevaluatedadverseeventcostsinadditiontothoseofacquisitionand cleaning. CONCLUSIONS: Highqualitystudiesassessingthecost-effectivenessof reusingsingle-usesurgicaldevicesarerare,partlyduetodifcultiesindatacollection.Theeconomicevidencejustifyingthisprocedureisthereforeinconclusive, anditsvaluefordecisionmakingshouldbelow.TheEuropeanCommissionhave highlightedthelackofeconomicevidenceasamajorbarriertopolicydecisionmakingintheirreportfromAugust2010.Thereisaneedforlarger,longer-term studiesadequatelyassessingalleconomicaspectsofreprocessingdisposabledevices,includingconsiderationofthecompromisedsafetyandfunctionalityofdevices. PSU22 APATIENT-BASEDQUESTIONANIRETOASSESSOUTCOMESOFFOOTSURGERY: VALIDATIONINTHECONTEXTOFSURGERYFORHALLUXVALGUSDawsonJ1,JenkinsonC1,ChurchmanD2 1UniversityofOxford,HealthServicesResearchUnit,Oxford,Oxfordshire,UK,2IsisInnovation Ltd.,Oxford,Oxfordshire,UKOBJECTIVES: Todevelopandvalidateaquestionnaireforcompletionbypatients havingfootsurgery. METHODS: Weevaluatedliteratureforexistingfootpainand disabilityquestionnairesandshort-listedtheManchesterFootPainandDisability Questionnaireforitssuitabilityasthebasisofanoutcomesmeasureforfootsurgery.Thiswasthenassessedinthecontextofhalluxvalguscorrectivesurgery. Interviewswithpatientsledtoinitialchanges,resultingin20candidatequestionnaireitemswithveresponsecategorieseach.Theseweretestedinaprospective studyof100patients(representing138footoperations)undergoinghalluxvalgus correctivesurgery. RESULTS: Analysisofunderlyingfactorstructure,dimensionality,internalreliability,constructvalidityandresponsivenessofthequestionnaireitemsinrelationto1)SF-36generalhealthsurvey,and2)AmericanOrthopaedicFoot&AnkleSociety(AOFAS)halluxclinicalscaleledtotheexclusionof someitems(2itemsdemonstratingooreffectsandafurthertwoitemsthatdemonstratedpoorfactorstructure)andresultedinanal16itemquestionnaire,the 'Manchester-OxfordFootQuestionnaire'(MOXFQ).TheMOXFQconsistsofthree domains/scales:Walking/standing'(sevenitems),Pain'(veitems)andSocial interaction'(fouritems)eachhavinggoodmeasurementproperties.Allthreedomainswereunidimensional. CONCLUSIONS: Thenew16-itemMOXFQhasgood measurementpropertiesassessedinthecontextofoutcomesassessmentofsurgeryforhalluxvalgus.FuturestudiesshouldassesstheMOXFQinthecontextof surgeryforarangeofotherfootandankleconditions. PSU23 ASSESSINGTHEOUTCOMEOFSHOULDERSURGERY:DEVELOPMENT, VALIDATIONANDRESPONSIVENESSOFTHEPATIENT-REPORTEDOXFORD SHOULDERSCORE(OSS)DawsonJ1,FitzpatrickR1,CarrAJ2,ChurchmanD3 1UniversityofOxford,HealthServicesResearchUnit,Oxford,Oxfordshire,UK,2Universityof Oxford,NufeldDepartmentofOrthopaedics,RheumatologyandMusculoskeletalSciences, Oxford,Oxfordshire,UK,3IsisInnovationLtd.,Oxford,Oxfordshire,UKOBJECTIVES: Todevelopandvalidateaquestionnaireforcompletionbypatients havingshoulder(otherthanstabilisation)operations. METHODS: APatientReportedOutcome(PRO)measurewascreatedfollowingreviewofavailablePROsand afterinterviewing20patientsattendinganoutpatient'sshoulderclinic.The drafted22-itemPROwasinitiallytestedinapilotstudywithanewsetof20patients.TheoriginalPROwasmodiedafterthepilotstudywhichwasthentested ontwofurthergroupsofpatientsresultinginthenal12-itemPRO.Aprospective studyof111patientswasundertakenbeforeoperationandatfollow-upsixmonths later.EachpatientcompletedthenewquestionnaireandtheSF36.Somelledin theStanfordHealthAssessmentQuestionnaire(HAQ).Anorthopaedicsurgeon assessedtheConstantclinicalassessmentscore. RESULTS: Thesinglescorederivedfromthedevelopedquestionnairehadahighinternalconsistency(Cronbach'salphaof0.89pre-operativeand0.92post-operative).Reproducibility,examinedbytest-retestreliability,wasfoundtobesatisfactory.Thevalidityofthe questionnairewasestablishedbyobtainingsignicantcorrelationsintheexpected directionwiththeConstantclinicalassessmentscoreandtherelevantscalesofthe SF36andtheHAQ.Sensitivitytochangewasassessedbyanalysingthedifferences betweenthepreoperativescoresandthoseatfollow-up.Changesinscoreswere comparedwiththepatients'responsestopostoperativequestionsabouttheircondition.ThestandardisedeffectsizeforthenewquestionnairecomparedfavourablywiththatfortheSF36andtheHAQ.Thenewquestionnairewasthemost efcientindistinguishingpatientswhosaidthattheirshoulderwasmuchbetter fromallotherpatients. CONCLUSIONS: TheOxfordShoulderScoreprovidesa measureofoutcomeforshoulderoperationswhichisshort,practical,reliable, validandsensitivetoclinicallyimportantchanges. PSU24 AGREEMENTBETWEENPROMISDEPRESSION,SLEEPDISTURBANCE,AND FATIGUECATMEASURESANDSF-36SCORESINACLINICALCOHORTNerenzDR,PietrantoniL,SchultzL,ObeidL,SwartzA,RubinfeldI,VelanovichV HenryFordHospital,Detroit,MI,USAOBJECTIVES: PROMISmeasureshavebeentestedincommunitypopulations,but havebeenmuchlessstudiedinclinicpatientpopulations.WeexaminedthecorrelationsbetweenthePROMISSleepDisturbance,Fatigue,andDepressionmeasuresandscalescoresontheSF-36amongpatientsbeingseeninclinicpriorto surgery. METHODS: Aseriesofpatientsbeingseenforpre-surgicalconsultation wererecruitedfromHenryFordHospitalsurgeryoutpatientclinic.Theycompleted asetofPROMISCATmeasures(SleepDisturbance,Fatigue,Depression)usingthe AssessmentCenter,andalsocompletedtheSF-36surveyaspartoftheregular clinicalcareprocess.SF-36dimensionswerescoredona0-100scale.Spearman correlationcoefcientswerecalculatedastheprimarymeasureofagreementbetweentheSF-36andthePROMISCATmeasurements. RESULTS: TherewasreasonablevariabilityonallthreePROMISmeasures.ThethreePROMISmeasureswere signicantlycorrelatedwitheachother(r 0.38to0.57).AllSF-36domainscores werehighlycorrelatedwiththePROMISSleepandFatigueCATmeasures,except fortheSF-36Role-Emotionalscale.SeveralSF-36domainscoreswerehighlycorrelatedwiththePROMISDepressionCATmeasure.Thecorrelationcoefcients betweenthePROMISDepressionscaleandtheSF-36domainswithwhichitwas signicantlycorrelatedwereessentiallyequivalent.ThePROMISFatigueandSleep DisturbancemeasuresbothhadahigherdegreeofcorrelationwiththeSF-36VitalityscalethanwithotherSF-36domainscales. CONCLUSIONS: Inapopulationof clinicpatientsbeingseenpriortosurgery,scoresonthreePROMISCATmeasures werehighlycorrelatedwithscoresonmostdomainscalesoftheSF-36,providing evidenceforthevalidityofthePROMISmeasuresinthissetting.Theconsistent levelofcorrelationacrossvariousmeasuressuggestedthatallmaybereectinga generalsenseofgoodversuspoorhealthinthiscohortofpre-surgicalpatients. PSU25 IDENTIFYINGIMPROVEMENTOFPRESSUREULCERPREVENTIONATTHE POINT-OF-CAREWITHMULTIPLEMETHODSINTHEDEPARTMENTOFSURGERYPadulaWV1,MishraMK2,WeaverCD3,YilmazT4,SplaineME5 1UniversityofColoradoHealthSciencesCenter,Denver,CO,USA,2Dartmouth-HitchcockMedical Center,Lebanon,NH,USA,3StateUniversityofNewYorkUpstateMedicalUniversity,Syracuse, NY,USA,4StateUniversityofNewYorkStonyBrook,StonyBrook,NY,USA,5TheDartmouth InsituteforHealthPolicyandClinicalPractice,Lebanon,NH,USAOBJECTIVES: Tocompareresultsofregressionandstatisticalprocesscontrolchart analysisforhospital-acquiredpressureulcers(HAPUs).Todeterminewhichanalyticalmodelsarebestinterpretedfortheclinicalmicrosystem. METHODS: Data wereextractedfrombillingdataandpatientrecordsatatertiary-carefacilityinthe USfrom2004-2007forparametersassociatedwithHAPUincidenceincluding:age, gender,medicalhistory,pressureulcerlocation,length-of-stay,andpressureulcer riskassessmentwiththeBradenscale.PrimaryoutcomemeasureswereHAPU incidencerate,daysbetweendiagnosesofHAPUsinaninpatientsurgicalward, andproportionofcompletedBradenscalesforadmittedpost-operativepatients. Theseoutcomeswererstanalyzedinlinearprobabilitymodelandtestedfor heteroskedasticity,thenasapaneldataset.Finally,datawerettostatistical processcontrolchartsforacloserlookatperformanceinclinicalmicrosystems. RESULTS: Among43,844hospitalinpatients,therewere337totalincidencesof HAPUshospital-wide.Aprobitregressionmodelpredictedthecorrelationofage, gender,andlength-of-stayonHAPUincidence(R2 0.096).Paneldataanalysis determinedthatforeachadditionaldayinthehospital,therewasa0.3%increase inthelikelihoodofdevelopingaHAPU.Ap-chartofHAPUincidenceshoweda meanincidencerateof1.17%thatremainedinstatisticalcontrol.Basedonthe g-chart,theaveragetimebetweeneventsforthelast25HAPUswas13.25days. Therewasa55-dayperiodbetweentwoincidencesduringtheobservationperiod. Thep-chartaddressingBradenScaleassessmentsshowedthat40.5%ofallpatients wererisk-stratieduponadmission. CONCLUSIONS: Standardregressionanalysis islimitedtohospital-widedata,andshouldnotbeusedtointerpretdiseaseoutcomesattheleveloftheclinicalmicrosystem.Statisticalprocesscontrolcharts amplifypatientoutcomesatthepointofcare,andareusefulforguidinganddevelopingqualityimprovementinterventions.A90VALUEINHEALTH14(2011)A1–A214

PAGE 99

POSTERSESSIONII: DISEASE-SPECIFICSTUDIES Diabetes/EndocrineDisorders–ClinicalOutcomesStudies PDB1 TREATMENTPATTERNSANDOUTCOMESAMONGOLDEREXENATIDEBID USERSCOMPAREDTOINSULINGLARGINEUSERSGaeblerJA1,SchultzJF2,BlickensderferA3,WentenM1,RiedelA4,HarleyC5 1AmylinPharmaceuticals,Inc.,SanDiego,CA,USA,2UniversityofMinnesota,Duluth,MN,USA,3AmylinPharmaceuticals,Inc.,Chestereld,MO,USA,4i3Innovus,EdenPrairie,MN,USA,5i3 Innovus,PaloAlto,CA,USAOBJECTIVES: Weevaluatedtreatmentpatterns,hypoglycemicevents,glycemiccontrolandmedicalcostsforpatients 60ywithtype2diabetesmellitus(T2DM)newly initiatingtreatmentonexenatideBID(ExBID)orinsulinglargine(IG). METHODS: Medical,pharmacyandlabdataforcommercialandMedicareAdvantageenrolleeswere obtainedfromadministrativeclaimsfromalarge,nationalhealthplan.Subjectswere identiedfromMay2005toDec2008with6-mobaselineand12-mofollow-upperiods. Subjectswithanyinsulintreatmentduringbaselinewereexcluded.Propensityscore (PS)matching(1:1)ofExBIDandIGsubjectswasusedtocreatebalancedcohorts. Logisticregressionmodelswereusedtoanalyzemedicationadherence(Medication PossessionRatio(MPR) 80%),therapypersistence(gapof60daysintreatment discontinuation),anyacutehypoglycemicevent,andglycemiccontrol(follow-upA1C levelsof 7%). RESULTS: Thenalmatchedstudysampleincluded3,263subjectsper cohort;averageagewas65y(SD4.9);48%werefemale;83%wereenrolledinacommercialhealthplan.Inthefollow-upperiod,ExBIDpatientsexperiencedsignicantly fewerhypoglycemiceventscomparedtoIGpatients(15vs.40,p 0.001).Subjectsin theExBIDcohortwerealsomorelikelytoobtainanMPR 80%(OddsRatio(OR) 1.93, CL1.72-2.17),lesslikelytodiscontinuetherapy(OR 0.28,CL0.25-0.31)andlesslikelyto haveanyacutehypoglycemicevent(OR 0.48,CL0.26-0.88)duringthefollow-upperiod.ForthosewithvalidA1Clevelsinbothbaselineandfollow-upperiods(N 669; ExBID 359,IG 310),ExBIDsubjectsweremorelikelytoachieveA1C 7%(OR 1.62, CL1.11-2.35)comparedtoIGsubjects.Associatedmedicalcostsweresimilarforboth groups.Resultswerecomparableforpatients 65y. CONCLUSIONS: OlderExBID subjectsweremorelikelytoadheretotherapyandachieveA1C 7%andlesslikely todiscontinuetherapyorexperienceanyacutehypoglycemiceventthanolderIG subjects. PDB2 EVALUATINGONE-YEARSTABILITYOFTHETOTALILLNESSBURDENINDEXBillimekJ1,CantrellR2,GreeneldS1,KaplanS1 1UniversityofCaliforniaIrvine,Irvine,CA,USA,2EliLillyandCompany,Indianapolis,IN,USAOBJECTIVES: TheTotalIllnessBurdenIndex(TIBI)isawell-validated,patientreportedmeasureofthepresenceandseverityofmedicalconditionscomorbidto anindexconditionandisanimportantcomponentofthePotentialforBenet ConceptualModel,showntopredictresponsetotreatmentindiabetes.Theobjectiveofthisresearchwastodeterminetheone-yearstabilityoftheTIBI. METHODS: Weidentiedpatientswithtype2diabetesenrolledintheReducingRacialDisparitiesinDiabetesCoachedCare(R2D2C2)studywhohadcompletedthe47-itemTIBI atbaselineandoneyearfollow-up,andwhoseantihyperglycemicmedicationregimenhadnotchanged(n 364).Forthissample,weestimatedstandarderrorof measurement(SEM)fortheTIBIandcomparedchangesinTIBIscoresusingpaired t-testsandlimitsofagreement. RESULTS: SEMfortheTIBI-47was0.90.Scoresfor baselineandfollow-upwere3.52and2.88respectively(range 0-13),meanimprovement 0.64,withintheSEM.82.2%ofrespondentshadachangeintheTIBI scoreof 3points.Minorimprovements( 5%ofscalerange)wereobservedin transientsymptoms(e.g.uwithcough,shortnessofbreath,etc.)versusdiagnoses (e.g.congestiveheartfailure,stroke,etc.).Allchangesinsymptomswerewithin SEMforrespectiveTIBIsubdimensions. CONCLUSIONS: TIBIscoresappearedtobe highlystableforthisstudysampleovertheone-yearobservationperiod.Small changesobservedinsymptomseverityitemsmaybeduetoeffectiveclinicalmanagement.Furtherresearchisneededtodistinguishobserveddifferencesdueto treatmentresponsefrommeasurementerror. PDB3 LONG-TERMANTIPROTEINURICEFFECTOFALISKIRENINDIABETICPATIENTS WITHPERSISTENTALBUMINURIADESPITECHRONICACEIORARBTREATMENTHuangW1,MersyJH1,LevinP2 1GreaterBaltimoreMedicalCenter,Baltimore,MD,USA,2BayWestEndocrinologyAssociates, Baltimore,MD,USAOBJECTIVES: Studiesshoweddualblockadeofrenin-angiotensin-aldosteronesystemwithACEIandARBhadproducedlittlegainwithsignicantsideeffectsin proteinuricpatients.Hereweevaluatedthereal-worldeffectivenessoflong-term antiproteinuriceffectsofanoveldualblockadewiththedirectrenininhibitor aliskirenadditiononchronicACEIorARBtreatmentindiabeticpatientswithpersistentalbuminuria. METHODS: WeretrospectivelycollecteddatafromtheelectronicdatabaseofBaywestEndocrinologyAssociates.Aliskiren150-300mgdaily wasaddedonchronicACEI/ARBtreatment(meanknownduration6years).The Cockroft-Gaultformulawasusedtoestimatetheglomerularltrationrate(eGFR). TheWilcoxonSignedRankstestandKendall'stau-bcorrelationanalysiswereused tocalculatestatisticalsignicance. p 0.05wasconsideredsignicant. RESULTS: Twenty-four(meanage64years)patientswereselected,withameanobservation time15monthsbeforeand20monthsafteraliskirenaddition.Therewasasignificantriseinurinaryalbumin-to-creatinineratio(Ualb/Cr)[unit:mg/g,146.3(16.11978.7)vs.51.0(34-1170.3)]andSBP(unit:mmHg,158 19vs.142 21)inthese patientsatbaseline(i.e.,whenAliskirenwasadded)whencomparedwiththatof15 monthsbefore( p 0.05).AndtherewasasignicantdeclineinalbuminuriaandSBP 20monthsafteraliskirenaddition:reducedmedianUalb/Crwas73.3%,reduced meanSBPwas24mmHg( p 0.001).However,reductioninUalb/Crwasnotrelated toreductioninSBP( p 0.05).NochangeswerefoundintheDBP,HbA1c,serum creatinineandeGFR.Asmallincrementofserumpotassiumlevelswasdetected afteraliskirenaddition(unit:mmol/L,4.5 0.6vs4.4 0.4mmol/Latbaseline, p 0.05).Duringthestudy,noadverseevents,includingsymptomsofhyperkalemiaandhypotension,weredocumented. CONCLUSIONS: Aliskirenadditionto ACEI/ARBmayhaveextraantiproteinuriceffectsthatmightbeindependentonBP reductionandheldwithoutreboundaslongas1.5yearfollow-up. PDB4 MODELINGHEALTHANDECONOMICOUTCOMESASSOCIATEDWITH EXENATIDEONCE-WEEKLYVERSUSINSULINANDPIOGLITAZONETREATMENT FORTYPE-2DIABETESGaeblerJA1,Soto-CamposG2,AlperinPE3,HoogwerfBJ4,WintleM1,MaggsD1,HanJ1, BlickensderferA5 ,PencekR1,BruhnD6,PeskinBR2 1AmylinPharmaceuticals,Inc.,SanDiego,CA,USA,2Archimedes,Inc.,SanFrancisco,CA,USA,3Archimedes,Inc.,SanDiego,CA,USA,4EliLillyandCompany,Indianapolis,IN,USA,5Amylin Pharmaceuticals,Inc.,Chestereld,MO,USA,6EliLillyandCompany,SanDiego,CA,USAOBJECTIVES: Exenatideonce-weekly(ExQW)isaGLP-1receptoragonistthatimprovesglycemiainpatientswithtype2diabetes(T2DM)whileelicitingpotential weightlossandimprovementincardiovascularriskfactors(bloodpressure(BP)and plasmalipids).Inpublishedtrials,ExQWresultedinsuperiorreductioninA1Ccomparedtomaximumdailydosesofsitagliptinandpioglitazone(Pio)onmetformin(Met) background,andtotitratedinsulinglargine. METHODS: WeusedtheArchimedes Model,avalidated,clinicallydetailedmodelofphysiology,disease,andhealthcare delivery,toexplorepotentiallong-termExQWbenetsandcosts.Wesimulated20yof treatmentinavirtualpopulation(n 19,885)basedonindividualswithT2DMdrawn fromNHANESwhowereonMet / sulfonylureas(meanage59y,BMI33kg/m2,wt 93kg,durationT2DM9y,baselineA1C8%).Theeffectsofthreetreatmentregimens weremodeledatsimulationstart:1)advancementtoinsulinatA1C 8%(treatto targetA1C 7%),2)additionofPio,and3)additionofExQW.ExQW'seffectonA1C, weight,BP,andlipidswasderivedfromfourphase3ExQWtrials.Medicalcosts(inpatient,outpatient,ambulatory,treatments)werederivedfromtheMedicareCurrent BeneciarySurvey,MedicarePartDdata,drugstore.com,andpublishedliterature. SinceExQWisinvestigational,antidiabetictherapycostswereexcluded. RESULTS: At 20y,nalmeanA1Cwas 7%inallarms.ComparedtoinsulinandPio,respectively, ExQWwasassociatedwithrelativereductionsinnalmeanweight(5%vs.bothcomparators),incidenceofrstMajorAdverseCardiovascularEvent(8.2%and3.3%),and CHF(5.1%and15.0%).Allarmsshowedcomparablebenetincontrollingneuropathy, butExQWshowedsignicantlygreaterreductionsinrenalcomplications.Asearlyas 5y,ExQWdemonstratedtotalcost-savingsof$545and$379perlife-yearvs.insulin andPio,respectively. CONCLUSIONS: Thesesimulationssuggestthatthebenetsof ExQWmaytranslateintoclinicallyandeconomicallymeaningfulreductionsinlongtermoutcomes. PDB5 THEIMPACTOFANTIDIABETIC-INDUCEDHYPOGLYCEMIAONCLINICAL OUTCOMESANDRESOURCEUTILIZATIONAMONGVETERANSWITHTYPE-2 DIABETESMELLITUS(T2DM)ZhaoY,ShiL,FonsecaV,CampbellC TulaneUniversity,NewOrleans,LA,USAOBJECTIVES: Toexaminetheimpactofantidiabetic-inducedhypoglycemiaon clinicaloutcomesandresourceutilizationamongT2DMpatientsintheVeterans Affairs. METHODS: Thisretrospectivecohortstudyusedelectronicmedicalrecords between01/01/2004and09/01/2010fromtheVeteransIntegratedServiceNetwork (VISN)16datawarehouse.Patientswererequiredtohaveatleast2recordsofT2DM ICD-9-CMdiagnosis(250.xxexceptfor250.x1and250.x3).The rstdispensedateof anewantidibeticagent(indexdrug)weredenedastheindexdate.Nohypoglycemia diagnosiswasallowedduringtheone-yearpre-indexperiod.ThehypoglycemiacohortandcontrolcohortweredenedbyICD-9-CMdiagnosisofhypoglycemia(250.8, 251.0,251.1and251.2)withintheindex-treatmentperiodandnohypoglycemiaduring theone-yearpost-indexperiod,respectively.Clinicaloutcomesincludedhemoglobin A1c,cardiovasculardiseases(CVD)andmicro-vascularcomplications.Resourceutilizationincludedhospitalizationsandemergencyroom(ER)visits.Clinicaloutcomes werecomparedbyCoxregressionmodels,controllingforage,race,indexdrug,renal functions,andbaselinevariables:A1c,Charlsoncomorbidityindex(CCI),andresource utilization. RESULTS: Among42,437T2DMpatients,369patientsofthehypoglycemiacohortand42,068patientsofthecontrolcohortdifferedinracialandmarital status,baselineCCIandresourceutilization.Thehypoglycemiacohortwasmore likelytoreceiveinsulin/sulfonylureaasindexdrug.Thepost-indexA1cwasnumericallyhigherinthehypoglycemiacohortthancontrolcohort(10.12vs9.87,p 0.0602).ThehypoglycemiacohortwasmorelikelytodevelopCVD(HR 1.32,95%CI: 1.13-1.54)and40%morelikelytodevelopmicro-vascularcomplications,compared withthecontrolcohort.RisksofhospitalizationandERvisitwerehigherforthe hypoglycemiacohortthancontrolcohort(18.2%vs.9.2%,32.3%vs.20.7%,both p-values .0001,respectively). CONCLUSIONS: Patientswithhypoglycemiamay leadtoworseclinicaloutcomesandhigherrisksofhospitalizationandERvisitthan thosewithout. PDB6 LONG-TERMEFFECTIVENESSOFMANAGINGDIABETESWITHTHECHRONIC CAREMODEL:SIMULATIONSPERFORMEDUSINGARCHESZgiborJC1,KuoS1,SchuetzCA2,RobertsMS1,CohenMD2 1UniversityofPittsburgh,Pittsburgh,PA,USA,2Archimedes,Inc.,SanFrancisco,CA,USAA91VALUEINHEALTH14(2011)A1–A214

PAGE 100

OBJECTIVES: TheChronicCareModel(CCM)isaframeworkforenhancinghealth caredeliveryforchronicillnessesthroughsixelements:self-managementsupport, deliverysystemdesign,communityresources,organizationalsupport,decision support,andclinicalinformationsystems.TheCCMwasimplementedatUSAF WilfordHallMedicalCenterfrom2006through2008,butlittleisknownaboutits long-termeffectiveness. METHODS: WeusedtheARCHeSinterfacetotheArchimedesmodeltoforecasttheeffectivenessoftheCCMcomparedtothecurrent levelsofcareintheUS(CLC)over20yearsforpatientswithtype2diabetes.We modeledtheCCMinterventioneffectsonhealthoutcomesanddiseaseprogression ofdiabetesbasedonempiricalbiomarkerchangeresultsfromthree-yearmilitarybasedclinicaldata.Theprimaryanalysisfocusedontheeffectivenessintermsof diabetesrelatedhealthoutcomes,whileasecondaryanalysisconsideredcosteffectivenessfromahealthcaresystemperspective. RESULTS: Comparedwith CLC,therelativereductionsin20-yearKaplan-MeierincidenceduetotheCCM were10.3%,3.8%,5.8%,6.0%,11.1%formyocardialinfarction,proliferativediabetic retinopathy,blindness,footulcer,andfootamputation,respectively.TheCCM resultedina6.7%relativereductionincoronaryheartdiseasedeath,anda0.6% relativereductioninall-causedeath.TheCCMhadnocompellingbenetsfor reducingtheincidenceofstrokeornephropathy. CONCLUSIONS: Wefoundthe CCMlowerstheriskofmorbidityandmortalityfromdiabetesinamilitarypopulation.OurresultsusingARCHeSsimulationsarebroadlyconsistentwithprevious MarkovmodelstudiesbutextendthoseresultsbyprovidingamoredetailedunderstandingofthespecichealthoutcomesaffectedbytheCCM,aswellasthe impactoncomorbiditiesassociatedwithdiabetes. PDB7 DIETTHERAPIESINPATIENTSWITHTYPE-2DIABETES:AMIXED-TREATMENT COMPARISONOFRANDOMIZEDCONTROLTRIALSHarringtonAR,MaloneD TheUniversityofArizona,Tucson,AZ,USAOBJECTIVES: Type2DiabetesMellitus(TTDM)contributestothebody'sinabilityto produceasufcientamountofinsulininordertosustainglucoselevelsnecessary forenergyproduction.Thestudyaimistodeterminetheassociationbetween multiplediettherapiesandtheclinicalproleofadultpatientswithTTDM. METHODS: Weconductedamixed-treatmentcomparisonmeta-analysisofrandomizeddiettherapiesforTTDMpatients.Therapiesincludedhighcarbohydrate andisocaloric(HCI);highcarbohydrateandhypocaloric(HCH);lowcarbohydrate andisocaloric(LCI);andlowcarbohydrateandhypocaloric(LCH).Asystematic literaturesearchwasconductedthroughDecember2010.Theprimaryoutcomes measuredwerehemoglobinA1c(HbA1c),high-densitylipoprotein(HDL)cholesterol,low-densitylipoprotein(LDL)cholesterol,andtriglycerides.Theseoutcomes weredichotomizedbaseduponAmericanDiabetesAssociation(ADA)standards.A dietwasconsideredsuccessfulifthepatient'sHbA1cwas 7%,HDLwas 40 mg/dL,LDLwas 100mg/dL,ortriglycerideswere 150mg/dL.Mixed-treatment comparisonmeta-analysiswasusedtocombinedirect,within-trial,andbetweentrialcomparisonswithindirecttrialevidencefromothertrialswhilemaintaining randomization.Resultsarereportedasoddsratios(OR)and95%credibleintervals (CI). RESULTS: Intotal,10studies(21treatmentarmsand340patients)wereincludedintheanalysis.ForbothHbA1candtriglycerides,theHCHdietyieldedlower measurementscomparedtotheHCIdiet(OR:0.89CI:0.04,20.86;OR:0.58CI:0.02, 15.55,respectivelyforHbA1candtriglycerides)andLCIdiet(OR:0.47CI:0.02,7.45; OR:0.15CI:0.001,6.28,respectivelyforHbA1candtriglycerides).LDLandHDL cholesterolmeasurementswerenotfoundtobestatisticallysignicantlydifferent amongthefourdiettypes. CONCLUSIONS: Thisstudyillustratednostatistically signicantdifferencesexistedbetweenthefourtypesofdietswithrespecttofour clinicaloutcomes.FurtherresearchisnecessarytoidentifyanoptimaldietcombinationforpatientswithTTDM. PDB8 EFFECTIVENESSANDADHERENCEWITHFIXED-DOSECOMBINATION(FDC) VERSUSCOADMINISTEREDDUALTHERAPY(CDT)ANTIHYPERGLYCEMIC REGIMENS:AMETA-ANALYSISHanS,DaviesMJ,LentoK,RadicanL,ZhangQ MerckSharp&DohmeCorp.,WhitehouseStation,NJ,USAOBJECTIVES: TocomparetheeffectofantihyperglycemicFDCsonA1CandadherencerelativetoCDTsinclinicalpractice. METHODS: A2-partsystematicliterature reviewandmeta-analysiswereperformedtocomparetheA1Cresponseandadherencebetweenthe2drugregimens.Inclusioncriterialimitedarticlestostudies thatcomparedequivalentdrugcomponentswithinFDCandCDT.Searchesused Medline,Embase,andCochranedatabases.Ofthe1246identiedabstracts,152 articleswerereviewed,and8mettheinclusioncriteria.Resultswereextractedand pooledinameta-analysis,usingarandom-effectsmodel.Unreportedstandard deviationswereimputedaccordingtotheCochraneHandbook.Cohortcomparisonsweredescribedasmeandifferences(MD)with95%condenceintervals(CIs). RESULTS: Atotalof37,173patientscomprisedthe16reportedcohortcomparisons. FivecomparisonsdescribedA1CreductionsinFDCandCDTcohorts,andametaanalysisrevealedagreaterreductioninA1CwithFDC(MD, 0.5%[95%CI: 0.8, 0.3%]).Elevencomparisonsevaluatedmedicationadherence(measuredbymedicationpossessionratio[MPR])betweentheFDCandCDTcohorts.These11cohort comparisonsweredividedinto3subgroupsduetodifferentstudydesigns.Five comparisonsdescribedMPRforFDCandCDTcohorts,withhigheradherencewith FDC(MD,12%[95%CI:5,19%]).Threecomparisonsexaminedpatientswho switchedfrommonotherapytoFDCorCDT,withhigheradherenceforpatients whoswitchedtoFDC(MD8%[95%CI:6,10%]).Threecomparisonsdescribedresults forpatientswhoswitchedfromCDTtoFDCorstayedonCDT,withhigheradherenceforpatientswhoswitchedtoFDC(MD5%[95%CI,3,7%]). CONCLUSIONS: Use ofantihyperglycemicFDCswasassociatedwithlowerA1ClevelsandhigherMPRs comparedtoCDTinpatientswithT2DMinclinicalpractice. PDB9 IMPACTOFEXENATIDE,PENINSULINANDVIAL/SYRINGEINSULINON PATIENTOUTCOMESINADIABETESPOPULATIONINTHEUNITEDSTATES:A RETROSPECTIVEDATABASEANALYSISOFPERSISTENCEANDFIRST-YEAR COSTSRashidN1,McCombsJ1,FosterS2,MillerLA2 1USCSchoolofPharmacy,LosAngeles,CA,USA,2EliLillyandCompany,Indianapolis,IN,USAOBJECTIVES: TheObjectiveofthisresearchistocomparehealthcarecostsand medicationpersistenceforpatientsinitiatingtreatmentusingexenatide,peninsulinorvialinsulin. METHODS: Commercialhealthplandata[2004to2008]were usedtoidentifyepisodesofanti-diabeticdrugtherapywhichwerethenclassied accordingtotreatmenthistory:rstobservedtreatment;restartingaprevioustherapy[90-daygapinalltreatment];switchingtherapy;andaugmentationtherapy. Allexenatideandinsulinepisodeswereselectedforthisanalysis.Multivariate statisticalmethodsadjustedfordemographics,drugusehistory,priormedicalcare use,comorbidmedicalconditions,andprescriptiondrugprole.Severalsensitivity analyseswereconducted. RESULTS: 213,701episodesofanti-diabeticdrugtherapy wereidentiedofwhich7,031patientsinitiatedusingexenatide,21,011usedvial insulin[VI]and422usedpeninsulin[PI].Timetoallcausediscontinuation[TTAD] wasmeasuredfortheindexdrugandalldiabetic-relateddrugs.Peninsulin[PI]was discontinued91daysearlierthanexenatide[p 0.0001],whilevialinsulin[VI]was continued18dayslongerthanexenatide[p 0.01].PIpatientsdiscontinuedall anti-diabeticdrugs34daysearlierthanexenatidepatients[p 0.05]whileexenatideandVIpatientsexhibitedsimilarTTADforalldrugs.Relativetovialinsulin, exenatideusewasestimatedtoreducerst-yearmedicalcostsby $4,629 [p 0.0001]whichwassufcienttooffsethigherprescriptiondrugcostsforexenatide[ $519,p 0.0001].Similarcostresultswerefoundcomparingexenatideto peninsulinbutwerenotstatisticallysignicant,likelyduetothelimitedsample sizeforpeninsulin.Theseresultswereconrmedusingpropensityscorematching estimationandwererobustacrossepisodetype. CONCLUSIONS: Patientsinitiating drugtherapyusingexenatideincurlowerpost-treatmentcoststhansimilarpatientswhoinitiatedtreatmentusinginsulin. PDB10 ASSESSINGTHEABILITYOFPATIENTREPORTEDMEASURESTODISCRIMINATE PROXIESFORTREATMENTRESPONSECantrellR1,BrennemanS2,KaplanS3,ShenW1,BrekkeL2,BillimekJ3,PaczkowskiR1, HaynesV4 ,GreeneldS3,BueschingD1 1EliLillyandCompany,Indianapolis,IN,USA,2i3Innovus,EdenPrairie,MN,USA,3University ofCaliforniaIrvine,Irvine,CA,USA,4i3Innovus,BaskingRidge,NJ,USAOBJECTIVES: ThePotentialforBenetconceptualmodelisaframeworktoidentify subgroupsofpatientswithdifferentialpotentialforresponsetotreatment.Within thiscontext,wedevelopedtheENSEMBLEMinimumDataset(MDS),acollectionof patientreportedmeasures,toassesspatientheterogeneity.Theobjectiveofthis researchistoprovideevidencethatENSEMBLEMDSmeasurescanbeusedtoform subgroupsthatareclinicallyandstatisticallymeaningfulintheirabilitytodiscriminateproxiesfortreatmentresponse. METHODS: Across-sectionalsurveywas mailedto3380type2diabetespatientsidentiedfromalargeadministrative claimsdatabase.ENSEMBLEMDSmeasuresincluded:TotalIllnessBurdenIndex (TIBI),EQ-5D,PsychologicalHealthQuestionnaire(PHQ-4),PerceivedStressScale (PSS-4),PerceivedSocialSupport(MOS-SS),MacArthurScaleofSubjectiveSocial Status(M-SSS),MacArthurIncomeQuestion(MIQ),BarrettSimpliedMeasureof SocialStatus(BSMSS),Self-reportedDepression(DQ),andSelf-ReportedHealth (SRH).Proxiesfortreatmentresponsewerepatientglobalimpressionofdisease severity(PGIS),patientglobalimpressionofimprovement(PGII)andoveralldisease burden.WeassessedtheabilityofENSEMBLEMDSmeasurestodiscriminate amonglevelsofPGIS,PGII,andburdenviaunadjusted,adjusted,andtrend analyses. RESULTS: Atotalof724patientsrespondedwithcompletesurveydata andwerecategorizedinto3levelsonPGIS,PGII,andburden.TIBI,EQ-5D,PHQ-4, PSS-4,MOS-SS,andM-SSSdiscriminatedacrossthelevelsofPGIS,PGII,andburden (p 0.05)forallanalyses.DQdiscriminatedburdenforallanalyses(p 0.01)and PGIIandPGISforunadjustedandtrendanalyses(p 0.05).BSMSSonlyshowed signicanttrendforburden(p 0.05). CONCLUSIONS: Resultsofthisstudyprovide evidencethatthecomponentmeasuresoftheENSEMBLEMDSdiscriminateamong levelsofdiseaseseverity,improvement,andburden.Futureresearchincludestestingthesemeasuresinotherdiseaseareas,testingalternativepatientreported measures,andcreatingacompositescoringalgorithm. PDB11 COMPARISONOFMULTIVARIABLE-ADJUSTEDLOGISTICREGRESSIONWITH PROPENSITYSCORE-MATCHED,PROPENSITYSCORE-STRATIFIED,AND PROPENSITYSCORE-ADJUSTEDLOGISTICREGRESSIONMODELSKhozaS,BarnerJC,RichardsK UniversityofTexasatAustin,Austin,TX,USAOBJECTIVES: Tocomparethemultivariable-adjustedlogisticregressionmodel withthepropensityscore-matched,propensityscore-stratied,andpropensity score-adjustedlogisticregressionmodelsinestimatingtheeffectofexposureto antidepressantagentsinincreasingtheriskoftype2diabetesmellitus. METHODS: AretrospectivecohortstudyusingtheTexasMedicaidprescriptionclaimsdatabasewasconductedfromJanuary1,2002toDecember31,2009.Patientsaged18-64 yearswithnewprescriptionsforantidepressants(exposedgroup)orbenzodiazepines(unexposedgroup)andwithoutdiabetesatcohortentrywereincludedinA92VALUEINHEALTH14(2011)A1–A214

PAGE 101

thestudy.Propensityscores,whichpredictedexposuretoantidepressantagents, wereusedtocreatepropensityscore-matched,propensityscore-stratied,and propensityscore-adjustedlogisticregressionmodels. RESULTS: Atotalof44,715 patientsformedthestudysample.Theriskestimatesvariedacrossdifferentanalyticmethods.Thepropensityscore-matchedlogisticregressionmodelyieldedthe highestriskestimate(RelativeRisk[RR] 1.452;95%CondenceInterval[CI]:1.276 1.651),followedbythemultivariable-adjustedlogisticregressionmodel(RR 1.319; 95%CI:1.0671.630)andthepropensityscore-stratiedlogisticregressionmodel (RR 1.153;95%CI:1.0331.287).Thepropensityscore-adjustedregressionmodel yieldedthesmallestriskestimate(RR 1.080;95%CI:0.9681.205). CONCLUSIONS: Propensityscoretechniquesusingpharmacyclaimsdatawithalimitednumberof covariatesyieldedvaryingestimatesofthetreatmenteffect.Thechoiceofthe propensityscoretechniquemayinuencethemagnitudeofthetreatmenteffect estimate. PDB12 USEOFANTIDEPRESSANTSANDTHERISKOFDIABETESMELLITUS:ANESTED CASE-CONTROLSTUDYKhozaS,BarnerJC,LawsonK,RascatiKL,WilsonJP,BohmanTM UniversityofTexasatAustin,Austin,TX,USAOBJECTIVES: Todeterminewhethercurrentuseofantidepressantsisassociated withincreasedriskoftype2diabetesmellituscomparedtoformeruse. METHODS: Anestedcase-controlstudywasconductedwithinacohortof35,552patientswith newantidepressantprescriptionsfromJanuary1,2002toDecember31,2009using TexasMedicaid.Patientsaged18-64yearsandwithoutdiabetesatcohortentry wereincludedinthestudy.Antidepressantexposurewasclassiedascurrentuse (within6monthofdiabetesdiagnosis)andformeruse( 6monthsbeforediabetes diagnosis).Conditionallogisticregressionwasusedtoaddressthestudyobjective. RESULTS: Atotal2,507casesofincidentdiabetesand10,028matchedcontrolpatientswereidentied.Theaverageagewas44.9(SD 13.0)andthemajority(75.6%) werefemale.Comparedtoformeruse,currentuseofantidepressantswasassociatedwitha2.1-foldincreaseintheriskofdiabetes(OddsRatio[OR] 2.108;95% CondenceInterval[CI]:1.5202.923).Comparedtoformeruse,currentuseof tricyclicantidepressants(TCAs;OR 2.279;95%CI:1.3443.865),selectiveserotoninreuptakeinhibitors(SSRIs;OR 2.241;95%CI:1.5393.263),serotonin-norepinephrinereuptakeinhibitors(SNRIs;OR 1.938;95%CI:1.1353.310),andOther (OR 1.769;95%CI:1.0902.872)wereassociatedwithanincreasedriskofdiabetes. Amongcurrentantidepressantusers,therewerenosignicantdifferencesinthe riskofdiabetesforSSRIs(OR 0.833;95%CI:0.5641.232),SNRIs(OR 0.787;95%CI: 0.4801.280),andOther(OR 0.673;95%CI:0.4281.057),comparedtoTCAs. CONCLUSIONS: Comparedtoformeruse,currentuseofantidepressantswasassociatedwithatwo-foldincreaseintheriskofdiabetesmellitus.Thisassociation wasalsoobservedwhencurrentusersofTCAs,SSRIs,SNRIs,andOtherantidepressantswerecomparedwithformerusers. PDB13 USEOFANTIDEPRESSANTSANDTHERISKOFDIABETESMELLITUS:A RETROSPECTIVECOHORTSTUDYKhozaS,BarnerJC,LawsonK,RascatiKL,WilsonJP,BohmanTM UniversityofTexasatAustin,Austin,TX,USAOBJECTIVES: Todeterminewhether:1)useofantidepressants,comparedtobenzodiazepines,increasestheriskoftype2diabetesmellitus;2)individualantidepressantclasses,comparedtobenzodiazepines,increasestheriskoftype2diabetes;and3)therearedifferencesintheriskoftype2diabetesamongantidepressant classes. METHODS: AretrospectivecohortstudyusingtheTexasMedicaidprescriptionclaimsdatabasewasconductedfromJanuary1,2002toDecember31, 2009.Patientsaged18-64yearswithnewprescriptionsforantidepressantagents (exposed)andbenzodiazepines(unexposed)andwithoutdiabetesatcohortentry constitutedthebasestudypopulation.Logisticregressionanalyseswereusedto addressthestudyobjectives. RESULTS: Atotalof44,715patientsformedthestudy sample.Ofthese,35,552wereexposedand9,163wereunexposed.Theaverageage was38.6(SD 14.2)andthemajority(69.3%)werefemale.Useofantidepressants wasassociatedwithanincreasedriskofdiabetes(RelativeRisk[RR] 1.396;95% CondenceInterval[CI]:1.1261.729)comparedtobenzodiazepineuse,aftercontrollingforage,gender,medicationadherence,persistence,numberofdiabetogenicmedications,ChronicDiseaseScore,treatmentduration,andyearofcohort entry.Comparedtobenzodiazepines,tricyclicantidepressants(TCAs;RR 1.445; 95%CI:1.0571.977),serotonin-norepinephrinereuptakeinhibitors(SNRIs; RR 1.593;95%CI:1.1602.186),andselectiveserotoninreuptakeinhibitors(SSRIs; RR 1.317;95%CI:1.0371.673),wereassociatedwithanincreasedriskfordiabetes.ComparedtoTCAs,therewerenosignicantdifferencesintheriskofdiabetes forSSRIs(RR 0.876;95%CI:0.6781.132),SNRIs(RR 1.003;95%CI:0.7341.371) andOtherantidepressants(RR 0.785;95%CI:0.5821.058). CONCLUSIONS: Overall,useofantidepressantswasassociatedwithanincreasedriskofdiabetesmellituscomparedtobenzodiazepineuse.Thisassociationwasalsoobservedwhen tricyclicantidepressants,selectiveserotoninreuptakeinhibitors,andserotoninnorepinephrinereuptakeinhibitorswerecomparedtobenzodiazepines. PDB14 USINGHEALTHOUTCOMESMODELINGTOASSESSTHEBENEFIT-RISKPROFILE OFEXENATIDEONCE-WEEKLYVERSUSINSULINGLARGINEFORPATIENTS WITHTYPE-2DIABETESGarrisonL1,BestJH2,RobertsonJC1,WentenM2,BruhnD3 1UniversityofWashington,Seattle,WA,USA,2AmylinPharmaceuticals,Inc.,SanDiego,CA, USA,3EliLillyandCompany,SanDiego,CA,USAOBJECTIVES: Understandingthebenet-riskprolesofmedicationsfortype2diabetes(T2D)isofhighpublichealthinterest.Weassessedthepotentialcomparativehealthbenetsandrisksofexenatideonceweekly(ExQW)comparedtoinsulin glargine(IG)undervariousassumptionsabouthypotheticalrisksfrompancreatic andthyroidcancerassociatedwitheithertreatment. METHODS: SafetyandefcacydatafromtherandomizedcontrolledDURATION-3trialwereincorporated intotheCOREDiabetesModeltopredictlong-termhealthoutcomes;pancreatic andthyroidcancerrisksfromSEER(generalpopulation);i3/MarketScanclaims databases(T2Dand/ortherapy)wereincorporatedintoderivedMarkovmodel. Incrementalnethealthbenet(INHB)wasestimatedinlifeyears(LYs)andqualityadjustedlifeyears(QALYs)(bothdiscountedat3%)fordifferenttimehorizons(base case 30years)anddifferentassumptionsaboutcancerrisks.Theincidencerateof pancreaticcancerwas1.17/10,000population/year.Inthebase-casescenariono increasedpancreaticcancerriskfromExQWandrelativerisk(RR)of1.25forIGwas assumed.Theriskforthyroidcancerwasassumedsameasingeneralpopulation andequalforExQWandIG.Intwoalternativescenariosforpancreaticcancerrisk: a)assumingIGRR 1.33andExQWhavingnoincreasedrisk,andb)assuming ExQWRR 1.07andIGRR 1.25. RESULTS: TheINHBforExQWversusIGwas0.118 Lysand0.239QALYsinbasecase;0.119LYsand0.239QALYsinscenarioa);and 0.118LYsand0.238QALYsinscenariob).Neitherscenarioa)orb)changedINHBby morethan1%.TheINHBforExQWremainedpositiveunderdifferentassumptions abouttimehorizonandrisklevels. CONCLUSIONS: Inallscenariosmodeled,the INHBforExQWwaspositivecomparedtoIG.Thisstudysuggeststhatthepotential benetsfromExQWinreducingestablishedT2Dcomplicationsareexpectedto outweighhypotheticalrisksfrompancreaticandthyroidcancer. PDB15 SHORT-TERMASSOCIATIONOFANTIDEPRESSANTDRUGUSE,LIFE-STYLERISK FACTORSANDNEW-ONSETDIABETESWilkinsTL,SambamoorthiU WestVirginiaUniversity,Morgantown,WV,USAOBJECTIVES: Theobjectiveofthecurrentstudyistoassesstheshort-termassociationbetweenantidepressantdruguseandtheriskofnew-onsetdiabetesusing twoyearsofobservation. METHODS: Thisstudyusedthelongitudinaldataon 26,990adultsoverage21fromtheMedicalExpenditurePanelSurvey(MEPS),a large-surveyoffamiliesandindividualstoproducenationalestimatesofhealthcareuseandexpenditures.Wepooledlongitudinaldataforyears2004-2007.DepressionanddiabetesstatuswereidentiedfromtheMEPShouseholdlesand medicalconditionles.Antidepressantusewasderivedfromprescriptiondrug reports.Chi-squaretestsandlogisticregressionswereusedtoexaminethelink betweenantidepressantuseinteractedwithdepression,andnew-onsetdiabetes, aftercontrollingfordemographic,socio-economic,healthcareaccess,healthstatus,andlife-styleriskfactors(obesity,smoking,andphysicalactivity).Theindependentvariableswereenteredinblocks. RESULTS: Antidepressantsusewasreportedin11%ofindividuals.Inunadjustedmodels,theriskofnew-onsetdiabetes wassignicantlyincreasedforpersonsusingantidepressantswithdepression ( (AOR 2.12) comparedwiththosewhodidnothaveeither.Whenlifestylerisk factorswereenteredinthemodel,statisticalsignicancedisappeared.Independently,lifestyleriskfactorssignicantlyincreasedtheriskofnew-onsetdiabetes: hypertension,lipiddisorders,obesity,andnophysicalexercise. CONCLUSIONS: Wefoundnoassociationbetweennewonsetdiabetesandantidepressantuseand conrmedtheassociationbetweenlife-styleriskfactorsandnew-onsetdiabetes. Futurestudiesontheriskofnew-onsetdiabetesbydurationandintensityofantidepressantuseanddepressivesymptomsareneeded. PDB16 DIABETESMELLITUSTYPE2ANDT2DMPLUSHYPERTENSIONINBRAZIL:THE EPIDEMIOLOGICPROFILEOFTHEPOPULATIONREGISTEREDINTHE GOVERNMENTPROGRAMHIPERDIANascibenV,MeloTG BoehringerIngelheimBrazil,SaoPaulo,BrazilOBJECTIVES: Diabetesrepresentsaburdentothehealthcaresystemduetoits increasingratesofincidence,morbidityandmortality.Epidemiologicaldataiscrucialforpolicyanddecisionmakingtoclarifygapsinthecurrenthealthprogram andtodevelopnewones.WedecidedtoreviewthestatusofthepopulationregisteredatHiperdiaprograminordertobetterunderstandtherealsituationofthese diseases. METHODS: HIPERDIAisaprogramformonitoringhypertensive/diabetic outpatientsunderthepublichealthcaresystemcare.Basedonthatdatabase,we searchedthegeneralproleofthispopulationfrom2002until2010focusingon type2diabetes(T2DM)andT2DMplushypertension,andalsogeographicaldistribution,gender,ageandoverweight/obesityprevalencewereraised. RESULTS: From2002to2010,atotalof2,199,972peoplewithT2DMandT2DMplusHypertensionwereregisteredinthedatabase,with12.5%presentingT2DMonlyand87.5% withbothconditions,whichrepresentsabout1.5%ofthetotalpopulationunder thepublichealthcaresystem.ForpatientswithT2DMonlywefound114,836men and159,581women(41.8%and58.2%)andforpatientswithbothconditionswe found619,005menand1,306,550women(32.1%and67.9%).Regardingage,the majorityofthepatientsareinbetween50-70years-old.Almost50%ofthepatients aredistributedwithin5statesSP,MG,BA,PRandRJ.Overweightorobesitywas presentin48%ofthispopulation. CONCLUSIONS: Inaccordancewithepidemiologicaldata,mostofthediabeticpatientsinthisdatabasearemiddle-age,women andalmosthalfofthemareobeseoroverweight.Addedtothat,ourestimates aboutdiabetesprevalenceweredifferentfromthenumberreportedbytheBrazilianMoHat2008(9.7%).A93VALUEINHEALTH14(2011)A1–A214

PAGE 102

PDB17 ASSOCIATIONOFYOUNGERAGEWITHPOORGLYCAEMICANDCHOLESTEROL CONTROLINASIANSWITHTYPE-2DIABETESINSINGAPORETohMPHS,WuCX,HengBH NationalHealthcareGroup,SingaporeOBJECTIVES: TheNationalHealthcareGroupPolyclinics(NHGP)isagroupof9 publicsectorprimarycareclinicsinSingapore.Thisstudyexaminesthefactors associatedwithpoorglycaemiccontrolinAsianpatientswithtype2diabetesmellitus(T2DM)inSingapore. METHODS: Thisisacross-sectionalstudyofpatients withT2DMwhoattendedthesameclinicin2009forthetreatmentofdiabetes. Demographiccharacteristics,medicaldiagnosis,clinicalparametersandlaboratoryresultswereextractedfromthegroup'sDiabetesRegistry(CDMS).Glycaemic (HbA1c)andcholesterol(LDL-c)controlwerecomparedwithageandlogisticregressionanalysiswasappliedtostudythefactorsassociatedwithpoorglycaemic controlusingHbA1ccut-offat8%. RESULTS: Amongthe58,057T2DMpatientswere morefemales(54%),disproportionatelymoreIndians(13%)andfewerChinese (71%)thanthegeneralpopulation.BothHbA1candLDL-cimprovedwithage.The meanHbA1cdecreasedgraduallyfrom8.16 1.74%( 40years)to6.94 0.99%(80 years)whilemeanLDL-cdroppedfrom2.84 0.80to2.56 0.70.TheIndianand MalaygroupshadsignicantlypoorerglycaemiccontrolcomparedtotheChinese, AdjOR1.66(95%CI:1.56-1.77)and1.53(95%CI:1.43-1.63)respectively.Othersignicantpredictorsofpoorglycaemiccontrolincludedthemalegender(AdjOR1.19; 95%CI1.19:1.14-1.25),presenceofmaculopathyorretinopathy,peripheralvascular disease,coronaryheartdisease,heartfailure,andbeingoninsulintherapy(AdjOR 8.00;95%CI:7.54-8.48).PatientswithpoorLDL-c(4.0 mmol/L)were4.2timesthe oddsofhavingpoorglycaemiccontrol(95%CI:3.78-4.66)whilethosewithGrade2 hypertensionwere1.5times(95%CI:1.35-1.76). CONCLUSIONS: YoungerT2DMpatientshadpoorerglycaemicandcholesterolcontrolthanolderpatients.Thosewith poorglycaemiccontrolalsohadcorrespondingpoorercholesterolandbloodpressurecontrol.Thesepatientshadahigherlifetimeriskofdevelopingmicro-and macro-vascularcomplicationsandshouldbetreatedmuchmoreaggressivelyto achieve"optimal"glycaemicandcholesterolcontrol. PDB18 FACTORSASSOCIATEDWITHROUTINEPROPERMONITORINGOFDIABETES CAREAMONGTHENON-INSTITUTIONALIZEDPOPULATIONINTHEUNITED STATES:ARETROSPECTIVEANALYSISOFTHE2007MEDICALEXPENDITURE PANELSURVEY(MEPS)ZhaoY,FonsecaV,ShiL TulaneUniversity,NewOrleans,LA,USAOBJECTIVES: Thisstudyaimedtoexaminetherateandpredictorsofdiabetes monitoringintheUS. METHODS: Thiscross-sectionalretrospectivestudywas conductedonarepresentative,non-institutionalizedsampleoftheUnitedStates population(the2007HouseholdComponent(HC)oftheMEPS).Accordingtothe AmericanDiabetesAssociation(ADA)2007practiceguidelines,properprovider monitoringisdenedasatleasttwoA1ctests,oneeyeandonefootexamination annually.HealthstatuswasmeasuredbySF-12.Alogisticregressionmodelwas usedtoexaminethepredictorsofpropermonitoring.Differencesinhealthstatus andmedicalexpendituresbetweenpatientswithandwithoutpropermonitoring wereexaminedusingt-tests.Estimateswereweightedtothetotalpopulation (WTP). RESULTS: Among1,747(WTP:19,320,394)patientswithdiabetes,80.64% hadatleasttwoA1ctests;63.29%hadaneyeexamination;and67.51%hadafoot examination.Thus,63.36%patients(WTP:14,065,289)receivedproperdiabetes monitoring.Olderpatients(OR:1.021,95%condenceinterval(CI):1.012-1.030), non-HispanicCaucasianscomparedwithAfricanAmericans(OR:1.236,95%CI: 0.933-1.636),patientswithahighereducationlevel(OR:1.211,95%CI:1.056-1.390), insurancecoverage(OR:2.216,95%CI:1.408-3.486),useoforalanti-diabeticdrugs (OR:2.935,95%CI:2.131-4.042)andinsulin(OR:3.453,95%CI:2.477-4.814)weremore likelytoundergothepropermonitoring.Wellmonitoredpatientshadahigher SF-12MentalComponentSummaryscore(50.09 0.37vs.48.51 0.45,p 0.05),buta lowerSF-12PhysicalComponentSummaryscore(39.95 0.34vs.42.28 0.47, p 0.05).Properlymonitoredpatientsspentsignicantlymoreontotalhealthcare services( $5,243),outpatientvisits( $1,023),andmedications( $1,204),respectively(allp-values 0.05). CONCLUSIONS: IntheUnitedStates,nearly40%patients withdiabetesdonotreceivetheproperdiabetesmonitoringaccordingtotheADA guidelines.Inadditiontoracialandsocioeconomicdisparities,anti-diabetics/insulinuse,mentalhealthstatus,physicalhealthstatus,andhealthcareexpenditure wereassociatedwithperformingmonitoring.Cost-benetoflong-termmonitoring shouldbestudied. Diabetes/EndocrineDisorders–CostStudies PDB19 ECONOMICASSESSMENTOFCONVERSIONTOINSULINPENDEVICESINA LONGTERMCAREFACILITYCHAINBazaloG1,WeissRC2,BouchardJ3,PerryR4,WendtF5 1ManagedSolutions,LLC,Conifer,CO,USA,2ManagedSolutions,LLC,Randolph,NJ,USA,3Novo Nordisk,Inc.,Princeton,NJ,USA,4SeniorPharmaStrategies,Hudson,FL,USA,5Senior PharmaStrategies,Burleson,TX,USAOBJECTIVES: Theobjectiveofthisstudywastodeterminetheeconomicimpactof apharmacyprogramtoconvertinsulinutilizationfrommulti-dosevialstopen deliverysystemsonalongtermcarefacilitychain. METHODS: Purchasingdatawas obtainedatthepatientlevelforbasalandshortactinginsulinsfromachainof75 skillednursingfacilitiesforthe12monthperiodendingJune2010.Dataincluded datedispensed,amountdispensed(mls),deliverysystem(penorvial)andamount paidtothedispensingpharmacy.Theinsulincostperpatient-dayforeachmonth wascalculatedastotalacquisitioncostforthemonthdividedbythenumberof patient-days.Theinsulincostperpatient-dayforeachstaywascalculatedasthe totalinsulinacquisitioncostdividedbythelengthofstayindays.Themeancost perpatient-dayforeachpatientstaysubsetbasedonpayertype,lengthofstayand deliverysystemused(penonly,vialonly,penandvialcombination)was calculated. RESULTS: Therewere2,405inpatientstaysoverthe12monthperiod, 70%coveredbyMedicareand29%byManagedCare.Two-thirdsofMedicarestays andoverthree-fourthsofmanagedcarestayswere30daysorless.Pendevice purchasesincreasedfromunder1%toalmost35%oftotalpurchasesoverthestudy periodduringwhichtheinsulincostperdaydeclinedfromover$10perpatient-day to$4.Thecostperdayforvial-onlystays($7.84)andcombinationvialandpenstays ($7.79)were72%higherthanpen-onlystays($4.54),despitea39%pricepremium permilliliterforpens.Differencesweremostmarkedforlengthsofstayunder30 days. CONCLUSIONS: Theincreaseinpendeviceusewasassociatedwithamarked decreaseininsulincostsonapatient-daybasis,particularlyforlengthsofstay under30days. PDB20 BUDGETIMPACTANALYSISOFTHEINTRODUCTIONOFSAXAGLIPTININTHE TREATMENTOFTYPE-2DIABETESINCHILEElgartJ1,CaporaleJ2,AielloEC3,GagliardinoJJ4,WaschbuschM3,JotimlianskyL3, ValenciaJE5 1NationalUniversityofLaPlata,LaPlata,BuenosAires,Argentina,2InstituteforClinical EffectivenessandHealthPolicy,BuenosAires,Argentina,3Bristol-MyersSquibb,BuenosAires, Argentina,4CENEXA,LaPlata,BuenosAires,Argentina,5Bristol-MyersSquibb,Bogota, ColombiaOBJECTIVES: ToestimatethebudgetimpactofSaxagliptinintroductionasatreatmentoptionforpatientswithtype2diabetesmellitus(DM2)comparedtothe presentsituation. METHODS: AnMSExcel-basedbudgetimpactmodelassuming coverageforonemillionpeople.Thetimehorizonwasthreeyearsandtheanalysis perspectivewasthatofthepublichealthcaresysteminChile.Pharmaceutical expensesofantidiabeticagentswereanalyzed,excludingothermedicalcosts.The costofantidiabeticagentswasbaseduponlistpricesadjustedtoco-payments, expressedin2009USdollars;theSaxagliptinpricewasconsideredtobeequalto thesitagliptinprice.ThemarketshareofthedifferentdrugswasbaseduponmarketstudiesanddataprovidedbyBristolMyersSquibb.Thebudgetimpactisreportedintermsofannualbudgetimpact,permemberpermonth(PMPM).Thecost ofpioglitazoneandrosiglitazonerelatedcardiovascularevents,aswellasthatof sulphonylureasrelatedhypoglycemiaeventswereexpressedasratesofoccurrenceperpatientperyearandcostperoccurrence. RESULTS: Theestimatednet budgetimpactfortheintroductionofSaxagliptinwasUS$70,723,US$162,885and US$251,574fortherst,secondandthirdyearrespectively;thecumulativenet budgetimpactwasUS$485,181.PMPMwasUS$0.0059,US$0.0136andUS$0.0209 eachyear,respectively.Thecumulativeimpactinthetotalannualbudgetforantidiabeticsrepresentedanincreaseof4.22%. CONCLUSIONS: Thebudgetimpactof addingSaxagliptininapopulationofonemillionpeopletothepublichealthcare systeminChileisminimalinpatientswithDM2.TheriseinpharmaceuticalexpensesderivedfromintroducingSaxagliptinintotheformularyisbalancedby savingsintermsofreductionofadverseeventsrelatedtothiazolidinedionesand sulfonylureas,aswellofloweringofinsulinrequirementsinanextendedtime horizon. PDB21 ACOSTCOMPARISONOFABASALBOLUSREGIMEN(INSULINGLARGINEAND INSULINGLULISINE)WITHACONVENTIONALPRE-MIXEDINSULINREGIMENIN TYPE-2DIABETESPATIENTS–THEGINGERSTUDYTillingC1,VoraJ2,KeechM3 1Sano-Aventis,Guildford,Surrey,UK,2RoyalLiverpool&BroadgreenUniversityHospitals, Liverpool,UK,3Pharmakos,StAlbans,Hertfordshire,UKOBJECTIVES: Thiscostanalysis,basedontheresultsoftheGINGERstudy,aimedto investigatewhetheranintensiedinsulinregimenisbettervaluetha na2injection perdayconventionalregimen. METHODS: GINGERwasa52weekmulti-national studyin310T2Dpatientsoninsulinforanaverageof5yearswithpoorglycaemic control.Itcomparedmealtimerapid-actinginsulinglulisine(IGL)andinsulin glargine(IG)oncedailywith2injectionsperdayofpre-mixedinsulin.UseofIGL/IG resultedinachangeofHbA1Cfrombaselinetoendpointof 1.31%and 0.80%for pre-mixedinsulin.CostswerecalculatedfromaUKNHSperspectiveusingMIMS November2010prices.InsulincostswerebasedontheuseofIGL/IG(ApidraSoloStarandLantusSoloStar)andbiphasicinsulinaspart(BIA,NovoMix30FlexPen) prelleddisposableinjectiondevices.Itwasassumedthatanewneedle,lancetand bloodglucoseteststripwereusedforeachinjectionwitha2Uprimingdoseof insulinbeforeeachinjection. RESULTS: TheannualdrugcostperpatientonIGL/IG washigherthanBIAat£692and£612respectivelywiththecostofmetformin similarforbothgroups.Thecostofneedles,lancetsandteststripswasmuchlower forBIAat£329comparedwith£537fortheIGL/IGgroup.Overallthetotalannual costperpatientfortheIGL/IGgroupwas£1243comparedwith£957forBIA.Over the52weekstherelativecostofa1%reductioninHbA1Cwas£949forIGL/IGand £1197forBIA,a1mmol/lreductioninFPGwas£518withIGL/IGand£563withBIA. SensitivityanalysesreplacingBIAbyinsulinlisproorisophaneinsulingavevery similarresults. CONCLUSIONS: AsimilarreductioninHbA1CandFPGcanbe achievedatarelativelylowercostwithIGL/IGincomparisonwithBIA. PDB22 INSULINGLARGINEPLUSOHASVERSUSBIPHASICINSULININTYPE-2 DIABETE S–ACOSTCOMPARISONTillingC1,OwensD2,KeechM3 A94VALUEINHEALTH14(2011)A1–A214

PAGE 103

1Sano-Aventis,Guildford,Surrey,UK,2UniversityHospitalLlandough,Penarth,UK,3Pharmakos,StAlbans,Hertfordshire,UKOBJECTIVES: AsType2Diabetes(T2D)progressesoralhypoglycaemicagents (OHAs)alonefailtomaintainbloodglucosecontrolandinsulinisrequired.LAPTOP wasamulti-nationalstudycomparingtheadditionofoncedailyinsulinglargine (IG,LantusSolostar)toglimepirideandmetforminwithtwicedaily30%regular/ 70%humanNPHinsulin(Actraphane30)withoutanyadditionalOHAsin371T2D patientsovera24weekperiod.IGplusOHAswassignicantlymoreeffectivein loweringHbA1C( 1.64%vs 1.31%).Acostminimisationanalysiswasundertaken usingLAPTOPdatatocomparethecostsofthe2regimens. METHODS: Theanalysis wasundertakenfromaUKNHSperspectivewithpricesfromMIMSNovember 2010.Costestimateswerebasedontheuseofnon-proprietaryOHAsandIGand biphasicinsulinaspart(BIA,NovoMix30FlexPen)prelleddisposableinjection devices.Itwasassumedanewneedle,lancetandbloodglucoseteststripwereused foreachinjectionwitha2Uprimingdoseofinsulinbeforeeachinjection. RESULTS: Costswerecalculatedoverthe24weekstudyperiod.Thetotalcostofdrugsper patientonIGplusOHAswasslightlylowerthanBIAat£130and£167respectively despitenoOHAuseintheBIAgroup(nalmeaninsulindosesat24weeksIG28.2U andBIA64.5U).Thecostofneedles,lancetsandteststripswasmuchlowerforIG plusOHAsat£76comparedwith£152fortheBIAgroup.Overallthecostperpatient for24weeksfortheIGplusOHAsgroupwas£206comparedwith£319forBIA,a differenceof35%.Sensitivityanalysesreplacingdisposablewithreusablepensand BIAbyotherbiphasicinsulinsgavesimilarresultstothebasecase. CONCLUSIONS: IncomparisonwithbiphasicinsulinthecostofIGplusOHAswas35%lessto achieveanequivalentreductioninHbA1C. PDB23 THEIMPACTOFTREATMENTMODIFICATIONONHEALTHCAREEXPENDITURE INPATIENTSWITHTYPE-2DIABETESINITIATINGEXENATIDEBIDORINSULIN GLARGINEPawaskarM1,BonafedeM2,JohnsonBH3,FowlerR3,HoogwerfBJ1 1EliLillyandCompany,Indianapolis,IN,USA,2ThomsonReuters,Cambridge,MA,USA,3ThomsonReuters,Washington,DC,USAOBJECTIVES: Toexaminetheimpactoftreatmentmodicationsonhealthcare expenditureforpatientswithtype2diabetes(T2D)initiatingexenatideBID(exenatide)orinsulinglargine. METHODS: AretrospectivedatabaseanalysiscomprisedofadultpatientswithT2Dwhoinitiatedexenatide(N 9197)orinsulin glargine(N 4499)between10/01/2006and03/31/2008with12monthspre-and18 monthspost-indexcontinuousenrollment.The2cohortswere1:1propensity scorematchedonbaselinedemographics,clinical,andresourceutilizationvariables.Treatmentmodicationwasdenedasthersteventintensication, switching,ordiscontinuationoftheindexmedication.Themeanhealthcareexpendituresinallpatientsandthoseexperiencingtreatmentmodicationswere estimatedusingmultivariateregression. RESULTS: Baselinecharacteristicsofpatientsonexenatide(n 3774)andglargine(n 3774)werewellbalancedafter matchingwithmeanageof57years,meanDeyoCharlsonComorbidityscoreof 1.6,andproportionatelymoremales(54%)inbothcohorts.Glargine-treated patientswere33%morelikelytomodifytheirtreatmentthanexenatide-treated patients(HazardRatio[HR]:1.33,p 0.0001).Comparedtoexenatide-treated patients,glargine-treatedpatientsweremorelikelytointensify(HR 1.72, p 0.0001),ordiscontinuetheirtreatment(HR 1.25,p 0.0001),butlesslikelyto switchtonewtherapy(HR 0.71,p 0.0001).Meanhealthcareexpenditurewas signicantlylowerforexenatidecomparedtoglargineafter18months(difference $1667,p 0.0001).Exenatide-treatedpatientswhocontinuedtheirtreatmenthadsignicantlylowerexpenditureof$1546(p 0.005)andthosewho intensiedtheirtherapyalsohadsignicantlylowerexpenditureof$2472 (p 0.001).Therewerenosignicantdifferencesinexpenditureforpatientswho switchedordiscontinuedtheirtherapy. CONCLUSIONS: Thelikelihoodoftreatmentmodicationandmeantotalhealthcareexpenditurevariedforpatientsinitiatingexenatideorglargineinreal-worldsettings.Exenatide-treatedpatientshad lowermeanhealthcareexpenditurecomparedtoglargine-treatedpatients.Exenatide-treatedpatientswhocontinuedorintensiedtheirtherapyalsohadassociatedreductioninmedicalexpenditure. PDB24 DIRECTHEALTHCAREANDINDIRECTWORKLOSSCOSTSASSOCIATEDWITH THEADDITIONOFROSIGLITAZONE(RSG)VERSUSSITAGLIPTIN(STG)THERAPY TOMETFORMIN(MET)LefebvreP1,SardaSP2,LalibertŽF1,RamamurthyP2,WeiR2,ArondekarB3,MendittoL3, MartinAA4 ,DuhMS2 1Grouped'analyse,LtŽe,Montreal,QC,Canada,2AnalysisGroup,Inc.,Boston,MA,USA,3GlaxoSmithKline,Philadelphia,PA,USA,4GlaxoSmithKline,Uxbridge,UKOBJECTIVES: Wecomparedhealthcareresourceutilization(HRU)andcostsassociatedwithadd-ontherapyofRSGversusSTGtoMET. METHODS: TypeIIdiabetes mellituspatients, 18years,initiatingRSGorSTG(rstdispensing indexdate) add-ontherapywithMETwereidentiedinthePharMetricsdatabase(1999-2008). Patientswerecontinuouslyenrolledfor 6monthspre-index(baseline)and12 monthspost-index,had 1dispensingforMETinthe6-monthpre-indexperiod, didnotuseinsulinorsulfonylurea,andweretreatedwithRSG METbefore05/01/ 2007orSTG METfor 6monthspost-index.All-causeanddiabetes-relatedHRU andannualcosts($2008)werereportedforthe12-monthfollow-upperiod.WorklosscostswereestimatedbyapplyinghourlywagefromtheBureauofLaborStatisticstomissedworkhours(hospitalization 8hours;outpatient/emergencyroom [ER]visit 4hours).Multivariateanalyseswereconducted,adjustingforbaseline demographics,comorbidities,andcosts. RESULTS: ComparedtoSTG METcohort (N 1,660)atbaseline,RSG METcohort(N 3,731)wasyounger(55vs.58years) withfewercomorbidities(CharlsonComorbidityIndex[0.26vs.0.34]),andhad lowertotalcosts(RSG MET:$7,875;STG MET:$9,412;costdifference $1,536, p 0.0043).FrequencyandcostsofhospitalizationsandERvisitswerenotdifferent atbaseline.Inthe12monthspost-indexperiod,all-causeHRUandcorresponding annualcostdifferencebetweenRSG METandSTG METcohortswasenlarged (RSG MET:$8,443;STG MET:$10,757;costdifference $2,314,p 0.0001).Afteradjustingforcovariates,thecostdecreaseassociatedwithRSG METremainedstatisticallysignicant(costdifference $1,248;costratio 0.87; P 0.0120).DiabetesrelatedadjustedincrementalcostsavingofRSG METoverSTG METwas$599 (costratio 0.83, P 0.0160).Theadjustedworklosscostwasalsolowerfor RSG METcomparedtoSTG MET(costdifference $22,costratio 0.93; P 0.0120). CONCLUSIONS: ComparedtothenewDPP-4agentSTGcombinedwithMET,RSG,a thiazolidinedione,combinedwithMETwasassociatedwithlowerall-causeand diabetes-relateddirecthealthcarecostsandindirectworklosscosts. PDB25 COSTOFALL-CAUSEANDCARDIOVASCULARDISEASE-RELATED HOSPITALIZATIONINPATIENTSWITHTYPE-2DIABETESTREATEDWITH EXENATIDEBID,SULFONYLUREAS,ORINSULIN:ARETROSPECTIVEANALYSIS OFTHELIFELINKDATABASEBestJH1,PelletierE2,SmithDB2 1AmylinPharmaceuticals,Inc.,SanDiego,CA,USA,2IMSHealthIncorporated,Watertown,MA, USAOBJECTIVES: Toassessthecostofall-causeorcardiovasculardisease-related hospitalizationinarealworldsettingamongpatientswithtype2diabetes prescribedexenatide,aGLP-1receptoragonist,comparedtopatientstreated withsulfonylureasorinsulins. METHODS: Analysesincludedpatientsinthe LifeLinkdatabaseinitiatinganewprescriptionforaglucose-loweringagent betweenJune1,2005andMarch31,2009,withoutaprescriptionforthesame agentintheprior9months.Patientswerefollowedfor12months.Intentionto-treatanalysesofcostsofall-causeandcardiovasculardisease-relatedhospitalization(myocardialinfarction,stroke,orcoronaryrevascularization),adjustedfordifferencesin 300clinicalanddemographiccharacteristics,were comparedusingpropensity-score-weightedmethods. RESULTS: Meanage(SD) was53(8)forpatientstreatedwithexenatide(n 14,400),55(12)forpatients treatedwithsulfonylureas(n 43,242),and52(12)forpatientstreatedwith insulin(n 17,627);43.4%,53.5%,and49.9%ofpatientstreatedwithexenatide, sulfonylureas,andinsulinweremale,respectively.Morepatientstreatedwith exenatidewereobese(17.0%)thanpatientstreatedwithsulfonylureas(7.4%)or insulins(10.5%).Patientstreatedwithexenatideweremorelikelytohavehyperlipidemia(67.9%)thanpatientstreatedwithsulfonylureas(49.9%)orinsulins (50.2%).Patientstreatedwithexenatidehadsignicantlylowermean(SD)allcausehospitalizationcoststhanpatientstreatedwithsulfonylureas($2,725 16,463Exenatide;$3,304 17,378sulfonylureas;p 0.001)orinsulin($3,211 17,969Exenatide;$4,849 21,110insulin;p 0.001).Patientstreatedwithexenatidealsohadsignicantlylowermean(SD)cardiovasculardisease-related hospitalizationcoststhanpatientstreatedwithsulfonylureas($428 6,174 Exenatide;$566 6,047sulfonylureas;p 0.05)orinsulin($470 6,841Exenatide;$726 7,012,insulin;p 0.001). CONCLUSIONS: Exenatidetreatment wasassociatedwithsignicantlylowercostsforall-causeandcardiovascular disease-relatedhospitalizationcomparedtotreatmentwithsulfonylureasor insulin. PDB26 ECONOMICEVALUATIONOFGLIMEPIRIDEANDGLIMEPIRIDE/METFORMINFOR TYPE-2DIABETESMELLITUSINMEXICOCarlosF1,LemusA2 1RACSaludConsultoresS.A.deC.V.,MexicoCity,Mexico,2Sano-Aventis,MexicoCity,MexicoOBJECTIVES: Toperformaneconomicevaluationoftheuseofgliperimide(GMP) andthexed-dosecombination(FDC)ofglimepiride/metformin(GMP/Met)inthe treatmentofpatientswithDM-2,fromtheMexicanPublicHealthSystem perspective. METHODS: TwoMarkovmodelsweredesignedforreectingdifferent treatmentsequences.Model1isfornotcontrolleddietandexerciseandmetformin intolerantpatients,whereGMPiscomparedtoglibenclamide(GBC)andthiazolidinedione(TZD).Model2isforpatientswhichthelifestylechangedandthemonotherapywithmetforminwasinsufcientforreachingalevelofHbA1c 7%,anda sulfonylurea/MetFDCiscomparedtoTZD/Met.Thedirectcostsofthedrugtreatmentandthehypoglycemiaepisodewerecalculated.Thedosageandefcacyof theoralantidiabeticagentsandinsulinwereestablishedbasedontheliterature reviewandlocalpractice.Theunitcostswereelicitedfromofcialsources.The timehorizonwasthreeyears,dividedinquarterlycycles. RESULTS: Therapywith GMPwasdominantversusthetreatmentsequenceswhichincludedabeginning therapywiththiazolidinedione.ThesavingsafterthreeyearswithGMPorGMP/ MetinsteadofusingTZDrangedfromUS$288.7toUS$632.1perpatient.Although GBChasalowcost,itisassociatedwithanincreaseinthemortalityrateand hypoglycemiaevents.AFDCofGMP/Metcausedanincrementalcostof20%versus GBCandmetforminseparately.Theaveragecostsperadditionallifeyearobtainby usingGMPinsteadofGBCrangedfromUS$3,074.9toUS$3,261.4.Theprobabilistic sensibilityanalysisshowsthatGMPandGMP/Metresultedashighlycost-effective inapproximately90%ofthesimulations. CONCLUSIONS: Glimepiridemonotherapy(model1)andtheadministrationofGMP/Met(model2)representhighly cost-effectivehealthinterventionsregardingtheuseofglibencamideanditisdominantversustheuseofthiazolidinedione.A95VALUEINHEALTH14(2011)A1–A214

PAGE 104

PDB27 COST-EFFECTIVENESSANALYSISOFMETFORMINCOMBINEDWITH SAXAGLIPTINVERSUSMETFORMINCOMBINEDWITHSULFONYLUREASIN TYPE-2DIABETESPATIENTSINCHILEElgartJ1,CaporaleJ2,AielloEC3,WaschbuschM3,JotimlianskyL3,GagliardinoJJ4, ValenciaJE5 1NationalUniversityofLaPlata,LaPlata,BuenosAires,Argentina,2InstituteforClinical EffectivenessandHealthPolicy,BuenosAires,Argentina,3Bristol-MyersSquibb,BuenosAires, Argentina,4CENEXA,LaPlata,BuenosAires,Argentina,5Bristol-MyersSquibb,Bogota, ColombiaOBJECTIVES: Todeterminethecost-effectivenessrelationofaddingSaxagliptinto Metformintherapy(SAXA MET)comparedtoaddingSulfonylureas(SULF MET), inpatientswithtype2diabetesmellitus(DM2)whohavefailedtoachieveadequate glycemiccontrolwithmetformin. METHODS: Adiscreteeventsimulationmodel (Cardifflongtermcost-utilitymodel)withaxedtimeincreasebasedonUKPDS68 wasusedtosimulatediseaseprogressionandtoobtainanestimateofthetreatment'seconomicandhealthconsequencesinDM2patients.ClinicalefcacyparametersforSaxagliptinwereobtainedfromtheliterature;drugacquisitioncosts, adverseevents(AEs)andmicrovascularandmacrovascularcomplicationswere takenintoaccount.Thetimehorizonwas20yearsandtheperspectivewasthatof thepublichealthcaresysteminChile.CostswereexpressedinUSdollars(2009), withanannual3.5%discount. RESULTS: Alowernumberofnon-fataleventswere foundfortheSAXA MET-treatedgroupversustheSULF MET-treatedgroup.Additionally,themodelpredictedalowernumberoffatalmacrovascular(132.3vs. 136.0)andmicrovascular(19.6vs.19.7)eventsfortheSAXA MET-treatedgroupvs. theSULF MET-treatedgroup.ThetotalcostoftheSAXA METcohortwaslower thantheSULF METcohort:US$15.006.011andU$S14.557.581,respectively.TreatmentwithSAXA METresultedinahighernumberofQALYs(9,794vs.9,594)and LYs(23,068vs.23,019)forthe1000patients'cohortthantreatmentwithSULF MET; theincrementalanalysisperQALYandLYgainedwas-US$2,243andUS$ 9,182 respectively(dominant). CONCLUSIONS: ResultssuggestthattheadditionofSaxagliptintometformintherapyisdominantcomparedtotheadditionofsulfonylureas;therefore,thisinterventionwouldrepresentanefcientuseofhealthresourcesforDM2patientsinChile. PDB28 COSTSAVINGSINTYPE-2DIABETES(T2D)WITHINSULINGLARGINE COMPAREDWITHINSULINDETEMIRINTHEUNITEDKINGDOMTillingC1,EvansLM2,KeechM3 1Sano-Aventis,Guildford,Surrey,UK,2UniversityHospitalLlandough,Penarth,UK,3Pharmakos,StAlbans,UKOBJECTIVES: Theobjectivewastodeterminewhethertheuseofinsulinglargine (IG,Lantus)couldresultincostsavingsincomparisonwithinsulindetemir(ID, Levemir)inT2DpatientsintheUK. METHODS: Clinicaldatafortheanalysiswas takenfro ma1yearmulti-nationalstudy(Rosenstock2008)in582insulin-na•ve patientstreatedwithIGorIDasadd-ontherapytooralglucoseloweringdrugs.IG wasadministeredoncedailyinlinewithit'slicenceandIDonce(44%ofpatients)or twicedaily(55%)accordingtotreatmentresponse.Meandailyinsulindoseswere 40UforIGand71UforID.Baselinecharacteristicswerecomparableonstudyentry. After1yearglycaemiccontrol,weightgain,adverseeventsandriskofhypoglycaemiaweresimilarforbothgroupssoacostminimisationanalysiswasundertaken. CostswerecalculatedfromaUKNHSperspectiveusingMIMSNovember2010 prices.InsulincostwasbasedonIGSoloStarandIDFlexPenprelleddisposable injectiondevices,whichareofsimilarcost.Itwasassumedthatanewneedle, lancetandbloodglucoseteststripwereusedforeachinjectionwitha2Upriming doseofinsulinbeforeeachinjection. RESULTS: Theannualcostperpatientfor needleswas£43forIGand£66forID.Thecostoflancetsandteststripswas£122for IGand£190forID.Theannualcostofinsulinwas£426forIGand£758forID.The totalannualcostperpatientofadministeringIGwas£591comparedwith£1014for ID.Sensitivityanalysesforweightgain,IGdeviceandinsulindoseconrmedthe robustnessofthebasecaseresults. CONCLUSIONS: IGmayrepresentamorecost effectiveoptionforT2DpatientsintheUKrequiringbasalinsulinanaloguetherapy withpotentialannualcostsavingsof42%(£423/patient)comparedwithID. PDB29 DIRECTMEDICALCOSTOFHYPOGLYCEMIAAMONGPATIENTSWITHTYPE-2 DIABETESINTHEUNITEDSTATESCurkendallS1,ZhangB2,OhK1,WilliamsS3,PollackM3,GrahamJP2 1ThomsonReuters,Washington,DC,USA,2Bristol-MyersSquibb,Plainsboro,NJ,USA,3AstraZenecaPharmaceuticalsLP,Wilmington,DE,USAOBJECTIVES: Hypoglycemiaisacommontreatmentrelatedsideeffectofdiabetes, whichcanbeassociatedwithfrequentemergencyroomuseandhospitalization. Theobjectiveofthisstudywastoassessthedirectmedicalcostsofindividual hypoglycemiaeventsinaUSpopulationoftype2diabetes. METHODS: Patients 18yearswithtype2diabetesdiagnosedduringtheperiod2003to2008were selectedfromthe MarketScan¨ResearchDatabases andfollowedfromtheirrst diabetesdiagnosisinthestudyperioduntiltheendofcontinuouscoverageor December31,2008,whichevercamerst.Allindividualhypoglycemiaeventsidentiedbyaclaim(ICD-9-CM250.3,250.8,251.0,251.1,251.2)onauniquedatewere countedforeachpatient.Directmedicalcostswerecalculatedperhypoglycemia eventbytreatmentsettinganddiabetesdrugregimen. RESULTS: Atotalof2.4 milliondiabetespatientswereselected.Duringfollow-up,91,595ofthesepatients experienced169,248hypoglycemiaevents.Costswerehighestforhypoglycemia eventsidentiedintheemergencyroom(ER)-to-inpatientsetting($10,362/event), followedbyinpatient($7,317/event),ER($701/event),andoutpatient($285/event) settings.Patientstreatedwithaninsulinonlyregimenhadthehighestdirectmedicalcostsofhypoglycemia($9.49perpatientpermonth).Amongpatientstreated withnon-insulinregimens,theestimatedhypoglycemiacostsforpatientsonsulfonylureaswere$3.87perpatientpermonthand$0.84forotherOADs. CONCLUSIONS: Directmedicalcostsassociatedwithhypoglycemiaamongtype2 diabetespatientsvariedbytreatmentsettinganddrugregimen.Patientstreated withinsulinhadthehighestdirectmedicalcosts.Withinnon-insulinregimens, patientsonsulfonylureashadhighercoststhanthoseonnon-sulfonylureaOADs. Treatmentstrategiesthatprovideeffectiveglycemiccontrol,withalowerpotential forinducinghypoglycemiashouldbeconsideredinthemanagementoftype2 diabetes. PDB30 ACROSS-SECTIONALSURVEYONECONOMICBURDENOFTYPE-2DIABETES MELLITUSPATIENTSWITHORALANTI-DIABETICDRUGSTHERAPYINCHINAJiL1,YeQ2,ChangJH2,LiuGG3 1PekingUniversityPeople'sHospital,Beijing,China,2NovoNordisk(China)PharmaceuticalsCo., Ltd.,Beijing,China,3PekingUniversity,Beijing,ChinaOBJECTIVES: Thestudyassessedtheeconomicburdenincurredbytype2diabetes mellitus(T2DM)patientswithoralanti-diabeticdrugs(OADs)therapyinChina. METHODS: Across-sectionalsurveywasconductedbetweenDec3rd,2008andJuly 31st,2009at75urbanhospitalsin9citiesinChina.Atotalof9577T2DMpatients withOADstherapycompletedtheself-administeredquestionnairesduringtheir clinicalvisits.Theinformationcollectedfromquestionnaireincludeddemographiccharacteristics,utilizationofOADs,glycemiccontrol,diabetescomplications,annualdirectmedicalcostsandproductivitylostassociatedwithT2DM. RESULTS: Themeanage( SD)was59.5 12.7years,51.1%weremaleandthemean diabetesduration( SD)was7.9 6.3years.Only7.9%and15.3%ofthepatients achievedHbA1c 6.5%,andHbA1c 7.0%respectively.Theaveragetotalmedical expenditureswasCNY9127.2peryear,ofwhichCNY6478.5wasspentonT2DM. TheannualexpenditureofOADswasCNY4279.4,accountingfor66.1%ofthetotal annualT2DMtreatmentexpenditure.Theannualaveragediabetes-relatedcosts foroutpatientandhospitalizationwereCNY5998.9and4265.8respectively.T2DM patientsdiagnosedlessthan3years,3-5years,5-10yearsandmorethan10years hadanannualaveragetreatmentcostofCNY4783.0,5829.5,6248.4,and8183.8 respectively.TheaverageannualproductivitylostduetoT2DMwas29.9working days.ItresultedinanestimatedaverageopportunitycostofCNY2367.1percapita (The2008averageannualincomeofurbanresidentswasCNY28898).Thedirect treatmentcostofT2DMandopportunitycosttogetheraccountedfor56%ofpatients'disposableincomein2008(CNY15781). CONCLUSIONS: ThestudydemonstratedthehigheconomicburdenandinadequateglucosecontrolamongT2DM patientswithOADstherapy.ThedirectmedicalcostofT2DMincreaseswithdiseaseprogression.Ithighlightstheimportanceofimplementingcost-effectivetherapeuticregimenstoimprovediabetesmanagement. PDB31 BENEFITSANDCOSTSOFCONDUCTINGSPONSOREDCLINICALTRIALSINA PUBLICLYFUNDEDNEWZEALANDHOSPITAL:PRE-TRIAL,TRIALAND FOLLOW-UPMurphyL1,MaquireW2 1UniversityofTasmania,ManukauInstituteofTechnology,Auckland,NewZealand,2University ofTasmania,Hobart,Tasmania,AustraliaOBJECTIVES: (1)Toidentifyandanalysethebenetsandcostsofsponsoredclinical trialsinapubliclyfundedNewZealandhospital,fromtheperspectiveof(a)the researchunitand(b)thedistricthealthboard;(2)Toidentifypatternsofcostsand benetsduringthepre-trial,trialandfollow-upstagesofaclinicaltrial. METHODS: Wedrawonthedatarelatingtocasesandcontrolsinahealthoutcomesstudyof twophaseIIIclinicaltrialstoassesspreventativemedicationforpatientsatriskof seriouscardiovascularevents.ThedataincludeMinistryofHealthdata,TheCentre forClinicalResearchandEffectivePractice(CCRep)protandlossstatements, CountiesManukauDistrictHealthBoard(CMDHB)ChronicCareManagementdata, trialprotocolsanddatafromatrialhealthoutcomeco-study.Weidentifypatterns ofcostsandbenetsduringtheenrolment,trialandfollow-upstagesofaclinical trial. RESULTS: Theresearchunitandthehealthboardbothderiveeconomicbenetsfromthesetrials.Themagnitudeofthebenetsdiffersdependingonthe perspectivetaken.Theresearchunithastotaleconomicbenetsrangingfrom US$470andUS$846perparticipantwhichitisabletoreinvestintootherresearch projects.Wendthebenetsfromtheperspectiveofthehealthboardarepositive butsmallUS$15-$45perparticipantovertheeightyearperiod.Savingsfrom treatmentcostavoidancearelowerthananticipated.Wendtwoperiodswhere costsarehighest(1)duringenrolmentand(2)attheendofthetrial. CONCLUSIONS: Theremaybeeconomicadvantagesinactivelyencouragingsponsoredclinical trialswithinNewZealandpublicallyfundedhospitals.Thisstudyprovidesevidenceofdifferingpatternsofcostsandbenetsduringthethreestagesofaclinical trialenrolment,trialandfollow-up. PDB32 ECONOMICEVALUATIONOFRANIBIZUMABFORTHETREATMENTOFDIABETIC MACULAREDEMAINCANADAHaigJ1,LawrenceD1,BarbeauM2,BlouinJ2,CruessA3 1i3Innovus,Burlington,ON,Canada,2NovartisPharmaceuticalsCanada,Inc.,Dorval,QC, Canada,3DalhousieUniversity,Halifax,NS,CanadaOBJECTIVES: Diabeticmacularedema(DME)isanophthalmologiccomplication thatdevelopsinasubsetofpatientswithdiabeticretinopathyandthatcausesloss offunctionality,reducedhealth-relatedqualityoflife,visualimpairment,andif leftuntreated,blindness.Acost-effectivenessanalysiswasconductedtomeasureA96VALUEINHEALTH14(2011)A1–A214

PAGE 105

andcomparetheeconomiccostsandclinicaloutcomesassociatedwithranibizumabmonotherapyversuslaserphotocoagulationaloneforthetreatmentofDME inCanada. METHODS: Cost-effectivenessofranibizumabtoaCanadianhealthcare systemwasanalyzedusingaMarkovmodelbasedondatafromtheRESTOREclinical trial.Afteroneyearoftreatment,theRESTOREtrialreporteda6.1(p 0.0001)letter improvementinbestcorrectedvisualacuity(BCVA)withranibizumabcomparedto laserphotocoagulation.Themodelincluded8healthstatesasdenedbyBCVA,and followedacohortof63yearoldpatientswithDMEduetoeithertype1ortype2 diabetesoveratimehorizontoamaximumof20years.Year1healthstatetransitions andallhealthstateutilitieswerederivedfromtheRESTOREtrial.Fromyear2onwards, healthstatetransitionswerebasedonxedprobabilitiesofimprovingorworsening BCVA.ResourceuseandcostswerecollectedfromtheRESTOREtrial,publishedliterature,orstandardCanadiansources. RESULTS: Comparedtolaserphotocoagulation,patientsreceivingranibizumabmonotherapyaccruedanadditional1.44years withoutlegalblindness(BCVA 35letters).Also,patientsreceivingranibizumab accruedanadditional0.19quality-adjustedlifeyears(QALYs)withatotalincrementalcostofapproximately$8,500,resultinginapproximately$44,000perQALY gained. CONCLUSIONS: Comparedtolaserphotocoagulation,ranibizumabmonotherapyshowscost-effectivenesswithincommonlyacceptedcostperQALY thresholds.Inaddition,thisanalysispredictstheuseofranibizumabforDMEwill resultinmoreyearswithoutlegalblindness. PDB33 EVALUATINGTHECOST-EFFECTIVENESSOFSWITCHINGFROMINSULIN GLARGINETOINSULINDETEMIRINPATIENTSWITHTYPE-2DIABETESINA CHINESESETTING:AMODELINGSTUDYBASEDONTHEPREDICTIVESTUDYYangL1,LayAL2,ChangJH3 1PekingUniversity,Beijing,China,2NovoNordiskA/S,Copenhagen,Denmark,3NovoNordisk (China)PharmaceuticalsCo.,Ltd.,Beijing,ChinaOBJECTIVES: Toevaluatethelong-termhealthandeconomicoutcomesofInsulin Detemir(IDet)therapyinuncontrolledpatientswithtype2diabetesmellitus (T2DM)switchingfrominsulinGlargine(IGlar)intheChinesesetting. METHODS: ThepublishedandvalidatedCOREDiabetesModelwasusedtomakethelong-term (30years)projectionofhealtheconomicoutcomes.ThepatientdemographicinformationandclinicalendpointswerederivedfromSouthKoreasub-analysisof thePREDICTIVEtrial.Thestudywasalarge,multi-centre,6monthsobservational studyassessingthesafetyandefcacyofIDet.HbA1cwasreducedof 0.2% (p 0.05)byswitchingfromIGlartoIDet.BaselineriskfactorsandracialcharacteristicdatawereobtainedfromChinesecohortstudies.Themarketretailpricesof medicationswerecalculatedtoestimatetreatmentcosts.ThediabetesmanagementandcomplicationscostswereobtainedfromChinesepublisheddataand adjustedto2010valuesusingtheChineseConsumerPriceIndex.Anannualdiscountingrateof3%wasusedforbothhealthandeconomicoutcomes.One-way sensitivitiesanalysiswasperformed. RESULTS: ThetreatmentofIDetconverted fromIGlarwasprojectedtoreducethecumulativeincidencesofDM-relatedcomplications.Backgroundretinopathy,end-stagerenaldisease,ulcer,myocardialinfarctioneventswerereduced0.685%,0.167%,0.243%,0.482%respectively.Therapy conversiontoIDetwasprojectedtoimprovelifeexpectancyby0.061year,andwas associatedwithimprovementsof0.484qualityadjustedlifeyear(QALY).Thecosts ofcomplicationswerereducedbyCNY5,595,resultinginatotaldirectmedicalcost savingofCNY2,869. CONCLUSIONS: TherapyconversionfromIGlartoIDetin T2DMpatientscoulddelaytheonsetofdiabetescomplications,wasassociated withimprovementsinlifeexpectancyandQALYs,andreduceddirectcostover patientlifetimes.IDetwasshowntobeadominanttreatmentoptioninpatients withT2DMinadequatelycontrolledwithIGlarinChinesesetting. PDB34 PHARMACOECONOMICEVALUATIONOFAPHARMACIST-MANAGEDDIABETES CLINICTanPS,PtT TheNationalUniversityofMalaysia,KualaLumpur,MalaysiaOBJECTIVES: Theaimofthisresearchwastoassessthecost-effectivenessofpharmaceuticalcare(PC)programme(relativetocontrol)forpatientswithtype2DM. METHODS: Atotalof222patientswererecruitedatapharmacist-manageddiabetesclinicofagovernmenthospitalinMalaysiaandrandomlyallocatedtointerventionandcontrolgroup.Patientsintheinterventiongroup(n 111)wereprovidedwithPC,whereas,patientsinthecontrolgroup(n 111)receivedonlythe usualpharmacyservice.Clinicalandeconomicoutcomesofpatientswereevaluatedatbaselineandaftersixmonths. RESULTS: Therewasnosignicantdifferenceinthedemographicandclinicalcharacteristicsatbaselinebetweentheinterventionandthecontrolgroup.Signicantreductioninglycosylatedhaemoglobin (HbA1c)wasobservedfrombaselineto6-monthintheinterventiongroup.(Mean standarddeviations 9.93 1.76%versus8.83 1.85%, p 0.05).Althoughthe totalcostsperpatientweresignicantlyhigherfortheinterventiongroup (RM681.07versusRM542.64; p 0.014),thecosteffectivenessratiowaslowerinthe interventiongroup(RM619.15versusRM3617.60/1%reductioninHbA1c).Theincrementalcost-effectivenessratioforHbA1cwasRM145.72/%. CONCLUSIONS: In conclusion,incorporationofPCintothemanagementoftype2DMcanhavea denitive,positiveimpactonglycaemiccontrolandleadtomorecost-effective treatment. PDB35 UTILIZATIONOFPHYSICIAN,NURSINGANDDIETICIANSERVICESBYTYPE2 DIABETICPATIENTSATTENDINGPRIMARYCARECLINICSINSINGAPORETohMPHS,WuCX NationalHealthcareGroup,SingaporeOBJECTIVES: Physicians,NursingCareManagers(CMs)andDieticiansarepartof thehealthcareteamlookingafterpatientswithdiabetesmellitusatthe9public sectorprimarycareclinicsintheNationalHealthcareGroup(NHG)inSingapore. Thispaperstudiestheassociationbetweenthelevelofglycaemiccontrolandthe utilizationoftheseprofessionalservicesbytype2diabetesmellitus(T2DM)patientsattheseclinics. METHODS: ThisstudyincludedallpatientswithT2DMwho hadattendedthesameclinicsin2009foratleast12monthsforthetreatmentof diabetes.DatawasextractedfromtheNHGDiabetesRegistry(CDMS).Thenumber ofoutpatientclinicvisitstothePhysicians,CMsandDieticiansinayearwascomparedbyageanddegreeofglycaemiccontrolusingthemeanglycatedhaemoglobin(HbA1c)fortheyear,"Optimal"being 7%,"Acceptable"7-9%and"Poor"9 %. RESULTS: Therewere58,057T2DMpatientswithmorefemales(54%)anddisproportionatelymoreIndians(13%)andfewerChinese(71%)thanthegeneralpopulation.MeanHbA1cwas7.42 1.27%whichdecreasedgraduallywithagefrom8.16% ( 40years)to6.86%(90 years).TheannualPhysicianattendancesfordiabetes consultincreasedfrom4.1 1.5("Optimal"HbA1ccontrol)to5.2 2.2("Poor"control),representing85-89%ofallPhysicianattendancesfortheyear.Proportionof patientsseenbytheCMsincreasedfrom6.8%("Optimal"control)to52.5%("Poor" control),andsimilarforDieticians,anincreasefrom1.3%("Optimal"control)to 9.5%("Poor"control). CONCLUSIONS: T2DMpatientswith"Optimal"glycaemic controlhadsignicantlyfewerclinicvisitstothehealthcareteamthanthosewith "Poor"control,translatingtoloweroverallhealthcarecostsforthepatientsandthe healthcaresystem.Moreattentionisneededtoimprovethecareofthepatients with"Poor"glycaemiccontrol,inordertoachievebetterlong-termhealthoutcomesandreducehealthcareutilizationandcosts. PDB36 ANECONOMICMODELOFTHEEFFECTSOFQUICKRELEASEBROMOCRIPTINE MESYLATEVERSUSALTERNATIVEORALANTIDIABETICDRUGSON HOSPITALIZATIONSANDCOSTSMenzinJ1,DuczakowskiC1,FriedmanM1,NeumannP2,ScrantonR3 1BostonHealthEconomics,Inc.,Waltham,MA,USA,2TuftsUniversitySchoolofMedicine, Boston,MA,USA,3Veroscience,LLC,Tiverton,RI,USAOBJECTIVES: Quickrelease(QR)-bromocriptinemesylateisanewtreatmentfor type2diabetes(T2DM)thathasbeenshowntoreduceHbA1candcardiovascular (CV)-relatedhospitalizationsoveroneyear.Theobjectiveofthisstudywastoestimatetheeconomicimpactof(QR)-bromocriptineversusotheroralantidiabetic drugsasadd-ontherapyforpatientswhofailedinitialtreatment. METHODS: A decision-analysismodelwasdesignedtocompareoutcomesamongQR-bromocriptine,pioglitazone,rosiglitazone,andsitagliptinoveroneyearwhenusedas add-ontherapy.Foreachdrug,changesinHbA1clevelswerederivedfromproduct labelsandusedtoderivetheexpectednumberofhospitalizationsfordiabetesrelatedcomplicationsbasedonpublishedequationslinkingHbA1clevelstohospitalizations.RatesofCV-relatedevents(majoradversecardiacevents,congestive heartfailure,angina,andrevascularization)wereextractedfromclinicaltrialreportsofantidiabeticdrugsinT2DMandwereexpressedonanannualizedbasis. HospitalcostsperadmissionwereestimatedfromtheHealthcareCost&UtilizationProject,whiledrugcostswerebasedonwholesaleacquisitioncost.Apayer perspectivewasassumedanddirectmedicalcostswereexpressedin2009USD. RESULTS: PatientstreatedwithQR-bromocriptinehadlowerdiabetes-relatedhospitalizationcosts($2,017)thanthosereceivingrosiglitazone($2,038),andhigher costscomparedtopioglitazone($1,928)andsitagliptin($1,989).Patientsreceiving QR-bromocriptinehadlowerCV-relatedhospitalizationcosts($426)thanthose treatedpioglitazone,sitagliptin,orrosiglitazone($523,$708,and$729,respectively).AnnualdrugcostswerelowerforpatientsreceivingQR-bromocriptine ($2,122)comparedtopioglitazone($2,605)andsitagliptin($2,282)andhigherthan forthosereceivingrosiglitazone($1,977).Overallone-yearcostswereestimatedto be$4,565,$5,056,$4,979,and$4,745forQR-bromocriptine,pioglitazone,sitagliptin, androsiglitazonerespectively. CONCLUSIONS: Ourndingssuggestthat,overone year,T2DMpatientstreatedwithQR-bromocriptineasanadd-ontherapyareexpectedtohavelowercoststhanthosereceivingpioglitazone,sitagliptin,and rosiglitazone. Diabetes/EndocrineDisorders–Patient-ReportedOutcomes&Preference-Based Studies PDB37 MAIL-ORDERPHARMACYUSEANDMEDICATIONADHERENCEAMONG MEDICAREPARTDBENEFICIARIESWITHDIABETESZhangL,ZakharyanA,StocklKM,HaradaAS,CurtisBS,SolowBK PrescriptionSolutions,Irvine,CA,USAOBJECTIVES: ToexaminemedicationadherenceamongMedicarePartDbeneciarieswithdiabetesandexplorewhetherthereisanyassociationofusingmail-order pharmacy(vs.retailpharmacy)withbetteradherenceinthispatientpopulation. METHODS: Usingadministrativepharmacyclaimsdata,weconductedaretrospectivecohortstudyontheMedicarePartDbeneciarieswhonewlyinitiatedoral anti-diabetictreatmentbetweenJuly1,2008andDecember31,2008.Theprimary outcomeofinterestwasmedicationadherencetooralanti-diabeticsduringthe benetyearof2009,whichwasmeasuredusingtheproportionofdayscovered (PDC).Mail-orderpharmacyuserswerematchedtoretailpharmacyusersviapropensityscoring,controllingforpatient'sdemographicandclinicalcharacteristics. RESULTS: Atotalof22,546patientswithdiabeteswereidentied.TheaveragePDC was0.60andonly41.6%ofthestudypopulationwasadherent(denedasPDC 0.8) withoralanti-diabeticmedicationsduringcalendaryear2009.Thematchedsampleincluded1361patientsfromeachcohort.ComparedwiththeretailpharmacyA97VALUEINHEALTH14(2011)A1–A214

PAGE 106

group,mail-orderpharmacyusersdemonstratedasignicantlyhigherPDC(0.68 vs.0.61;P 0.001)throughoutthebenetyear.Morepatientsinthemail-order pharmacygroup(49.7%)wereadherentwiththeiroralantidiabeticmedications comparedto42.8%intheretailpharmacygroup. CONCLUSIONS: Adherencewith oralanti-diabeticmedicationsamongMedicarePartDbeneciariesissuboptimal. Patientsusingmail-orderpharmacywerelikelytohavebetteradherencethan thosewhousedretailpharmaciesfortheirmedicationrells.Thecausalrelationshipbetweenmail-orderpharmacyuseandadherence,however,shouldbefurther examinedinarandomizedstudysetting. PDB38 ACOMPARISONOFINSULINADHERENCEINPATIENTSWITHTYPE-2DIABETES INITIATINGTHERAPYWITHINSULINDETEMIRFLEXPENORNPHINSULININ AVIALBuysmanE1,ConnerC2,LiuF1,AagrenM3,BouchardJ4 1i3Innovus,EdenPrairie,MN,USA,2NovoNordisk,Inc.,Redmond,WA,USA,3NovoNordisk, Inc.,Princeton,NJ,USA,4NovoNordisk,Inc.,Plaistow,NH,USAOBJECTIVES: Non-adherencetoinsulintherapyinpatientswithtype2diabetes presentsaseriouschallenge.Potentialexplanationsfornonadherencemayincludeaversiontoinsulinself-injectionandfearofhypoglycemicevents.Inclinical trails,insulinanalogshaveshowntoreducetheriskofhypoglycemiceventsversus humaninsulins,andarecentreviewsuggeststhatinsulindeliveredviaapen devicemayresultingreateradherenceversusvialandsyringe.Thisstudywas conductedtocomparetheadherenceratesofpatientsinitiatingbasalinsulintherapywithinsulindetemir(IDet)FlexPen¨versusthoseinitiatingbasalinsulintherapywithNPHviavialandsyringe. METHODS: DataweregatheredfromalargeUS nationalpayerretrospectiveclaimsdatabase,andincludedonlypatientswithtype 2diabetesthatinitiatedbasalinsulintherapywitheitherIDetFlexPen¨orNPHin vials.Patientswithclaimsforanyothertypeofinsulin,otherthantheindexinsulin formulationsduringthe12-monthobservationperiodwereexcluded.Patients weredenedasbeingadherenttotherapyiftheyhadamedicationpossession ration(MPR)ofatleast0.80inthe12-monthfollowupperiod. RESULTS: TheIDet FlexPen¨cohort(n 1082)andtheNPHvialcohort(n 794)wereofsimilarage (54.06vs.53.13,p 0.134);however,theIDetFlexPen¨cohorthadalowerproportionoffemalepatients(44%vs.55%,p 0.001)andfewerpatientswithoutahistory ofpre-indexOADs(9%vs45%,p 0.001),thantheNPHvialcohort.Aftercontrolling forimportantconfounders,patientsinitiatinginsulintherapywithIDetFlexPen¨ were39%morelikelytoachieveanMPRof0.80orgreaterversuspatientsinitiating insulintherapywithNPHvial(95%CI:1.04-1.85). CONCLUSIONS: Theseresults suggestthatadherencemaybeimprovedforpatientsinitiatingbasalinsulintherapywithIDetintheFlexPen¨versusNPHinavial. PDB39 ADESCRIPTIVEANALYSISOFTREATMENTADHERENCEWITHGROWTH HORMONE:FINDINGSFROMANATIONALMANAGEDCAREPOPULATIONIN THEUNITEDSTATESFincherC1,KozmaC2,MeleticheDM3,VelezF3 1EMDSerono,Inc.,AnnArbor,MI,USA,2IndependentResearchConsultant/AdjunctProfessor, UniversityofSouthCarolina,St.HelenaIsland,SC,USA,3EMDSerono,Inc.,Rockland,MA,USAOBJECTIVES: Todescribeadherencewithgrowthhormone(GH)treatmentbyage groupamongnewlyinitiatedpatients. METHODS: Somatropinpharmacyclaims wereselectedfromJuly1999throughMay2008inPharMetrics,anationalmanaged caredataset.Continuouseligibilityfor6monthsbeforeand12monthsafterthe rstGHprescription(i.e.,indexdate)wasrequired.Patientswereexcludedifthey had:claimsforbrandsofsomatropinusedforHIV-associatedwasting;potential dataentryerrorsinkeyvariables;orhadaclaimforGHproductnolongeronthe market.Adherencewasassessedusingmedicationpossessionratios(MPRs)denedasthesumofdayssupplydividedby360daysofobservationandcappedat 100%.EvaluationsbyagegroupincludedaverageMPR,percentageofpatientswith atleast80%MPR,andpercentageofpatientswhostayontherapy(i.e.,persistence) foratleast80%oftheobservationperiod. RESULTS: Atotalof1355patientsreceivingGHmetstudycriteria;64.0%malesand36.0%females.ThepercentageofpatientsreceivingGHbyagecategorywasasfollows: 4(5.2%),4to 13(40.9%),13to 18(34.8%),18 (19.1%).AverageMPRwasgreatestinthe 4agegroup(87.2;SE 2.8),followedby4to 13(82.9;SE1.0),13to 18(80.5;SE1.1)and18 (68.1;SE1.5). Patientsinthe18 groupshadsignicantlylowermeanMPR(68.1%SD27.2)than allotheragegroups(allabove80%;p 0.0001).Thepercentageofpatientswithan MPRofatleast80%was71.4%,69.3%,66.1%and44.8%,respectively.Approximately 89%ofthe4to 13agegroupremainedontherapy(i.e.,persisted)formorethan 80%oftheobservationperiodversusonly64.5%ofthe18 group. CONCLUSIONS: Adherencewithgrowthhormonedecreasesasageincreases.Thereisopportunity, however,forimprovementinallagegroups. PDB40 EXAMININGADHERENCEWITHMEDICATIONSUSEDINTREATINGDIABETIC PERIPHERALNEUROPATHICPAINANDTHEIRASSOCIATIONWITHORAL ANTIDIABETICMEDICATIONADHERENCEOladapoAO,BarnerJC,RascatiKL,StrasselsS UniversityofTexasatAustin,Austin,TX,USAOBJECTIVES: Toexamineadherencetomedicationsusedinmanagingpainfuldiabeticperipheralneuropathy(PDPN)anddeterminetheirassociationwithoral antidiabeticmedication(OAD)adherence. METHODS: AretrospectivecohortanalysisusingtheTexasMedicaidprescriptionclaimsdatabase.Studyparticipants wereadult(30-64years)MedicaidbeneciariesprescribedOADandPDPNmedications.DatawereextractedfromJune1,2003toOctober31,2009.Thestudy'sresearchobjectiveswereto:1)provideadescriptionofPDPNandOADmedicationuse amongthestudysubjects;2)determineifPDPNmedicationadherencediffers amongindividualPDPNmedication(i.e.,TCAs,gabapentin,pregabalinandduloxetine);and3)determineifPDPNmedicationadherenceisrelatedtopost-indexOAD medicationadherencewhilecontrollingforcovariates.Adherencewasmeasured usingmedicationpossessionratio(MPR). RESULTS: Atotalof4277patientsmetthe study'sinclusioncriteria.TheoverallmeanMPR( SD)forPDPNmedicationswas 75.4%( 23.9%).MeanMPRdifferedsignicantlyamongindividualPDPNmedications(p 0.0001).MeanMPRwashighestforduloxetine(85.6% 18.2%)andwas lowestforpregabalin(69.4% 24.9%).TheoverallmeanMPR( SD)forOADmedicationsdecreasedsignicantly(p 0.0001)from73.0%( 24.3%)inthepre-index periodto64.5%( 25.6%)inthepost-indexperiod.Aftercontrollingforcovariates, non-adherers(i.e.,MPR 80%)toPDPNmedications,comparedtoadherers(i.e., MPR 80%),weresignicantlylesslikelytobeadherenttoOADmedications[Odds Ratio(OR) 0.626;95%CI 0.545-0.719]. CONCLUSIONS: Patients'adherenceto PDPNmedicationswasassociatedwithadherencetoOADmedications.Patients whowereadherenttoPDPNmedicationsweremoreadherenttoOADmedications. PDB41 TREATMENTGAPSAMONGGROWTHHORMONEUSERSINANATIONAL MANAGEDCAREPOPULATIONINTHEUNITEDSTATESFincherC1,KozmaC2,MeleticheDM3,VelezF3 1EMDSerono,Inc.,AnnArbor,MI,USA,2IndependentResearchConsultant/AdjunctProfessor, UniversityofSouthCarolina,St.HelenaIsland,SC,USA,3EMDSerono,Inc.,Rockland,MA,USAOBJECTIVES: TodeterminetreatmentgaptrendsamongrsttimeGHrecipients usingalarge,USmanagedcaredatabase.Toourknowledge,thisistherstanalysis ofgapsinGHtreatmentusingalargesampleofprescriptionclaims. METHODS: SomatropinpharmacyclaimswereselectedfromJuly1999throughMay2008in PharMetrics,anationalmanagedcaredataset.Continuouseligibilityfor6months beforeand12monthsaftertherstGHprescription(i.e.,indexdate)wasrequired. Patientswereexcludediftheyhad:claimsforbrandsofsomatropinapprovedfor humanimmunodeciencyvirus-associatedwastingandshortbowelsyndromein patientsreceivingspecializednutritionalsupport,potentialdataentryerrorsin dayssupply,quantitydispensedorallowedamounts,hadaclaimforGHproduct nolongeronthemarket,orincorrectdosesbasedondataentryerrors.Thenumber andpercentageofpatientswithtreatmentgapsofatleast60dayswerebasedon reporteddayssupplydispensingdates. RESULTS: Atotalof1355patientsreceiving GHmetstudycriteria.Thestudypopulationconsistedof64.0%malesand36.0% females.ThepercentageofpatientsreceivingGHbyagecategorywereasfollows: 4(5.2%),4to 13(40.9%),13to 18(34.8%),18 (19.1%).Gapsintherapyof60days werecalculatedandresultsshowedthat27.0%ofGHpatientsdiscontinuedtheir medicationforatleast60dayswithin1yearoffollow-up. CONCLUSIONS: Gapsin therapyoccuramongapproximately27%ofGHusersduringaone-yeartimeperiod,regardlessofageorgender.Reasonsforthistherapeuticgapremainunknown. PDB42 ADHERENCETOROUTINEORSTRUCTUREDGLUCOSEMONITORING:A QUALITATIVEEVALUATIONFROMTHESTEPSTUDYBergstromF1,WhittemoreR2,SchaeferC1,ChristopherCA3,AmstutzL4,WagnerRS4 1Covance,Gaithersburg,MD,USA,2YaleUniversity,NewHaven,CT,USA,3Covance,SanDiego, CA,USA,4RocheDiagnosticsOperations,Inc.,Indianapolis,IN,USABACKGROUND: EvidenceregardingthevalueandutilityofSMBGininsulin-na•ve type2diabetes(T2DM)hasbeenmixed.Failuretopersonallyunderstandthevalue associatedwithSMBGresultsand/orshareSMBGdatawithhealthcareproviders (HCPs)maycontributetopoortreatmentadherence.TheStructuredTestingProgram(STeP)study,aprospective,cluster-randomized,multi-centeredclinicaltrial, demonstratedthatstructuredSMBGsignicantlyimprovesglycemiccontroland qualityoflifeinnon-insulin-treatedT2DM.Inthestudy,483poorly-controlled (HbA1c 7.5%),insulin-na•veT2DMpatientswererandomizedtost