<%BANNER%>

Application of logistic regression for signal detection and risk assessment of macrolide-associated torsade de pointes.

www.ispor.org ( Publisher's URL )
University of Florida Institutional Repository
MISSING IMAGE

Material Information

Title:
Application of logistic regression for signal detection and risk assessment of macrolide-associated torsade de pointes.
Physical Description:
Conference Papers
Creator:
Ayad Ali
Conference:
ISPOR 16th Annual European Congress
Publisher:
ISPOR
Place of Publication:
Lawrenceville, NJ, USA
Publication Date:

Notes

Abstract:
OBJECTIVES: Torsade de pointes (TdP) is an identified risk for antibacterial treatment with macrolides, but the risk might vary across individual macrolides. This study evaluates the association between macrolides and TdP from the FDA Adverse Event Reporting System (FAERS). METHODS: Disproportionality analysis of FAERS between 1997 and 2012 was conducted for azithromycin, clarithromycin, erythromycin and Telithromycin with TdP preferred term of the Medical Dictionary for Regulatory Activities using logistic regression. Logistic reporting odds ratio (LROR) and corresponding 95% confidence intervals (CIs; LR05-LR95) are estimated for individual macrolides. Estimates are compared with Multi-Item Gamma Poisson Shrinker data mining algorithm. Estimates ≥ 2 yield significant safety signals. Lack of overlap between estimate-specific CIs indicates efficient approach for signal detection. RESULTS: A total of 318 TdP events were reported for macrolides. The majority of reports were for erythromycin (n= 122) and clarithromycin (n= 114), followed by azithromycin (n= 74) and telithromycin (n= 8). Most of events were reported for females with median age of 62.8 years. 98% of TdP were serious events (10% of those were deaths). Death contributed to 18%, 10% and 5% of serious TdP events for azithromycin, erythromycin and clarithromycin, respectively. Telithromycin didn’t have death TdP reports. Significant TdP signals were identified for macrolides. LROR (LR05-LR95) were: erythromycin 48.1 (40.4-55.9); clarithromycin 15.3 (12.1-19.1); azithromycin 13.9 (11.1-17.2); and telithromycin 6.79 (3.20-13.5). Except for telithromycin, no overlap was observed between estimate-specific CIs between analysis methods. CONCLUSIONS: Macrolides are associated with more than expected reporting of TdP compared to other drugs and events in the database; however, the risk was highest in erythromycin and lowest in telithromycin. Fatal TdP was highest in azithromycin, and telithromycin didn’t have fatal TdP. The findings show differential risk of TdP across macrolide antibiotics, emphasizing the need for monitoring cardiac rhythms in patients treated with macrolides.
Acquisition:
Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Ayad Ali.
Publication Status:
Published
General Note:
Suggested Citation: Ali AK. Application of logistic regression for signal detection and risk assessment of macrolide-associated torsade de pointes. Value in Health. November 2013; 16(7):A512 Abstract No. PCV1.

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:
IR00003527:00001

MISSING IMAGE

Material Information

Title:
Application of logistic regression for signal detection and risk assessment of macrolide-associated torsade de pointes.
Physical Description:
Conference Papers
Creator:
Ayad Ali
Conference:
ISPOR 16th Annual European Congress
Publisher:
ISPOR
Place of Publication:
Lawrenceville, NJ, USA
Publication Date:

Notes

Abstract:
OBJECTIVES: Torsade de pointes (TdP) is an identified risk for antibacterial treatment with macrolides, but the risk might vary across individual macrolides. This study evaluates the association between macrolides and TdP from the FDA Adverse Event Reporting System (FAERS). METHODS: Disproportionality analysis of FAERS between 1997 and 2012 was conducted for azithromycin, clarithromycin, erythromycin and Telithromycin with TdP preferred term of the Medical Dictionary for Regulatory Activities using logistic regression. Logistic reporting odds ratio (LROR) and corresponding 95% confidence intervals (CIs; LR05-LR95) are estimated for individual macrolides. Estimates are compared with Multi-Item Gamma Poisson Shrinker data mining algorithm. Estimates ≥ 2 yield significant safety signals. Lack of overlap between estimate-specific CIs indicates efficient approach for signal detection. RESULTS: A total of 318 TdP events were reported for macrolides. The majority of reports were for erythromycin (n= 122) and clarithromycin (n= 114), followed by azithromycin (n= 74) and telithromycin (n= 8). Most of events were reported for females with median age of 62.8 years. 98% of TdP were serious events (10% of those were deaths). Death contributed to 18%, 10% and 5% of serious TdP events for azithromycin, erythromycin and clarithromycin, respectively. Telithromycin didn’t have death TdP reports. Significant TdP signals were identified for macrolides. LROR (LR05-LR95) were: erythromycin 48.1 (40.4-55.9); clarithromycin 15.3 (12.1-19.1); azithromycin 13.9 (11.1-17.2); and telithromycin 6.79 (3.20-13.5). Except for telithromycin, no overlap was observed between estimate-specific CIs between analysis methods. CONCLUSIONS: Macrolides are associated with more than expected reporting of TdP compared to other drugs and events in the database; however, the risk was highest in erythromycin and lowest in telithromycin. Fatal TdP was highest in azithromycin, and telithromycin didn’t have fatal TdP. The findings show differential risk of TdP across macrolide antibiotics, emphasizing the need for monitoring cardiac rhythms in patients treated with macrolides.
Acquisition:
Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Ayad Ali.
Publication Status:
Published
General Note:
Suggested Citation: Ali AK. Application of logistic regression for signal detection and risk assessment of macrolide-associated torsade de pointes. Value in Health. November 2013; 16(7):A512 Abstract No. PCV1.

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:
IR00003527:00001


This item is only available as the following downloads:


Full Text

PAGE 1

A512 VALUE IN HEALTH 16 (2013) A323…A636 nel during standard-of-care visits. Although this is the traditional approach, it has several practical shortcomings: infrequent or irregular standard-of-care visits; patients switching health care providers; transient nature of some safety events; treatment at non-study sites, and the lack of care-seeking for the safety events. An alternative study design that overcomes several of these shortcomings includes direct-to-patient contact (DPC) and follow-up. OBJECTIVES: The purpose of this study is to describe patients short and long term management patterns including the detection and verification of events of interest by contacting patients event treating physician (ETP). METHODS: This is multinational multicenter observational prospective longitudinal cohort study involving 10,600 patients in Europe and South America. Patients were enrolled at hospital and followed up by telephone by a specialized DPC unit at 6 weeks then quarterly up to 2 years after hospital discharge. Data collected include clinical outcomes, safety events of interest, economic burden and quality of life. RESULTS: A total of 88,276 phone calls to patients were performed, achieving a reachable rate of 88.5% and an interview successful rate of 84.6%. 87.2% of the initial sample completed the 2-year follow-up period: 5.7% of the sample died, 6.0% voluntarily withdrew and 0.9% was withdrawn to participate to a clinical trial. Only 19 patients were never reached (0.2%). In total, 2144 events of interest were detected and a medical verification from ETPs was obtained in 72.3% of the case. CONCLUSIONS: This study demonstrates that DPC along with a proactive pharmacovigilance system achieved very high response rates from both patients and ETPs. It represents an effective alternative methodology to collect patients data, including safety data, in real life settings. PCV3 ANTITHROMBOTIC STRATEGIES AND ADVERSE OUTCOMES IN ACUTE CORONARY SYNDROME AND ATRIAL FIBRILLATION: A COMMUNITY STUDYChamberlain A M 1, Gersh B J 1, Klaskala W 2, Mills R M 2, Alonso A 3, Weston S A 1, Roger V L 1 1Mayo Clinic, Rochester, MN, USA, 2Janssen Research & Development, LLC, Raritan, NJ, USA, 3University of Minnesota School of Public Health, Minneapolis, MN, USAOBJECTIVES: Atrial fibrillation (AF) is frequent in acute coronary syndromes (ACS) and impacts the need for subsequent antithrombotic treatment. However, whether associations of antithrombotic treatment with strokes, bleeding, and death differ between ACS patients with and without AF remains uncertain. METHODS: Olmsted County, MN residents with incident myocardial infarction or unstable angina during 2005-2010 were classified according to AF status prior to or during the index ACS hospitalization. Antithrombotic strategy at discharge was categorized into double/ triple vs. no/single and warfarin vs. no warfarin therapy. Cox regression was used to examine associations between treatment and strokes, bleeding, and mortality and to test whether AF differentially impacted these associations. Models were ageand sex-adjusted and weighted for the inverse probability of the propensity score for therapy. RESULTS: A total of 244 of 1140 incident ACS patients had AF (ACS + AF), and of these patients, 49% were discharged on double/triple antithrombotic therapy and 33% received warfarin. In 896 ACS patients without AF, 63% were discharged on double/triple therapy and 4% on warfarin. Over a median of 4.2 years, 95 strokes, 297 bleeds, and 255 deaths occurred. Antithrombotic strategy was not associated with the risk of stroke or bleeding. Among all ACS patients, mortality was similar between strategies; however, ACS + AF patients on double/triple therapy (vs. no/single therapy) showed a trend toward a lower risk of death [HR (95% CI): 0.64 (0.41-1.02) for ACS + AF, 1.25 (0.83-1.87) for those without AF; pinteraction= 0.03]. For warfarin vs. no warfarin, those with AF exhibited a lower mortality risk with warfarin [HR (95% CI): 0.46 (0.29-0.73) for ACS + AF, 2.01 (0.94-4.32) for those without AF; pinteraction< 0.01]. CONCLUSIONS: Altogether, in community ACS patients, antithrombotic strategy is not associated with the risk of stroke, bleeding, or death. However, in the subset of ACS patients with AF, a warfarin strategy is associated with a lower risk of death. PCV4 COMORBIDITY INFLUENCE IN THE CONTROL OF ARTERIAL HYPERTENSIONSicras-Mainar A 1, Fernandez J L 2 1Badalona Serveis Assistencials, Badalona, Spain, 2Primary Care Center Doctor Vicen Papaceit, La Roca del Valles, SpainOBJECTIVES: Determine the impact of co-morbidity in controlling arterial hypertension in clinical practice situation. METHODS: Cross-sectional study, conducted from retrospective review of patient records obtained from primary care. We included patients 30 years who demanded attention during the year 2012. The main measures were: demographics, control ( 140/90 mmHg) blood pressure (BP), evolution time, co-morbidity (general and specific) protocol using cardiovascular events (CVE), visits and referrals to specialists. Statistical analysis and multiple linear regression for correction of the models, p < 0.05. RESULTS: We recruited 900 patients (mean age 67.4 years, women: 51.1%). The 64.8% (CI: 61.7% -67.9%) of subjects had a proper control of the BP. The average duration of hypertension was 6.0 years and use 4.6 protocol entries / year. The average number of visits was 15.3 and 0.7 referrals. The years of evolution were associated with age (r = 0.323). The control patients showed a higher average BP of ischemic heart disease (11.1% vs. 6.9%) and CVE (19.4% vs. 14.8%), p < 0.05. The BP control was associated with women (OR = 1.4) and the presence of CVE (OR = 1.8), p < 0.02. The number of diagnoses (morbidity) was associated with age ( = 0.14) and the number of referrals ( = 0.08), p < 0.05 (coefficient R2= 40%). CONCLUSIONS: Optimal control of BP is high. The profile of these patients is women in a situation of secondary prevention. PCV5 HIGH DOSAGES OF STATINS IN HIGH CARDIOVASCULAR RISK PATIENTS IN A REAL WORLD SETTING: TREATMENT PATTERNS AND OUTCOMESDegli Esposti L Sangiorgi D Buda S CliCon Srl, Ravenna, ItalyOBJECTIVES: International guidelines recommend the use of high dosages statins (in particular, Rosuvastatin and Atorvastatin) in high cardiovascular (CV) risk patients. The aim of this study was to assess the degree of implementation of the while after first year of treatment, controls tend to become less frequent, mostly every-three-months (39.9%). When recurrence occurs, main retreatment criteria were: presence of macular fluid (93.5%), loss of ETDRS letters (50.5%; 68.8% with 5 letters), macular hemorrhage (46.3%) and increase of central-retinal-thickness (45.4%). Single injection of anti-VEGF retreatments was applied in 48.6% of recurrences, while 23.6% were not treated with anti-VEGF. CONCLUSIONS: A relatively high adherence to SERV guidelines was followed by Spanish retinologists regarding diagnosis, treatment and control of AMD. However, limitations have been identified on patterns of follow-up and retreatment criteria that may compromise therapeutic and clinical outcomes. PSS58 TREATMENT PATTERNS OF RANIBIZUMAB AND AFLIBERCEPT FOR THE MANAGEMENT OF WET AGE-RELATED MACULAR DEGENERATION IN ROUTINE CLINICAL PRACTICE IN THE UNITED STATESFerreira A 1, Sagkriotis A 1, Milnes F 1, Olson M 1, Hill J 2, Lu J 2, Makin C 2 1Novartis Pharma AG, Basel, Switzerland, 2IMS Health, Plymouth Meeting, PA, USAOBJECTIVES: To compare the frequency of injections for ranibizumab and aflibercept for the treatment of wet age-related macular degeneration (wAMD) in a realworld setting. METHODS: Claims from the IMS Integrated Data Warehouse were used to identify treatment-nave patients with wAMD receiving 1 prescription for ranibizumab or aflibercept between 1 November 2011 and 31 March 2012 and with a follow-up of 365 days. Patients were considered to be treatment-nave if they had received no anti-VEGF treatment for > 180 days prior to the index date. Mean number of injections received within 365 days of treatment initiation and the proportion of patients receiving 3 loading doses within 3 months were calculated and compared for patients starting therapy with ranibizumab or aflibercept. No adjustments for bilateral treatment were performed. RESULTS: Patient characteristics were similar for patients receiving ranibizumab (N = 2,799) or aflibercept (N = 117) at the index date (median age: 82 years; % male, 36%). The mean (SD) number of anti-VEGF injections received by patients starting on ranibizumab or starting on aflibercept was similar: 5.63.5 and 5.42.8, respectively (p = 0.6593). In patients continuously treated with the same agent for 12 months, the mean number of injections was also similar across ranibizumab and aflibercept: 4.93.4 (n = 1,898) vs 5.22.6 (n = 87), respectively (p = 0.2705). Approximately two-thirds of patients with more than 3 injections received 3 loading doses within the first 3 months (ranibizumab, 65.6%; aflibercept, 64.3%). Results for patients who were treatment-nave for 365 days were consistent with the above results. CONCLUSIONS: There are limited data regarding real-world treatment patterns with aflibercept for the management of wAMD. Our results suggest that in routine clinical practice patients receive a comparable number of injections in the first year of treatment with both ranibizumab and aflibercept. Analysis of data from larger patient populations and for longer followup is warranted when these data become available. RESEARCH POSTER PRESENTATIONS … SESSION IV DISEASE-SPECIFIC STUDIES CARDIOVASCULAR DISORDERS … Clinical Outcomes Studies PCV1 APPLICATION OF LOGISTIC REGRESSION FOR SIGNAL DETECTION AND RISK ASSESSMENT OF MACROLIDE-ASSOCIATED TORSADE DE POINTESAli A K Eli Lilly and Company, Indianapolis, IN, USAOBJECTIVES: Torsade de pointes (TdP) is an identified risk for antibacterial treatment with macrolides, but the risk might vary across individual macrolides. This study evaluates the association between macrolides and TdP from the FDA Adverse Event Reporting System (FAERS). METHODS: Disproportionality analysis of FAERS between 1997 and 2012 was conducted for azithromycin, clarithromycin, erythromycin and Telithromycin with TdP preferred term of the Medical Dictionary for Regulatory Activities using logistic regression. Logistic reporting odds ratio (LROR) and corresponding 95% confidence intervals (CIs; LR05-LR95) are estimated for individual macrolides. Estimates are compared with Multi-Item Gamma Poisson Shrinker data mining algorithm. Estimates 2 yield significant safety signals. Lack of overlap between estimate-specific CIs indicates efficient approach for signal detection. RESULTS: A total of 318 TdP events were reported for macrolides. The majority of reports were for erythromycin (n = 122) and clarithromycin (n = 114), followed by azithromycin (n = 74) and telithromycin (n = 8). Most of events were reported for females with median age of 62.8 years. 98% of TdP were serious events (10% of those were deaths). Death contributed to 18%, 10% and 5% of serious TdP events for azithromycin, erythromycin and clarithromycin, respectively. Telithromycin didnt have death TdP reports. Significant TdP signals were identified for macrolides. LROR (LR05-LR95) were: erythromycin 48.1 (40.4-55.9); clarithromycin 15.3 (12.1-19.1); azithromycin 13.9 (11.1-17.2); and telithromycin 6.79 (3.20-13.5). Except for telithromycin, no overlap was observed between estimate-specific CIs between analysis methods. CONCLUSIONS: Macrolides are associated with more than expected reporting of TdP compared to other drugs and events in the database; however, the risk was highest in erythromycin and lowest in telithromycin. Fatal TdP was highest in azithromycin, and telithromycin didnt have fatal TdP. The findings show differential risk of TdP across macrolide antibiotics, emphasizing the need for monitoring cardiac rhythms in patients treated with macrolides. PCV2 DIRECT-TO-PATIENT CONTACT AND PROACTIVE PHARMACOVIGILANCE SYSTEM: A CASE STUDYWiederkehr S Seux M Michel J Fournie X Mapi, Lyon, FrancePost-authorization non-interventional safety studies typically enroll patients at sites and safety events are reported and collected as appropriate by site person-