<%BANNER%>

Understanding situations where people report cost-related medication nonadherence (CRN)

Permanent Link: http://ufdc.ufl.edu/UFE0041494/00001

Material Information

Title: Understanding situations where people report cost-related medication nonadherence (CRN) a qualitative study
Physical Description: 1 online resource (135 p.)
Language: english
Creator: Allen, Mark
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: access, adherence, cost, medication
Pharmaceutical Outcomes and Policy -- Dissertations, Academic -- UF
Genre: Pharmaceutical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Around 20% of the US population reports not acquiring a prescription because of cost, also referred to as cost-related medication nonadherence (CRN) leading to obvious adverse health effects. However, given how those with relatively high incomes report CRN, and given research evidence that people often have reasons for being nonadherent even with acquired medications, and given that many other factors affect health behavior, it seems likely that other factors may be playing a substantial role in medication-access decisions for those who report CRN. The main research objective was to gain better understanding of what people are experiencing in situations in which they would report CRN. A grounded theory approach was used to better understand what happens when people report CRN. The data from which the framework was based was gathered by interviewing 21 clients at a clinic serving the under-insured who reported CRN. Data coding, analysis, and framework development were performed by the primary investigator and a second independent, with the resulting framework representing the primary investigator s compiling of the two separate ones. A framework that describes what happens when people report CRN was developed. People were able to initially acquire their medication only if they were offered help in acquiring it, were willing to access known sources of help, or if they were willing to pay the given cost. For those not able to initially acquire their medication, whether or not they sought help depended on their belief in what help was available, and in a re-evaluation of how much they needed the medication. The vast majority of those who sought help eventually received help and were able to acquire their medication. Those who did not seek out help either lived without or substituted with an alternative. Reporting of CRN is not really a one-time event, but a dynamic process in which some people are able to eventually acquire their desired medication, with the key factor being whether or not they receive help in doing do. Factors previously shown to affect drug utilization (such as worries about side effects, addiction, regimen complexity, etc.) had little to no effect on whether or not the drug was acquired. This current research improves our understanding of what happens when people report CRN, and will help in the design of programs to improve access to needed medication.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mark Allen.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Kimberlin, Carole L.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041494:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041494/00001

Material Information

Title: Understanding situations where people report cost-related medication nonadherence (CRN) a qualitative study
Physical Description: 1 online resource (135 p.)
Language: english
Creator: Allen, Mark
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: access, adherence, cost, medication
Pharmaceutical Outcomes and Policy -- Dissertations, Academic -- UF
Genre: Pharmaceutical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Around 20% of the US population reports not acquiring a prescription because of cost, also referred to as cost-related medication nonadherence (CRN) leading to obvious adverse health effects. However, given how those with relatively high incomes report CRN, and given research evidence that people often have reasons for being nonadherent even with acquired medications, and given that many other factors affect health behavior, it seems likely that other factors may be playing a substantial role in medication-access decisions for those who report CRN. The main research objective was to gain better understanding of what people are experiencing in situations in which they would report CRN. A grounded theory approach was used to better understand what happens when people report CRN. The data from which the framework was based was gathered by interviewing 21 clients at a clinic serving the under-insured who reported CRN. Data coding, analysis, and framework development were performed by the primary investigator and a second independent, with the resulting framework representing the primary investigator s compiling of the two separate ones. A framework that describes what happens when people report CRN was developed. People were able to initially acquire their medication only if they were offered help in acquiring it, were willing to access known sources of help, or if they were willing to pay the given cost. For those not able to initially acquire their medication, whether or not they sought help depended on their belief in what help was available, and in a re-evaluation of how much they needed the medication. The vast majority of those who sought help eventually received help and were able to acquire their medication. Those who did not seek out help either lived without or substituted with an alternative. Reporting of CRN is not really a one-time event, but a dynamic process in which some people are able to eventually acquire their desired medication, with the key factor being whether or not they receive help in doing do. Factors previously shown to affect drug utilization (such as worries about side effects, addiction, regimen complexity, etc.) had little to no effect on whether or not the drug was acquired. This current research improves our understanding of what happens when people report CRN, and will help in the design of programs to improve access to needed medication.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Mark Allen.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Kimberlin, Carole L.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041494:00001


This item has the following downloads:


Full Text

PAGE 1

UNDERSTANDING SITUATIONS WHERE PEOPLE REPORT COST-RELATED MEDICATION NONADHERENCE (CRN): A QUALITATIVE STUDY By MARK A. ALLEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010 1

PAGE 2

2010 Mark A. Allen 2

PAGE 3

3 To Michelle, Remy, and Quincy

PAGE 4

ACKNOWLEDGMENTS I would like to acknowledge my wife for supporting me on many levels throughout this whole process (and our entire tenure together), and to my two daughters who have amplified my life since they have appeared. I would also like to thank my multiple parents, grandparents, and siblings who have fo stered an environment in which learning and scholarly activities are encouraged and valued. Additionally, I appreciate the extensive ti me and effort put forth by my advisor and chairperson, Carole Kimberlin, and my dissertation committee members Richard Segal, Earlene Lipowski, and Allyson Hall. Without their guidance and constructive criticism, this research project w ould not be what it is. I would also like to thank several people at Community Clinic, In c. (CCI) for letting me use one of their clinics to recruit subjec ts for this study: R onald Greger, CCI medical director; Margaret Chang, medical director Silver Spring clinic; and Cenia Galvez, manager Silver Spring clinic. Research of any kind would come to a standstill without organizations such as CCI helping researc hers collect data and providing logistical support. Finally, I would like to thank the American Foundation of Pharmaceutical Education (AFPE) which funded this study. Their funding was obviously essential for the completion of this project. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDG MENTS .................................................................................................. 4LIST OF TABLES ............................................................................................................ 8LIST OF FI GURES .......................................................................................................... 9ABSTRACT ................................................................................................................... 10 CHAPTER 1 INTRODUC TION .................................................................................................... 12Problem Stat ement ................................................................................................. 15Theoretical Perspective .......................................................................................... 17Research Pu rpose .................................................................................................. 20Research Q uestions ............................................................................................... 22Significance fo r Pharma cy ...................................................................................... 222 LITERATURE REVIEW .......................................................................................... 25Current State of CRN in the United States .............................................................. 25Prevalence Es timates ....................................................................................... 25Consequences of CRN ..................................................................................... 26Identification and Characterizati on of Those Reporting CRN .................................. 28Concepts that have been Related to CRN .............................................................. 30Piette et al. 2006 .............................................................................................. 30Andersen & Adays Behavioral Model of Health Services Use, Phase 5 (Andersen & Davi dson, 2007) ....................................................................... 31Other Concepts from the Heal th Behavior Li terature ........................................ 33Health Belief Model (Rosenstock 1966; Janz & Be cker, 1984) ................. 34Social Cognitive Theo ry (Bandura, 1986) .................................................. 35Summary .......................................................................................................... 35Using Grounded Theory in Dr ug Use Res earch ..................................................... 36Summary ................................................................................................................ 403 METHOD S .............................................................................................................. 47Sampling ................................................................................................................. 48Data Collection / In terview Gu ide ............................................................................ 52Data Codi ng ............................................................................................................ 55Data Anal ysis .......................................................................................................... 57Theory Construction ............................................................................................... 58Scientific Rigor ........................................................................................................ 59 5

PAGE 6

4 RESULT S ............................................................................................................... 61Sample .................................................................................................................... 61General Description .......................................................................................... 61Demographi cs .................................................................................................. 61Sample Size as Determined by Sa turation of Majo r Concept s ......................... 63Acquiring Medi cation............................................................................................... 63Overview of the New Framew ork ............................................................................ 64Attempt to Acquire Desi red Prescrip tion ................................................................. 65Help Fact ors ..................................................................................................... 66Other persons provision of help to passive patient ................................. 66Patient willingness to access known sources of hel p ................................. 67Monetary Fa ctors ............................................................................................. 68Out-of-pocket pre scription co st .................................................................. 68Willingness to pay ...................................................................................... 73Factors not Shown to Impact Prescription Ac quisition ...................................... 75General beliefs about medication ............................................................... 76Other utilization-re lated fact ors .................................................................. 78Response to Not Acquiri ng Prescrip tion ................................................................. 79Knowledge / Belief that Help is Av ailable ......................................................... 79Perceived Level of Need for t he Prescribed M edication ................................... 80Perceived consequences of not taking the pre scription ............................. 81The time period for which they desired the pre scription ............................. 83The perceived availability of alternat ives ................................................... 85Case studies demonstrat ing perceiv ed need ............................................. 88People Seeking Help or Additional Re sources ....................................................... 90Money from Fri ends/Family .............................................................................. 91Money from Church .......................................................................................... 91Location of W here to Go................................................................................... 92Working More to Ge t Extra Ca sh ...................................................................... 92Looking for a Program / General Assi stance .................................................... 93Adequate Provisio n of He lp .............................................................................. 93Summary .......................................................................................................... 94Access Resu lt ......................................................................................................... 94Summary ................................................................................................................ 955 DISCUSSI ON ......................................................................................................... 98Sample .................................................................................................................... 98Dynamic Nature of CRN ......................................................................................... 99Comparison with Other Framewor ks .................................................................... 101Help Fact ors ................................................................................................... 101Monetary Fact ors ........................................................................................... 102Perceived N eed .............................................................................................. 103Knowledge / Belief that He lp is Ava ilable ....................................................... 107Seeking Help or Addi tional Resour ces ........................................................... 107Factors Not Having an Effect in the New Framework ..................................... 108 6

PAGE 7

Summary ........................................................................................................ 110Implicatio ns ........................................................................................................... 110Practice Implic ations ...................................................................................... 110Physicians ................................................................................................ 112Pharmacist s ............................................................................................. 114Policy Implicat ions .......................................................................................... 117Research Implications .................................................................................... 118Insurance does not necessa rily equal a ccess .......................................... 118Access related factors and other ut ilization related factors ...................... 119CRN is not a one-time, di chotomous event .............................................. 120Limitations ............................................................................................................. 122Summary of Conclusions and Implicatio ns ........................................................... 123Conclusion s .................................................................................................... 123Dynamic nature of CRN ........................................................................... 123Getting help (or additional resource s) was the difference between getting and not getting t he prescripti on ................................................. 123Access not affected by other ut ilization-related factors ............................ 124Perceived need was based on three fact ors ............................................ 124Implicatio ns .................................................................................................... 124Future research should distinguis h between access and other utilization related fact ors ....................................................................................... 124Insurance does not necessarily mean access .......................................... 125Short-term solutions ................................................................................. 125Physicians can offer help, but should not be relied upon to do so ........... 125Pharmacists are in a unique position and have a unique skill set to help people acquire their prescriptions ......................................................... 126Summary .............................................................................................................. 126APPENDIX INTERV IEW GUID E .............................................................................. 127LIST OF REFE RENCES ............................................................................................. 129BIOGRAPHICAL SK ETCH .......................................................................................... 135 7

PAGE 8

LIST OF TABLES Table page 2-1 Estimates of CRN by di fferent res earcher s ........................................................ 412-2 Factors associated with great er likelihood of reporting CRN .............................. 422-3 Questions to determine cost-rela ted medication underuse and resulting medication affordability used by different researchers ........................................ 432-4 Characteristics of studi es using grounded theory ............................................... 453-1 Characteristi cs of populat ion .............................................................................. 60 8

PAGE 9

LIST OF FIGURES Figure page 2-1 Conceptual framework of factors affecting cost-related medication underuse (Piette et al., 2006a) ........................................................................................... 462-2 Andersen and Adays B ehavioral Model of Health Systems Use, Phase 5 (Andersen & Davi dson, 2007)............................................................................. 464-1 Framework showing factors af fecting medicati on acquisi tion ............................. 97 9

PAGE 10

Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for t he Degree of Doctor of Philosophy UNDERSTANDING SITUATIONS WHERE PEOPLE REPORT COST-RELATED MEDICATION NONADHERENCE (CRN): A QUALITATIVE STUDY By Mark A. Allen May 2010 Chair: Carole L. Kimberlin Major: Pharmaceutical Sciences Around 20% of the US population reports not acquiring a prescription because of cost, also referred to as cost-relat ed medication nonadherence (CRN) leading to obvious adverse health effects. However, gi ven how those with relatively high incomes report CRN, and given research evidence that people often have reasons for being nonadherent even with acquired medications, and given that many other factors affect health behavior, it seems likely that other factors may be pl aying a substantial role in medication-access decisions for those who report CRN. The main research objective was to gai n better understanding of what people are experiencing in situations in which they would report CRN. A grounded theory approach was used to better understand what happens when people report CRN. The data from which the framework was based was gathered by interviewing 21 clients at a clinic serv ing the under-insured who reported CRN. Data coding, analysis, and framework develop ment were performed by the primary investigator and a second independent, with t he resulting framework representing the primary investigators comp iling of the two separate ones. 10

PAGE 11

11 A framework that describes what happens when people report CRN was developed. People were able to initially ac quire their medication only if they were offered help in acquiring it, were willing to acce ss known sources of help, or if they were willing to pay the given cost. For those not able to initia lly acquire their medication, whether or not they sought help depended on their belief in what help was available, and in a re-evaluation of how much they needed the medication. The vast majority of those who sought help eventually received help and were able to acquire their medication. Those who did not seek out help either lived without or substituted with an alternative. Reporting of CRN is not really a one-time event, but a dynamic process in which some people are able to eventually acquire t heir desired medication, with the key factor being whether or not they rece ive help in doing do. Factors previously shown to affect drug utilization (such as worries about side effe cts, addiction, regimen complexity, etc.) had little to no effect on whether or not the drug was acquired. Th is current research improves our understanding of what happens w hen people report CRN, and will help in the design of programs to improve access to needed medication.

PAGE 12

CHAPTER 1 INTRODUCTION This study focuses on learning more about the situations in which people report that they are not purchasing medication becau se of cost, a form of cost-related nonadherence (CRN). Patients reporti ng that they had not initially purchased a prescription because of cost were interviewed. Interviews were analyzed using a grounded theory approach to determine what factors were mo st important in their decision-making process. Having access to medication is an obvious first step in people taking said medication, so knowledge of these factors will be helpful in designing future programs to improve access to medicati on for this population. Approximately 20% of people r eport that they take less prescription medication than recommended because of cost (Piette, Heis ler, & Wagner, 2004a; Piette, Heisler, & Wagner, 2004b; Heisler, Wagner, & Piette, 2004a; Piette & Heisler, 2004c; Safran et al., 2002; Steinman, Sands, & Covinsky, 2001; Piette, Heisler, & Wagner, 2006b; Kenndey & Erb 2002; Reed & Hargraves 2003). Underuse of medication because of cost can lead to poorer health outcomes (espec ially in the case of chronic diseases) (Heisler et al., 2004b; Tamblyn et al ., 2001; Soumerai, Ross-Degnan, Avorn, McLaughlin, & Choordnovskiy, 1991; Kennedy & Er b, 2002), Cost concerns may also be forcing people to develop coping strategi es such as forgoing paying for other necessities (such as food or heat) in order to pay for medications (Piette et al., 2004b; Heisler et al., 2004a; Safran et al., 2002). Given that health care expenditures in general and costs for prescription drugs are both increasing, and that there is not a mandated safety net for those facing cost pressures, the number of people who are cutting back on medications because of cost is not likely to decrease in the near future. 12

PAGE 13

An easy assumption that the main reas on people report not purchasing medication because of cost is that medica tion is too expensive for them to afford. It stands to reason that if medications were made c heaper, then more people would be able to pay for them and in turn, more people actually wo uld pay for their prescribed medications. Is reducing medication cost all that needs to be done in order to minimize CRN? While the answer may seem obvious, the data suggest that other factors in addition to cost are being considered for those who report CRN. First of all, there are those with relatively high incomes who are reporting CRN, even though they could probably afford to pur chase most prescribed medications. In various studies examining a nationally repr esentative sample of people over 50, CRN in the previous year is reported in those with annual incomes >$60,000 in 13% of adults with diabetes treated with hypoglycemic medication (Piette et al., 2004b), in 16% of adults with chronic illnesses (Piette, Heisler, & Wagner, 2004d), and in 15% of adults using both preventive and symptom-relief medica tions (Piette et al ., 2006b). While these populations are undoubtedly subject to other cost pressures, it is reasonable to assume that some of these indivi duals would be able to pay for their medication, and are choosing to not do so. Theref ore, there may be other factor s that are being taken into consideration. Additional evidence that people take into account more than just cost when deciding whether to purchase medication comes from the liter ature examining why people do not take their medicines, even afte r they have been acquired. In their synthesis of qualitative studies of medicinetaking, Pound et al. (2005) compiled a list 13

PAGE 14

based on their review of various reasons t hat have nothing to do with cost for why people do not take their medicine. These reasons include (and are not limited to): They weigh the benefits vs. t he adverse effects of the drug Taking the medicine does not fit in their daily routine They are nave scientists they try it or stop it, and see what works best Their condition is improving They cant tell the difference They substitute with something else They dont like medicine in general A negative self-identity co mes from taking the drug They are worried about addiction Similarly, various models of health behavio r, such as the Health Belief Model (Rosenstock, 1966; Janz & Becker, 1984), the Theory of Reasoned Action / Planned Behavior (Fishbein & Ajzen, 1975), and Soci al Cognitive Theory (Bandura, 1986), identify different constructs that have been shown to affect whether or not someone engages in a health behavior. While t hese theories have been used to explain utilization of health care services rather t han access, the factors th at affect utilization could also be factors affecting whether or not a person reports CRN. Someone might believe that the outcome of not taking a drug may not be very severe (perceived severity from the Health Belief Model (Rosenstock, 1966; Janz & Becker, 1984), so they may decide in such a situation to not purchase a prescribed drug to save the outof-pocket cost given the fact that they do not perceive t he consequences of not taking the medication to be severe. Therefore, for the reasons given above, it seems as though people may report CRN if asked, even though their choice to not purchase medication consists of a number of factors in addition to cost. Fo r example, suppose someone chooses to stop taking medication because they believe thei r condition is improving, and they do not 14

PAGE 15

need it anymore. They have a prescription for a refill that calls for a $5 copay, and they decide to not get it refilled. This pers on then reports CRN; but how much difference does the cost make? Will reducing the cost to $2 change what this person decides to do? The point of this is not simply to sa y that consumers of products take into account more factors than just the price: everybody would probably agree to that. The major point is that the peopl e are reporting that they ar e either not purchasing a prescription or underusing a current prescription explicitly because of cost. In other words, cost is the major issue, if not t he only issue involved in their nonadherence with their prescribed medication. Given how those with relatively high incomes report CRN, people often have reasons for being nonadher ent even with acquired medications, and that many other factors affect health behavior, it seems likely that other factors may be playing a substantial role in medicationaccess decisions for those who report CRN. Problem Statement The drug utilization process includes at least two steps; first one must be able to obtain the drug (access), then one must take t he drug (utilization). These two steps are obviously linked, as access to the drug is a necessary and sufficient condition for being able to utilize the drug. Ult imately, however, utilization of the drug is what really matters. When medication utilization is not achieved (i.e. when people are not adherent in taking the drug), how does one know what the problem is? C ould it be that the person did not obtain the drug (an access problem), and thus could not take it? Or could it be that the person was able to obtain the drug, but did not take it for some other reason (a utilization problem)? At this poi nt, it does not seem that researchers are differentiating between factors that affect access (such as income, insurance status, 15

PAGE 16

etc.) and factors that affe ct utilization once the drug has been obtained (such as side effects, dont like medicine in general, etc.) when trying to figure out the factors that affect medication utilization. The studi ed problem of CRN is undoubtedly an access issue, in that people are reporting that they are not taking their medication because the cost is prohibiting them from obtaining what they really need. Currently, there are at least two frameworks that attempt to relate all of the different factors (including cost-related ones) that affect whether or not one utilizes a medication : Piette, Heisle r, Horne, & Alexander (2006a), and Andersen & Adays Behavioral Model of Health Services Use, Phase 5 (Andersen & Davidson, 2007). Both of these will be examined in detail in Chapt er 2. The Piette et al. framework summarizes all of the factors that have been associated with repor ting of CRN. The Behavioral Model of Health Services Use cata logues these factors into larger constructs and hypothesizes about the relationships am ong these constructs and health behavior. Although these frameworks summarize many of the factors that affect CRN, they have two short-comings. The first is that as mentioned above when discussing reasons for not taking medication from the adherence/compliance literat ure, there could be many health beliefs or behaviors that have littl e to do with accessing medication, yet could affect whether someone would report CRN. T hese are identified only generally in the two frameworks as beliefs, health practice s, or sociocultural influences. For example, if someone does not like to take medication in general and does not purchase a prescription for $10, they may report that they did not purchase the medication because of cost (an access issue), when in realit y, it is mostly their health beliefs that caused them not to purchase t he medication (a utilization iss ue). Based on the current 16

PAGE 17

frameworks, we have no idea which of these health beliefs or behaviors are affecting whether or not someone is not purchasing medication report edly because of cost, nor their relative importance. The other short-coming is that it would be difficult to det ermine which of the factors really come into play when a person is conf ronted with a situation in which they cannot afford their medication. For example, while the frameworks i ndicate that those of lower income will have a higher likelihood of reporting CRN compared to those with higher incomes, it cannot explain why some with higher incomes do not purchase medication, while those with lower incomes still do purchas e medication. So w hat factors are the most important ones that people take into consideration when deciding whether or not to purchase a medication? What specific problems do these individuals have in purchasing medication, such that policy an d resources could be used to help resolve them? Since CRN seems to affect all ty pes of people in the greater population (although some more than others), one needs to know the exact nature of the problems that are causing peopl e to report CRN. Theoretical Perspective Symbolic interactionism, a sociological per spective used to study human behavior, is the foundational basis for the method of grounded theory. To completely understand how grounded theory works, one must firs t understand symbolic interactionism. The term symbolic interactionism was created by the sociologist Herbert Blumer, and was intended to be used as a wa y of figuring out how to behave, or in essence live, in a particular society. One must figure this out, but how does one approach it? This perspective is based on three general principles (Blumer, 1969): 17

PAGE 18

1. "Human beings act toward things on t he basis of the meanings they ascribe to those things." 2. "The meaning of such things is derived from, or arises out of, the social interaction that one has with others and the society." 3. "These meanings are handled in, and modifi ed through, an interpretative process used by the person in dealing wit h the things he/she encounters. In essence, this describes the process in which people figure out what to do, or how to act given a certain situation. Befo re one knows what to do, one must first figure out the norms and rules for behavior given the society in which one lives. The only way to do this is by interacting with other peopl e or other situations, and interpreting the meanings that things have through continual iterations. As an example, imagine being dropped off in a place where one does not know anything about the culture, the language, or the social norms (for instanc e, someone from middleclass America being sent to an isolated rural village in a third wo rld country). How would one figure out how to behave, or what actions to take? Following the symbolic interactionist perspective, one would begin with relatively straightforwar d interactions with people, all the while interpreting and revising the meanings of ev erything that is happening to you. After enough revising, one is able to give meaning to things, and one can then act appropriately for that culture. The use of the symbolic interactionist perspec tive is therefore useful in situations where one is trying to learn more about a gr oup or a social process with which one is not familiar. While the relevant people ma y not be consciously thinking about the meaning they give to things brought about by their process of continual interaction and revision, by probing these people in depth, one can hope to learn as much as possible about the unknown group or social process. 18

PAGE 19

In social research, one is trying to dete rmine why people do things, or why people act the way that they do. T he use of the symbolic interactionist perspective to answer these questions has certain benefits; by eventually determining the meanings that people give to things, one can have a better (but not perfect) idea of how people act. Instead of just relying on past observations (every time X has happened, people did Y), one can learn about the meanings that people gave to X to figure out why they did Y, and hopefully apply this knowledge to sim ilar situations in which one is trying to determine human behavior. A drawback to this method is that it would take a large amount of interactions to determine how a society is structured which, from a sociological perspective, governs a societ ys norms and cultures, and thus how people behave. Nevertheless, the use of the symbolic interactionist perspective can be used to determine how people view the societys norms and cultures, regardless of the underlying social structure. Grounded theory is in essence the embodim ent of the use of the symbolic interactionist perspective to answer social research questions, where one is trying to figure out why people act as t hey do. Data are collected from a group of people that one is trying to learn more about, specifically how and why the individuals in this group behaved the way that they did. The dat a are coded and analyzed to determine the meanings that these individuals gave to people and things when responding to a certain situation. More and purposeful data ar e gathered, coded, analyzed, and constantly compared with previous data until one has figured out the rules that governed behavior for this group and for a particu lar situation. A theory th en emerges, grounded from the data collected, to explain what caused people to behave in a certain way. 19

PAGE 20

In this sense, the grounded theory method can be used anytime one is trying to determine how and why people behaved the way they did in a certain situation. In the case of this research, it is an appropriate way to try to figure out how and why people responded when faced with a situation in which they report that they did not buy a prescription because of cost. Using the grounded theory approach, one can collect data from these individuals and, by figuring out the meanings that they gave to certain things (through continuous data collection, codi ng, and analyzing), learn enough about them such that a theory can be devised to explain their behavior. Research Purpose The problem of cost-related medica tion nonadherence (CRN) has been shown to affect a significant percentage of the population (approximately 20%), and can lead to dire health and economic consequences. CRN can lead to poorer health outcomes (especially in the case of chronic dis eases), and may force people to develop coping strategies such as forgoing paying for other necessities (such as food or heat) in order to pay for medications. Gi ven that health care expenditures in general and costs for prescription drugs are both increasing, and t hat there is not a mandated safety net for those facing cost pressures, the number of people who are cutting back on medications because of cost is not likely to decrease in the near future. The current research into medication affo rdability differentiates people into one of two categories, either repor ting or not reporting CRN. There has been much research performed regarding the prevalence of CRN, and factors associated with CRN, but there is not currently a way to determine wh at factors are most important to people when deciding to fill or not fill a prescription. It may seem obvious that the easiest way to reduce CRN would be to reduce medication prices; however, given that people with 20

PAGE 21

relatively high incomes report CRN, and people who already have medication choose not to take it for various reasons, other factors besides cost may be having a major impact. Essentially, we do not know what the true nature of the problem is for those who report CRN, and whether it is just an issue of access, or if other factors (such as ones that affect utilization in a broader sense) come into play. The main purpose of this project is to gain better understanding of what people are experiencing in situations in which they would report CRN. This research needs to be done because we do not really understand all relevant aspects of the problem when people report CRN. Undoubtedly, some of the people who report CRN are people who need the medication, want to ta ke it, yet cannot afford it. These people need to be correctly ident ified, and programs need to be put in place to assist them in the most efficient way possi ble. Without performi ng this study and getting a much better idea of all of the factors t hat people take into consideration when purchasing medication, it will be impossibl e to differentiate people with true affordability problems as compared to the rest who are reporti ng CRN but are taking other factors into consideration as well when making the decision to purchase or not purchase a prescribed medication. What can be done to help the rest of this sub-population that is reporting CRN? Right now, we do not understand all of the components affecting why they are reporting CRN, and thus we do not know if they need financ ial help at all. For example, we may find a subset of people with good insurance coverage who believe that a $10 copay is too much to pay for any kind of medication, r egardless of benefits it could provide. This person would report CRN, but reducing the cost to $5 may not change this persons 21

PAGE 22

attitudes towards the benefit of medication. Does this person even exist right now? Unless we embark on this or a similar type of research project, we will not attain greater understanding. For those who are currently reporting CRN, there can be a wide variety of circumstances that affect their decision-making. Until we can identif y all of the factors involved in CRN, determine their relative importance, and come up with a way to differentiate all of the different types of people that require different solutions to ensure that they do not experience CRN, then we really cannot solve the general problem of CRN. Research Questions 1. When people report not purchasing medi cation because of cost, what do they mean? 2. What differentiates instances wher e the same individuals purchase some prescribed medications and instances where they dont purchase other medications because of cost? Significance for Pharmacy Right now, many people report that t hey do not have access to essential medications because of cost. We know this based on population-based surveys that ask a simple, dichotomous question regarding underuse of medication because of cost. Media reports of Americans going without prescriptions because of cost have become commonplace (Felland & Reschovsky, 2009). Costs are going up, and people are reporting that cost is a problem. Currently, the problem exists and will seemingly get worse over time. Without a clear understanding of the possibly multiple causal factors involved when persons report CRN, it will be di fficult to design effective, targeted interventions to help those in need. 22

PAGE 23

Until one understands the various dimensions of the probl em, it is going to be difficult to solve it. Ma ny people are reporting CRN, yet evidence from the literature suggests that other factors may be involved that are affecting peoples decision-making process when it comes to purchasing/using medications. What is the relative importance of these other factors for people who report CRN? Do these other factors overshadow factors related so lely to cost? For those people for whom other factors (besides just cost) are impor tant, what kind of policy or program can be designed to help them? Do they really have a problem that can be solved? Without this kind of information, it is going to be difficult to identify and help the people that really need financial assistance. This study provided a means of addressi ng the questions posed above. By talking to people in detail about experiences for wh ich they reported CRN (especially compared to experiences when they did NOT report CRN and purchased a prescribed medication), valuable insight was gained into a variety of factors that went into the medication-acquisition decision-making proce ss. Identification of these factors, and hopefully the inter-relationships between them will be useful in clarifying what the exact problems are as well as describing how to identify people with t hese problems and what can be done to help various people. Without th is knowledge, it is difficult to see how progress will be made in helpi ng people access necessary medication. The generic solution of simply reducing drug prices may not provide a comprehensive solution. Underlying all of this is the implicit assumption that we, as a society and as a profession of pharmacy, want to work in a positive way for people to improve their 23

PAGE 24

24 access to necessary medication. If this assumption is true, then performing this study was a preliminary step in knowing what acti ons to take in achieving this goal.

PAGE 25

CHAPTER 2 LITERATURE REVIEW This literature review is comprised of f our major sections. The first section deals with the current state of CRN; it is intended to show that the extent and severity of CRN constitutes a major problem that needs to be addressed. The second section discusses how current researchers have tried to identif y and characterize the population that is reporting CRN. This is important to know as it represents an attempt to describe the characteristics of this subpopulation. The third section di scusses existing concepts that have been shown to be related to CRN. Knowledge of these concepts will serve as a foundation from which the event ual theory will be built. The f ourth section examines the use of grounded theory in medical research, and concentrates on similar research concerning drug use decision-making that has been performed. This will provide support for the use of grounded theory in addressing the research questions in this investigation, and the met hodological approaches used. Current State of CRN in the United States Prevalence Estimates A significant percentage of the population reports CRN (Piette et al., 2004a; Piette et al., 2004b; Heisler et al., 2004a; Piette et al., 2004c; Safran et al., 2002; Steinman et al., 2001; Piette et al., 2006b; Kennedy & Erb, 2002; Reed & Hargraves, 2003; Kennedy et al., 2004; Kennedy & Morgan, 2006) While the exact definition of CRN and the target populations differ between the diffe rent researchers, an overall estimate of the percent of the general population reporting cost as a problem for accessing and using prescription medication is around 20%. Table 2-1 summarizes the percentages, populations, and definitions for CRN used by different researchers. While some of 25

PAGE 26

these prevalence estimates include thos e over 65 (whose cost problems have subsequently been improved by Medicare Part D), the estimates still serve the purpose of giving one a general idea of the prevalence of CRN. While there is a wide variety in t he prevalence estimates and populations considered in each study, CRN appears to a ffect approximately 20 % of the overall population, which is a sizeable percent of the U.S. population. Consequences of CRN Consequences of CRN can be roughly catego rized into two general categories: health related and economic. Many of the health-related consequences can be thought of as being a subset of general adherence-re lated consequences (which will not be discussed in depth here), but some res earchers have focused on health-related consequences specific to medica tion underuse because of cost. In a longitudinal analysis using individuals own reports of their medication use, Heisler et al. (2004b) report that, for those who restricted medication use because of cost compared to those who did not, there was a significant decline in self-reported health status (adjusted odds ratio 1.76, 95% CI 1.27-2.44), there we re higher rates of angina for those with cardiovascular disease (AOR 1.50, 95% CI 1.09-2.07), and there were higher rates of strokes or non-fatal heart attacks (AOR 1.51, 95% CI 1.02-2.25). In a sample of poor and elderly persons whos e use of essential dr ugs decreased by 9% (elderly) and 14% (poor) because of larger out-of-pocket costs, serious adverse events and emergency department visits associated with reduction in the use of essential drugs both increased (Tamblyn et al.,2001). Ad verse events associated with reduced use of essential drugs increased from 5.8 to 12.6 in the elderly cohort ( net increase of 6.8, 95% CI 5.6-8.0), and fr om 14.7 to 27.6 in the poor cohort (net increase of 12.9, 95% CI 26

PAGE 27

10.2-15.5). Emergency department visits per 10,000 person-months increased by 14.2 in the elderly cohort (95% CI 8.5-19.9) and by 54.2 in the poor cohor t (CI 33.5-74.8). In a study of elderly Medicaid patients in New Hampshire w here drug use declined by 35% because of increased cost-sharing payments due to a cap of reimbursable medications a patient can receive, rates of admission to nursing homes increased (RR to control group =1.8, 95% CI 1.2-2.6) (Soumerai et al., 1991). In adults with a disability who reduced medication use due to cost, 52.5% reported a health problem due to medication noncompliance, in cluding having pain or di scomfort, experiencing a worsening of a condition for which medici ne was prescribed, ex periencing a negative change in vital signs, having to go the doctor or emergency room, or having to be admitted to the hospital (Kennedy & Erb, 20 02). Given that the consequences of medication underuse for whatever reason are widely known, none of these results should be surprising. There are also economic consequences rela ted to responding to medication costs by shifting or not spending the same money elsewhere. Several researchers have examined coping strategies specific to respondi ng to medication cost pressures. Piette et al. (2004b) report that in a population of elderly diabetic patients faced with medication costs more than they could affo rd, 28% spent less on basic needs (such as food or heat), 14% increased debt, 10% borro wed money from a family member or friend, and 36% reported using at least one of the three coping strategies. In adults taking prescription medications for one of fi ve chronic diseases and faced with high costs, 22% cut back on necessities, 16% increased debt, and 31% used at least one coping strategy (Heisler et al., 2004a). In a probability sample of noninstitutionalized 27

PAGE 28

Medicare beneficiaries age 65 and older, 10% with coverage and 20% without coverage reported spending less on basic needs in or der to pay for prescription medication. (Safran et al., 2002) In summary, for this roughly 20% of the population that are r eporting CRN, there are severe health-related and economic outco mes. Although it is beyond the scope of this study, it seems logical to conclude that the economic consequences, for instance cutting back on other necessities, would eventua lly manifest in worse health outcomes. Therefore, the total econom ic and health-related consequences of forgoing proper medication use because of cost are likely to be even greater than what has previously been reported. Identification and Characteriz ation of Those Reporting CRN As was summarized in Table 2-1, most of the research done in this area has relied on one simple dichotomous question to det ermine the prevalence of medication underuse due to cost: Have you taken less medication than prescribed because of the cost in the last year? Sometimes a similar variant is used such as Did you not fill a prescription due to cost at least once in the la st year? While this single question is able to generally identify those with the problem of interest, it does not shed much light on the various characteristi cs of this population. In order to learn more about the characte ristics of this subpopulation as compared to the general population, researchers have us ed two basic approaches. The first is to identify demographic and health-related factor s that are associ ated with greater probabilities of reporting underus e of medication because of cost, taken from these large-scale surveys. These fact ors are summarized in Table 2-2. 28

PAGE 29

The table indicates that various demogr aphic characteristics have been identified that influence the likelihood that someone will report CRN. While most of these seem relatively obvious (lower income, no insu rance or drug coverage, higher OOP costs, higher number of medications, worse health), i dentification of these characteristics is only able to give one a general population-based view of t he characteristics of people who are reporting CRN. Unfortunately, knowledge of just these factors is not enough to clearly understand the problem that people are having in payi ng for medication. Missing is the understanding of psychosocial variable s and other factors that may affect ones decision to not obtain a medication. In addition, we do not understand the relative importance of the factors t hat people take into consi deration when making decisions about purchasing prescriptions. The second approach used to help gain a better understanding of this subpopulation, has been by asking about more s pecific behaviors related to underusing medication because of cost. Mo st of these additional behaviors are specific variants to the original underuse question; others refer to coping strategies used in response to financial pressures. Examples of these are summarized in Table 2-3. From the information in Tables 2-2 and 2-3, one is able to gain insight on some of the characteristics of this sub-population. Most of the results make sense; for example, those with lower incomes or no insu rance have a higher likelihood of reporting underuse of medication because of cost. Others reveal a more detailed insight, such as how the rates of reporting underuse change depend ing on the disease st ate, or whether the medication is pain-relieving vs. preventive. Additionally, the literature shows the extent of many coping strat egies that people use as a resu lt of concerns about paying 29

PAGE 30

for medication. The strategies that people were found to use in a number of studies included reducing the dose of medication tak en (e.g. delaying refills, skipping doses), spending less on other basic needs, borrowi ng money from friends and family, and using complementary or alte rnative medicines instead of a prescribed medication. Concepts that have been Related to CRN Even though the grounded theory approach, specified by Strauss & Corbin (1998) (outlined in the next chapter) in which the theory is disco vered only from the data collected, will be used, it still makes sense to examine the relevant literature to determine what concepts have been shown to be important in previous research in the topic area. As Strauss & Corb in state, Concepts derived from the literature can provide a source for making comparisons to data at t he dimensional level. (1 998, p.49) In this fashion, when collecting data, the investigator can be cognizant of these concepts so he can compare and contrast what he is disco vering with what has already been described. Piette et al. 2006 One of the main researchers examini ng cost-related non-adherence, John Piette of the University of Michigan, has attempt ed to create a concept ual framework outlining all of the factors and their inter-relationships that affect medica tion underuse because of cost. All of the demographic and health-related factors mentioned abo ve are contained in this framework in the following categories: financial pressures, patient characteristics, drug characteristics, and diagnostic charac teristics. The framework also includes additional categories such as regimen comple xity, clinician factors, and health system factors. This framework is shown in Figure 2-1. The factors identified in t he framework are grouped into di fferent factors, such as clinician factors or Rx characteristics, and represent Piette et al.s initial attempt to 30

PAGE 31

categorize and make sense of how all of the di fferent kind of factors are related. For example, prescriber incentives could hav e an effect on a clin icians therapeutic choice, which could therefore affect whether or not the patient is able to purchase a prescription. In doing a literature search of studies which have referenced this article (thus would indicate that this framework is bei ng used to conduct research), no studies were found that tried to use this framework to predict reporting CRN. This leads one to question its utility as a framew ork to guide further research or as a guide for making positive changes in helping to design intervent ions that would ultimately reduce CRN. Therefore, while this framewor k does a good job of identifying all of the factors that have been associated with whether or not someone reports CRN, one still needs further information to determine the relative importance of these factors, and thus to more fully understand the nature of the problem when people report CRN. Andersen & Adays Behavioral Model of Health Services Use, Phase 5 (Andersen & Davidson, 2007) The Behavioral Model of Health Services Use, over the years, has attempted to describe how people utilize health care services The model includes factors that affect whether or not people can access health care services, and factors that affect whether or not people utilize health care services; what they term potential access and realized access. Although they do not explic itly try to determine factors that would lead one to report CRN (as Piette et al. 1996 attempt to do), the model does identify characteristics that would influence ones potential access to medication. The framework is show n in Figure 2-2. 31

PAGE 32

The three main categories of factors have been defined as predisposing (basic demographics, individual beliefs, etc), enabling (income, insurance, etc.), and need (perceived need for services). As the model has evolved, it has expanded to look at these factors not only from an individual perspective, but al so to look at the context (health system, national health policy) in which an individual operates. Researchers have used this model to de scribe how access to medication affects utilization, more than how the diffe rent factors could affect acce ss to medication in itself. In other words, the factors that could affect access (mostly the enabling ones such as income and usual source of care) are in effe ct used as proxies as to whether or not one has access to medication. Thus, having insur ance (for example) is taken as a proxy for having access to medication. For example, Smith and Kirki ng, (1999) in their study of determinants of utilization of HIV drugs, use income, insurance status, and usual source of care as independent variables to help ex plain resultant utilizat ion. In one study examining who used anti-hypertensive drugs in an insured population, Stockwell, Madhavan, Cohen, Gibson, & Alderman (1994) us e the number of physician visits as a proxy for having access to t he medication. In another study examining drug use among the community-dwelling elderly (Fillenbaum et al., 1993), insurance status and adequacy of income are proxies used for havi ng access to the said drugs. In another study examining drug use among caregivers of veterans (Sleath, Thorpe, Landerman, Doyle, & Clipp, 2004), perce ived financial adequacy and prescr iption drug insurance were used as proxies for having access to the drugs. There does not seem to be much research on whether or not variables from the Behavioral Model, which are used as proxies for access, really do lead to true access at the individ ual patient level. 32

PAGE 33

It seems as though many of these factors could affect whether or not someone reports CRN. A predisposing factor such as ones health belief could affect the value that one associates with purchasing and using a prescribed drug. Besides income, health insurance coverage and the health syst em in which patients participate will undoubtedly influence how much t hey will be required to pay for their medication. In summary, this model of health system acce ss contains many factors that would probably be applicable to this study which exam ines factors that affect reporting CRN. But currently, even though the model attemp ts to combine the stages of access and utilization into one framework, no delineation is made regarding which factors influence access and which factors influence other aspects of utilization; studies using this model simply use having insurance as a proxy for having access. Other Concepts from the Health Behavior Literature Different health behavior theories have been proposed which try to explain why people engage or dont engage in health behaviors, implicitly assuming that access to the service is available (Munro, Lewin, Swart, & Volmink, 2007). Each of these different theories identifies various fa ctors that have been shown to be important in affecting ones health behavior. As was postulated in C hapter 1, some of these factors could be influencing whether or not a person purchases a prescription, even if the person is reporting that they are not purchasing the prescription because of cost. Each of these factors has been shown to affect one s health behavior; since cost-related nonadherence (CRN) could be considered as a health behavior, each of these factors could have an influence on this particular behavior. It therefore made sense to look for these various factors when collecting data and creating the new framework. 33

PAGE 34

To identify which constructs to specifically elicit information about, a review of health behavior theories (Munro et al., 2007) was consulted, which identified nine major theories and reviewed the evidence supporting each one. Each of these theories was then examined for relevant constructs; relevancy was determined based on previous knowledge of the liter ature regarding CRN (cost-relat ed medication non-adherence) and adherence. In other words, relevant construc ts were those which were hypothesized to possibly play a part in peoples decision-ma king in whether or not to purchase a prescription; not all constructs were consi dered. Once these constructs were chosen, questions regarding them were included in the interview guide. The purpose of asking open-ended questions to elicit information on these constructs was not to obtain the precise level of a defined construct, rather, it was to get a general idea of whether people took these factors into consideration when making their decision on whether or not to purchase a prescription. A descrip tion of each of the theories used and the constructs chosen from each follows. Health Belief Model (Rosenstoc k, 1966; Janz & Becker, 1984) The original Health Belief Model (Ros enstock, 1966) was based on patients levels of the following four core items: perceived susceptibilit y in getting the disease, perceived severity of the disease, perceiv ed benefits of enacting the health behavior, and perceived barriers to enacting the health behav ior. Several other factors have been added, but the general idea is that people weigh these four factors when determining whether or not they engage in a health behavior. Some of the concepts represented by the co nstructs might be coming into play for people when deciding whether or not to pay for a prescription. People may consider the perceived benefits of a drug when deciding whether or not to purchase it. Similarly, 34

PAGE 35

people may perceive that barriers exist t hat would not let them purchase the prescription. How people perceive the severi ty of the disease and the consequences of not taking the medication may also contribute to peoples decision-making process. Therefore, questions regarding the general concepts represented by these constructs were included on the interview guide. Social Cognitive Theory (Bandura, 1986) The concept of outcome expectations, r oughly defined as anticipatory outcomes of a behavior (Glanz, Lewis, & Rimer, 1997) s eemed to be relevant for this study. It seems that people need to believe that doing something such as purchasing a prescription, would result in a positive outcome. In an arti cle summarizing common constructs from a variety of theories, B andura (2004) asserts that the constructs of susceptibility, severity, and benef its from the Health Belief Model are, in fact, outcome expectations. Perceived susceptibility and severity are the negative outcomes expected if one does not adhere to treatment reco mmendations and perceived benefits are the positive outcome expectations. Since the outcome expectations construct seems to cross many different health beha vior theories, subjects were asked what they believed would happen if they did not take the spec ific medications they had been prescribed that they report ed not obtaining because of cost. Summary In summary, many different theories hav e been developed to try to explain and predict health behavior. However, in terms of medication taking behavior, none of them distinguish between the two dist inct stages of access and utilization; they all seem to focus of factors that relate to utilization gi ven that access already exists. Still, there could be factors that affect utilization that ar e also affecting access. For this reason, 35

PAGE 36

relevant concepts from different health beh avior models were included in the interview guide, to determine if they had any part in determining whether or not someone purchased a prescription. Using Grounded Theory in Drug Use Research Grounded Theory was developed in the 1960 s by two sociologists at the University of California at San Francisco who were performing research on the dying process. The novel concept was that the theory constructed to describe whatever process or situation studied should be grounded in the data that are collected. Their research led to two classic books using the Grounded Theory approach as well as a book describing the methodology, The Di scovery of Grounded Theory (Glaser & Strauss, 1967). As discussed in Chapter 1, the under lying foundation of Grounded Theory is the perspective of Symbolic Interactionism coined by Herbert Blumer (1969), which generally states that meanings for how people act are derived from social interaction, including the meaning that people give to ot hers symbols and actions. This discovery of how and why people act can form the basis of a theory for a particular process, if one employs the general Grounded Theory approach. In order to ensure that a grounded theory approach could be used to answer my research questions, similar research that has been performed regarding patients and the factors that came into play for differ ent types of drug use behaviors (mostly in the adherence realm) was reviewed. In examining this research, methodological decisions that had been made were highlighted to dete rmine what approaches worked well or not well, and thus could help me make t he most appropriate decisions on similar methodological issues. 36

PAGE 37

The following studies identified in T able 2-4 were identified with a PubMed search of the terms grounded theory AND d rug use. Studies that were most applicable to this investigation were revi ewed. These included research where the authors stated that a grounded t heory approach had been used, that they were trying to determine how patients made decisions regard ing a drug use behavior, that the data collection consisted solely of patient interviews, and that the studies took place in either North America or Western Europe. The eight studies listed in the table are not intended to represent all that are available, but do help determine whether or not grounded theory would be useful in answering the research ques tions in this investigation and also would help in making methodological decisions for this study. Table 2-4 summarizes the methodological choices and results achieved by each of the eight studies. Saturation? refers to whether or not t he authors reported that they had reached saturation, defined by Glas er & Strauss (1967). Wait to start coding? refers to whether or not the res earchers reported that they waited to begin coding until they had done a certain number of inte rviews, or if they started the coding process immediately. Theoretical sampling? refers to whether or not the authors changed their sampling strategy to a more purposeful one as interviews went on, in order to continue with the constant comparison necessary to generate a true grounded theory. Modified interview guide? refers to whether or not the researchers modified their interview guide as time went on, or if the interview guide remained constant throughout. The first thing that is evident from the re view is that either a set of major themes or a preliminary framework was able to be established by each study. It does give 37

PAGE 38

evidence that using a grounded theory approach c an lead to discovering major themes of how patients make drug-use decisions, some thing that I am attempting to do with this study. Following a grounded theory approach should therefore be able to generate at least major themes if not a preliminar y framework to help answer my research questions. Many of the other observations are met hodological. The first observation centers on the idea of saturation and the resultant sample size. In qualitative studies, a priori sample size is impossible to determine, since one is not testing a hypothesis using inferential statistics. Rather the concept of saturation is used, originally defined by Glaser and Strauss (1967, p.61) as that no additional data are being found whereby the sociologist can develop properties of the category. The firs t five studies in the table (Bower et al., 2006; Dolovich et al., 2008; Bajcar, 2006; Stack et al., 2008; Viswanathan & Lambert, 2005) were familiar with the concept of saturation when embarking on data collection, and determined after <20 interviews that they were not discovering any additional major themes, thus they stopped collecting data and decided that saturation had been reached. The sample sizes in these five studies is comparable to that determined in a study of saturation perfo rmed by Guest, Bunce, & Johnson (2006), where they determined after re-analyzing t heir 60 interviews that they had reached saturation in 18 interviews, if not 12. Conver sely, the final three studies (Carder et al., 2003; Dowell & Hudson, 1997; McCorry et al ., 2009) all collected data on a certain number of interviews without considering wh ether or not they had reached saturation. In the McCorry et al. (2009) study, they determined afterward that saturation had been reached before the last 10 inte rviews were performed. 38

PAGE 39

Another methodological observation center s on the use of theor etical sampling, defined by Glaser and Strauss (1967, p.45) as the process of data collection for generating theory whereby the analyst jointly collects, codes, and analyzes his data and decides what data to collect next and where to find them, in order to develop his theory as it emerges. This process of data collection is controlled by t he emerging theory. Three studies in the table (Dolovich et al ., 2008; Bajcar, 2006; Viswanathan & Lambert, 2005) all reported using this process, in whic h the nature of sampling is dictated by the data one wants to collect. In this way, the people that they interviewed in later stages were sampled in order to further clarify or refine a theme or framework that was emerging. In the other studies (Bower et al., 2006; Stack et al., 2008; Carder et al., 2003; Dowell & Hudson, 1997; McCorry et al., 2009) they selected a pool of interviewees that fit their elig ibility criteria and started in terviewing them in some order, with no mention of a purposeful sampling to clarify themes. Finally, there is the question of when to begin analyzing the data. It would seem that in using the approach spelled out by Glaser and Strauss (1967), one would start analyzing the data beginning with the first observation, as this would begin to identify any major themes, and may lead to any different sampling strategies fo r future data collecti on. The downside to using this approach is that t he first handful of data observati on may be too influential in shaping the eventual theory. Table 2-4 indi cates that only one study (McCorry et al., 2009) mentioned waiting to begin coding and anal ysis, and there is no indication that they did this in order not to be too heavily influenced by the first few interviews. They state that each of the three authors started coding five interviews each, which indicates that they collected most of t heir data before starting any kind of analysis. Based on this 39

PAGE 40

small sample, it does not seem as though ther e is any evidence that researchers, when using the grounded theory approach, wait before analyzing data. It seems as though if one follows the process and codes accurately, the threat of the first few interviews steering one in the wrong di rection is minimized. Summary In summary, there is still much that needs to be learned about the population reporting CRN, and the factors they take into consideration when deciding whether or not to fill a prescription. Based on national surveys, CRN is a condition that affects roughly 20% of Americans, and results in adverse health outcomes. Much of the research that has been performed has sought to determine the demographic characteristics and other behaviors that are associated with reporting CRN. While these do give some insight, it does not give one enough information to design targeted interventions to help these people access thei r medication. In addition, concepts from other frameworks used to explain health behav ior may also be related to reporting CRN, so these concepts should be explored in the data-gathering process. The use of grounded theory has been shown to be effective in doing research regarding factors that influence the taking of medication; essentially a question of why did people behaved the way they did. This research attempted to address a similar human behavior problem using grounded theory, but instead of examining utilization of medication, it dealt with discovering the fact ors that influenced whether or not someone purchased a medication. The following chapter will highlight how grounded theory was used to do this. 40

PAGE 41

Table 2-1. Estimates of CRN by different researchers Source Percentage Population Definition of CRN Piette et al. 2004a 18% Nationwide panel of adults >50 Taken less medication than prescribed because of the cost in the last year Piette et al. 2004b 19% Nationwide panel of adults >50 taking medication for diabetes, depression, heart problems, hypertension, or high cholesterol Taken less medication than prescribed because of the cost in the last year Heisler et al. 2004a 18% Nationwide panel of adults >50 taking medication for diabetes, depression, heart problems, hypertension, or high cholesterol Taken less medication than prescribed because of the cost in the last year Piette et al. 2004c 11.9% VA patients 24.8% Medicaid, non VA 15.2% private 22.0% Medicare 34.6% uninsured Nationwide panel of adults >50 taking medication for diabetes, depression, heart problems, hypertension, or high cholesterol Taken less medication than prescribed because of the cost in the last year Safran et al. 2002 11% with coverage 25% without coverage 10% probability sample from CMS records of eight states was surveyed Didnt fill prescription due to cost at least once in the last year Steinman et al. 2001 2% with coverage 3% with partial coverage 8% with no coverage Nationally representative sample of adults >70 from survey of Asset and Health Dynamics Among the Oldest Old (AHEAD) Taken less medication than prescribed because of the cost Piette et al. 2006b 23% Nationwide panel of adults >50 taking medication for diabetes, depression, heart problems, hypertension, or high cholesterol Taken less medication than prescribed because of the cost in the last year Kennedy & Erb 2002 33.9% National sample of adults with disabilities Disability FollowBack Survey (DFS) supplement of the National Health Interview Survey (NHIS) Any of the following, because of cost: Did not refill when ran out Used less often than prescribed Did not get when first prescribed Did not get entire prescription 41

PAGE 42

Table 2-1. Continued Source Percentage Population Definition of CRN Reed & Hargraves 2003 7.3% private insurance 22.8% public insurance 30.7% no insurance A nationally representative telephone survey of the civilian, noninstitutionalized population aged 18-64, supplemented by in-person interviews of households without telephones Did not purchase at least one prescription drug because of cost in 2001 Soumerai et al. 2006 29% disabled aged 55-64 13% aged >65 National sample of noninstitutionalized Medicare enrollees Did not obtain 1 or more medicines prescribed in the current year because it was too expensive Kennedy et al. 2004 5.9% Panel of US households (NHIS) Did not get prescription medication in the last year because they could not afford it Kennedy & Morgan 2006 9.9% Household phone survey Did not obtain a prescribed medication due to cost Table 2-2. Factors associated with greater likelihood of reporting CRN Source Demographic or health-related factors Piette et al. 2004a Age 50-54 Lower income Higher monthly OOP costs No prescription drug coverage Conditions (asthma, heart failure, depression, high blood pressure, high cholesterol, diabetes) Piette et al. 2004b Female Younger Higher number of medications Higher monthly OOP costs Diabetes Heisler et al. 2004a Female Younger Lower income Higher monthly OOP costs Piette & Heisler 2004c Medicare, Medicaid, no insurance (compared with VA patients) Safran et al. 2002 Poverty Chronic condition No prescription drug coverage Steinman et al. 2001 No prescription drug coverage Lower income Higher monthly OOP costs Lower self-rated health More co-morbid conditions Piette et al. 2006b Use of symptom relief medications(compared with preventive medications) 42

PAGE 43

Table 2-2 Continued Source Demographic or health-related factors Kennedy & Erb 2002 Younger Hispanic Lower income Uninsured More severe disability Higher number of prescriptions Reed & Hargraves 2003 Non-white More chronic conditions No insurance or public insurance Soumerai et al. 2006 Worse health More comorbidities No drug coverage Kennedy et al. 2004 Age (18-64) Female African-American Lower income No insurance or public insurance Table 2-3. Questions to determine cost -related medication underuse and additional questions used by different researchers Source Main definition of CRN Other items asked about by specific questions Additional features Piette et al. 2004a Taken less medication than prescribed because of the cost in the last year Took fewer pills or a smaller dose Did not fill a prescription at all Put off or postponed getting a prescription filled Used herbal medicines or vitamins when felt sick rather than take prescription medication Took medication less frequently than recommended to stretch out the time before getting a refill By different conditions Frequency (last month, last 12 months) Piette et al. 2004b Taken less medication than prescribed because of the cost in the last year For diabetes only Frequency (last month, last 12 months) Heisler et al. 2004a Taken less medication than prescribed because of the cost in the last year Spent less on basic needs such as food or heat to pay medication costs Borrowed money from family or friends to pay medication costs Increased credit card debt to pay for medications By different medications 43

PAGE 44

Table 2-3. Continued Source Main definition of CRN Other items asked about by specific questions Additional features Piette et al. 2004c Taken less medication than prescribed because of the cost in the last year Took fewer pills or a smaller dose Did not fill a prescription at all Put off or postponed getting a prescription filled Used herbal medicines or vitamins when felt sick rater that take prescription medication Took medication less frequently than recommended to stretch out the time before getting a refill Spent less on basic needs such as food or heat to pay medication costs Borrowed money from family or friends to pay medication costs Increased credit card debt to pay for medications Worried about how to pay for medications at least once per month By different conditions Frequency (last month, last 12 months) Safran et al. 2002 Didnt fill prescription due to cost at least once in the last year Skipped doses to make prescription last longer Spent less on basic needs such as food or heat to pay medication costs Frequency (one or more times, three or more times) Piette et al. 2006b Taken less medication than prescribed because of the cost in the last year By condition Preventive vs. symptom relief Kennedy & Erb 2002 Any of the following, because of cost: Did not refill when ran out Use less often than prescribed Did not get when first prescribed Did not get entire prescription Reed & Hargraves 2003 Did not purchase at least one prescription drug because of cost in 2001 Soumerai et al. 2006 Did not obtain 1 or more medicines prescribed in the current year because it was too expensive Skipped doses to make the medicine last longer Took less medicine to make the medicine last longer 44

PAGE 45

Table 2-3. Continued Source Main definition of CRN Other items asked about by specific questions Additional features Kennedy et al. 2004 Did not get prescription medication in the last year because they could not afford it Kennedy & Morgan 2006 Did not obtain a prescribed medication due to cost Table 2-4. Characteristics of studies using grounded theory Study General topic End product Sample size Saturation? Theoretical sampling? Wait to start coding? Bower, Frail, Twohig, & Putnam (2006) Choosing osteoarthritis medication 4 main themes 16 Yes, as decided by researchers No No Dolovich et al. (2008) Taking medication Conceptual model 18 Yes, as decided by researchers They report they did, but unclear No Bajcar (2006) Taking medication Model with 4 categories 10 Yes, in core category Yes No Stack, Elliot, Noyce, & Bundy (2008) Taking medication with diabetes and cardiovascular disease Set of themes 19 Yes, as decided by researchers No No Viswanathan & Lambert (2005) Taking antihypertensive medication Conceptual model 20 Yes, as decided by researchers Yes No Carder, Vuckovic, & Green (2003) Long-term medication use Set of themes 83 No No No Dowell & Hudson (1997) Taking medication Conceptual model 50 No No No McCorry, Marson, & Jacoby (2009) Taking antiepileptic medication 5 major themes 47 Yes, found nothing new in last 10 interviews No Waited until 15 interview s were complete 45

PAGE 46

46 Figure 2-1. Conceptual framewor k of factors affecting cost -related medication underuse (Piette et al., 2006a) Figure 2-2. Andersen and Aday s Behavioral Model of Heal th Systems Use, Phase 5 (Andersen & Davidson, 2007)

PAGE 47

CHAPTER 3 METHODS This chapter discusses the methodology behind the grounded theory approach that was used to answer the following research questions: 1. When people report not purchasing medi cation because of cost, what do they mean? 2. What differentiates instances where people purchase medication and when they dont? As mentioned in Chapter 1, a gr ounded theory approach was used to answer these research questions. Grounded theory is a qualitative research methodology that is defined as the discovery of theory from data (Glaser & Strau ss, 1967, p1) and is used in situations where one wants to gai n insight on why a particular process is happening, and what factors are causing things to happen. Grounded theory and its methodology were based on the ideas set forth by Glaser and Strauss in their seminal work The Discovery of Grounded Theory (1967) and on related works by these founding authors As time has progressed, the optimal process of doing grounded t heory building has been divided into two camps as represented by each of the two founding authors. While ther e is substantial agreement on many of the major points, they do repr esent two different approaches to conducting the research; thus one must decide which of the two to ultimately follow. The main difference between the two camps is how much knowledge a researcher should have before embarking on the project. Should the researcher try and enter the specific situation with as little knowledge as possibl e, so as to not introduce bias based on a priori concepts and frameworks, so the new theory can truly emerge? Or should one have extensive knowledge of the situation beforehand and use this knowledge to initially 47

PAGE 48

make sense of the data, but then be open enough to realize when new categories emerge so that a new t heory is not forced into a preexisting one, but rather can be discovered based on the data? The first app roach is the one advocated by Glaser (1978), while the second one is the approach advocated by Strauss (1987) and later by Strauss and Corbin (1990 & 1998). The Strauss & Corbin approach was followed in this study, with the understanding that the general principles laid out in The Discovery of Grounded Theory (1967) are truly what gu ide the grounded theory approach. The process Strauss and Corbin descri be looks at the continual interaction between sampling, data collection, coding, anal yzing, and finally theory construction. While all of these elements happen in conj unction with one another, they are separated below in order to provide more detail into each step. Sampling The type of sampling used in this investi gation is called theoretical sampling by Glaser and Strauss, and is defined in T he Discovery of Grounded Theory (1967, p.45): Theoretical sampling is the process of data collection for generating theory whereby the analyst jointly collects, codes, and analyzes his data and decides what data to collect next and where to find t hem, in order to develop his t heory as it emerges. This process of data collection is controlled by the emergi ng theory. Thus the people sampled in this study were those w ho reported that they had not purchased a medication because of cost at some point in the past 12 months, as this was the population of interest. The sampling was also a convenience sample of persons seeking primary care at a clinic that served a low-income population. Specific screening characteristics to identify the population of interest are given in Table 3-1. 48

PAGE 49

The reason that the second characteri stic was included above was to weed out those who have never purchased a medication before. While these never-users would undoubtedly give some insight, those who have both purchased a prescription and not purchased a prescription because of cost could id entify different factors that came into play in these two situations. It was thought that this ability to contrast different decisions on purchasing prescribed medications would help to identify which of the factors are the most important for the people participating in this study. Minors were excluded because many purchasing decisions would probabl y have been made by their parents or a responsible adult, and adults would be more li kely to provide be tter insight to the situation. With the recent advent of Medi care Part D in 2006 and its unique methods of providing prescription coverage to those over 65, any insight these Medicare eligible adults could provide would und oubtedly be influenced by the c haracteristics of the Part D plan they had chosen and the resultant bureaucratic issues involved. Since the purpose of the study was not to describe the degree of success of Part D, those 65 and over were excluded from the study. All subject recruitment took place at a single location, the Community Clinic, Inc. (CCI) clinic located in Silver Spring, MD. CCI is an organization accredited by the Joint Commission on the Accreditation of Healthca re Organizations (JCAHO) that provides high-quality primary care and health-related services for medically under-served persons. The population that they serve is those that are generally under-insured, and need a place where they can receive high-qualit y primary care. Most of the clients are Latino, and most clinic staff members are flu ent in both English and Spanish. The clinic 49

PAGE 50

provided permission for the investigator to us e the lunch room / conference room as a place to conduct the interviews. From the clinic waiting room, one could look down a short hallway past the reception desk to a conference room. Whenever there were at least two new clients in the waiting room, the inve stigator approached them as a group (with a lanyard containing his University of Flor ida ID) and made the following points: He was a graduate student doing a research project regarding people who had not purchased a medication because of cost. He was interested in interviewing people to learn about situations where this had happened to them. The interview would take about 15-20 minutes, could be done while waiting for your appointment, and concluded after your appointment if necessary. To thank you for your participation, you will get a $20 gift card from Safeway. If you would like to participate or have any questions, please come see the investigator in the conference room (poi nting down the hall wher e it was visible). At this point, the investigator would go back to the conference room and sit where he could be seen from the waiting room, and wa it for potential subjects to approach him. When they came, he answered their inquiri es about the study, as appropriate, and verified that they met the inclus ion criteria. If t he inclusion criteria were met, subjects were told in more detail what the purpose of the interview was and what they were being asked to contribute in terms of inform ation obtained for the interview so they could provide informed consent to participate in the study. The informed consent form was used as a template to convey this informati on. Once informed consent was given by the subject, the interview was performed in t he conference room. The subject had the option of leaving the door open or closing the door. 50

PAGE 51

The sampling occurred in roughly three basic steps as described by Strauss and Corbin (1998). The first step was called open sampling, with the resultant open coding, where the aim was to generally discover, name, and categorize phenomena according to their properties and dimensions. It was essentially the means by which the investigator was able to identify the major categories in volved in the process he was examining. The open sampling gradually evol ved into relational and variational sampling. As described by Strauss and Corb in, Sampling still proceeds on the basis of theoretically relevant concepts (categories), but the focus changes.the aim is to look for how categories relate to their subcategorie s as well as to further develop categories in terms of their properties and dimensions (Strauss & Corbin, 1998, p.209-210) The relational and variational sampling gradually evolved into discriminate sampling, where, given that the main categories have been es tablished, the focus is on examining concepts that will maximize opportunities for comparative analysis. (Strauss & Corbin, 1998, p.211) In this phase, more care was taken during the interviews to probe more deeply into the relati ve importance given to each of the main categories, in order to get a better idea on their decision-making process. The major difference that occurred during each of the thr ee phases mentioned above is that the focus of the interviews changed once the categories became better defined. In the open sampling phase, more time was spent in just letting the subjects talk about their experiences in a more openended fashion (approximately the first 8-11 interviews). In the variational sampling phase, now that the major categories were identified, the subjects were probed in mo re detail regarding s ub-categories and/or specific situations in whic h the categories could be further developed. In the final 51

PAGE 52

discriminate sampling phase, the focus was to make sure that all of the new subjects either fit into the existing schema, or that their experie nces could further enhance or define one or more of the categories. Data Collection / Interview Guide The data collected was in the form of se mi-structured interviews, as this was believed to be the most efficient way of co llecting the data. Observing people would not give the investigator insight into the fact ors that they consi der when purchasing a medication, nor were there any secondary sour ces (such as documents or records) that could give him this information. The interview consisted of three differ ent sections: an initial section with openended questions in which participants descr ibed their experiences of both not purchasing prescribed medications because of cost and of purchasing medication, a second section with more closed-ended questions that could give insight into various concepts that were believed to affect a per sons decision-making process, and a third section in which basic demographic information was collected. Each of these sections is described in more detail below. The in itial interview guide is included in the Appendix. The first section of questions began with di scussion of the most recent experience participants had about not purchasi ng or underusing a medicati on because of cost. The interviewer encouraged participa nts to talk about their exper ience and tried to gather whether or not there were any additional factors, besides co st, that entered into their decision-making process. If none were ment ioned, participants were directly asked about additional factors. Determ ining what these additional factors were, if any, was a central component of the theory to be constructed. The investigator also determined 52

PAGE 53

from their answers whether or not they could afford to purchase the medication and chose not to, or whether they truly c ould not afford the medication. Next, specific questions were asked regardi ng what the actual price was and how much they would have paid for the medica tion. The purpose of asking about the original price and the amount they were willing to pay to get an idea of how much of a difference there was between the two figures. Hearing that the actual prescription price was $15 and the person only wanted to pay $10, versus another scenario where the actual prescription price was $100 and the person only wanted to pay $5 gives valuable insight into each of the different scenarios. These questions helped to provide a clearer picture of the context w hen the person reported CRN. The next question concerned what they (o r someone else) did in order to reduce the price. By hearing what acti ons they undertook in reaction to the original price, it was hoped that more insight could be gained about what factors were most important in whether or not they ended up acqu iring the original prescrip tion, or if they had to do something else. Furthermore, I asked about what they would recommend to help them (or someone else) pay for a prescription the next ti me a similar situation arose. Since the ultimate application of the research findings is to try to assist people in accessing their needed medications, it made sense to inquire as to what they believed would be the most helpful to them. Unfo rtunately, most of the answers that were given for this (everybody should have coverage) were not as insightful as were initially hoped, so this line of questions was stoppe d after the open sampling phase. 53

PAGE 54

The final set of questions in this first se ction related to the participants last experience where they did purchase a medica tion. These responses were compared and contrasted with those fr om the experience when t hey did NOT purchase a medication to get a better idea of what factor s went into their decision-making process. If respondents were not able to articulate exactly why they did not purchase the medication in response to more open-ended questions, it was hoped that having them talk about this experience where they obtai ned a prescribed medication would make them think more carefully about how the two circumstances were different. The second major section of questions consisted of more directed questions aimed at measuring general concepts taken from the health belief and health education literature. As menti oned in Chapter 1, some of the c onstructs may be playing a part in peoples decision-making processes in det ermining whether or not to purchase medication. The general concepts asked abou t, and their relevance to the study, were discussed in Chapter 2. The final set of questions collected the basic demographic information as shown below: Age Race /ethnicity Gender Annual household income (<$20K, $20-<40K, $40-<60K, $60K) Number of peopl e in household Insurance coverage These questions come from the framework developed by Pi ette et al. (2006a), and are included because they have all been shown to be related to whether or not a person reports CRN. 54

PAGE 55

In using the theoretical sampling as de scribed above, the nature of the interview changed as the study evolved. In the open sampling portion, most of the questions were open-ended, giving people more of an opportunity to discu ss different factors that shaped their experiences. As major ca tegories and sub-cat egories emerged, the interview questions necessarily became more focused. So, while the investigator continued to acquire the personal and medica l characteristics as mentioned above for each person, the nature of the questions ch anged over time based on what the analysis of the data was showing. Given the relatively small number of in terviews that were thought to be needed, the investigator conducted all of the interv iews himself. He presented himself as an academic researcher with no ties to the pharmaceutical industry or any insurance company. Each interview was audio-reco rded only, and then transcribed by the investigator. Each transcription was doublechecked for accuracy by listening to the interview a second time and comparing the t aped interview with the transcript. In addition, after each interview, the investigator took notes regarding his general impressions of the interview, including what he believed t he major topics were and what he learned from the interview. After checking the transcription, he read through the transcribed interview and his notes to compare any differences between the two different ways in which the data from the interviews were processed. Data Coding During the initial or open coding phase, in which the transcripts from the first few interviews were analyzed in order to init ially develop categories, the original coding scheme was developed by utilizing the st andard coding paradigm spelled out by 55

PAGE 56

Strauss (1987, p.27-8) consisting of the following categories of things to look for in the data: 1. Conditions (e.g. circumstances su rrounding not getting a prescription filled because of cost and the last situati on where a prescription was filled) 2. Interaction among the actors (e.g. co mmunication with physician prescribing the medication, family situations that may have affected the decision about medication use) 3. Strategies and tactics (e.g steps the person took to tr y to obtain the medications, alternatives the person us ed to try to self-treat) 4. Consequences (what the effects were of not taking medication that was not obtained, what the effects were of taking the other medication when it was prescribed) A second investigator with extensive exper ience in performing qualitative research (and experience working in a pharmacy) was br ought in to independently code all of the interviews. Once the interviews were done, the two investigators reviewed how each interview was coded; discrepancies between the two were discussed and a consensus agreed upon. Currently, many software programs ex ist to help researchers code and analyze qualitative data, such as Atlas.ti or Nud.ist. In these software programs, it is easy to establish coding definitions and schemes, search for multiple instances where a code or combination of codes appears, and develop a hier archical structure of major categories and sub-categories. Usually these programs are of grea ter assistance when dealing with large amounts of data being coded and analyzed by large numbers of people, as an optimal way to stay organized. Given t he quantity of information (transcriptions of ~20 interviews) and the small number of pe ople doing coding and analysis, the primary investigator did not feel that it was nece ssary to use a software program. The second 56

PAGE 57

investigator, who worked independently of the primary investigator did use a software program to assist in c oding and analyzing the data. Data Analysis Using the grounded theory approach, data analysis was an iterative process. As one gathers new data, one continually (Morse & Richards, 2002): Codes it Compares it with previous data Analyzes it and identifie s themes or categories Discusses it with the expert and my committee Compares it with the literature Determines what data to gather next By using this process, major categorie s emerged to give the investigator a theoretical perspective on the si tuation of not purchasing a medication because of cost. To assist the investigator in this, memos, described by Strauss and Corbin as written recordsthat contain the products of analysis or directions for the analyst (Strauss & Corbin, 1967, p.217) were also wr itten. These were written ev ery four to five interviews, and assisted in developing theory as the data were being collected, coded, and analyzed. These memos also serve as a kind of a journal of the research study, to give evidence of how the final product was arrived at. The nature of t he memos changed as the sampling went from open to relational and discriminate sampling. The initial memos focused on just trying to understand the major factors that were involved. The later memos focused on the established fact ors, and attempted to determine the relationships between them. After each memo was written, it was shar ed with the investigators major advisor, who reviewed his memos and discussed the ca tegories and framework that was being 57

PAGE 58

developed. All analyses were also present ed to my committee for their comments and guidance in planning subseque nt data collection. This process continued until the categories that were explored attained a level of saturation, defined by Glaser & Strauss ( 1967, p.61) as that no additional data are being found whereby the sociologist can develop properties of the category. Essentially, this means that I had found out as much as possible about the various categories and sub-categories t hat I had chosen to focus on, and that additional data collection was not improving my understanding of the situation. As mentioned in Chapter 2, similar studies using grounded theory to learn more about drug-use decisionmaking reached saturation in 20 or fewer interviews, when the researchers were cognizant of saturation while collecting, coding, and analyzing their data. Saturation of the major categories in this study was reached in 21 interviews. Theory Construction As discussed above, the end result of this process was development of a theoretical framework consis ting of categories, sub-cat egories, and the hypothesized relationships between them. To ensure t hat the theoretical framework was generated from the data collected, t he second investigator created her own framework independently, which was then compared with the original fr amework. The resulting frameworks contained the sa me key categories, and essentially melded together to create the final framework. T he framework that resulted from this study can be used to guide future researchers in the study of CRN. Using this grounded theory approach, the framework came completely from the data collected. 58

PAGE 59

Scientific Rigor Any kind of research that is performed, be it quantitative or qualitative, must be done in a fashion such that reasonable steps are taken to maximize th e internal validity and the reliability of the results. For quant itative research, the choice (and proper execution) of the sample, study design, and instruments used for measurement make this process not necessarily easier, but at least more straightforward. For qualitative research, given that the sample is necessari ly not representative, the data collected not reproducible, and the coding and analysis dependent on the researcher, ensuring internal validity and reliability is still possible, yet more difficult to accomplish. Lincoln and Guba (1985) outline four crit eria that should be used to evaluate the scientific rigor of qualitative studies: credibility, transfe rability, dependability, and confirmability. Credibility, or whether the results and interpretations seem credible or have truth value (Lincoln and Guba, 1985, p.296) to those bei ng studied, should necessarily be achieved over the course of the proj ect. Since the constructed theory was based on the data gathered during the project, and data sampled later on in the study (during the discriminate sampling phase) were used to conf irm the theory, the re sults are thought to provide some truth in describing the exper iences of individuals interviewed in the study. Transferability refers to the ex ternal validity of the study. By definition, the results from qualitative research are not necessarily valid to differ ent populations, settings, etc. The goal of the qualitative researcher is therefore to explain as explicitly as possible the characteristics of the sample, how the samp le was recruited, and what questions were asked. In this manner, subsequent resear chers can compare their methods to mine 59

PAGE 60

60 and make appropriate judgments as to how t he results that were generated should be modified given the inherent di fferences in our methods. Dependability refers to how reliable the results are. Given the way that data are collected and analyzed in doing qual itative research, it is theoretically impossible to show that the resultant constructed theory is completely reliable. The investigator may have subtly altered interviews from one person to the next, or analyzed the transcript slightly differently on a different day. T he only way to get around this problem was to have an independent investigator examine my coding and anal ysis independently. In addition, the memoing process served as a way for any independent researcher to monitor my thought process. In this manner, the results were as reliable as possible, given the research methods used. Confirmability, or objectivit y, was achieved in the same manner as described in the previous paragraph. By explaining my methods, and by getting independent confirmation of data analysis and theory co nstruction, independent researchers could confirm the results that were achieved. Table 3-1. Characteristics of population Characteristic Operational definition Have not purchased a medication because of cost During the past 12 months, was there any time you did any of the following, because of cost: 1. Not purchase a prescription 2. Not purchase a refill 3.Not get entire prescription filled Have purchased a prescription before During the last 5 years, did you ever purchase a prescription? Age 18-64

PAGE 61

CHAPTER 4 RESULTS This chapter discusses the results of the research, namely the framework that was developed using the above-descri bed methodology. After initia lly discussing the sample from which the data was collected, exampl es will be given from the collected data illustrating how each component of the framework was discovered and ultimately constructed. Sample General Description As mentioned in the previous chapter, all of the respondents we re recruited from one location, a clinic in Silver Spring, MD that provides medical and pharmaceutical assistance to the under-insured. While the population was loosely defined by the original inclusion criteria (have reported CRN, have bought a prescription, aged 18-64), there seem to be other characte ristics of this population that should be defined, to make sure that further research that is desired to be done on a similar population can be sure of what that population really is. One must consider th at these are people who would go to a clinic to get medical help. T herefore, they had k nowledge of a medical condition, and had the ability to seek out me dical attention. This population also expressed the belief that going to the clinic (and being provided with a therapeutic plan) would ultimately help them, assuming t hat they could carry out the plan. Demographics As shown at the end of t he Interview Guide in the Appendix, a variety of demographic information was collected to not only better describe the population of this study, but also because various demogr aphic factors have been associated with 61

PAGE 62

reporting CRN in past research. Of the 21 respondents, 14 (67%) were female and seven (33%) were male. Ten (48%) identifi ed themselves as African-American, five (24%) as White, five (24%) as Latino/Latina/Hispanic, and one (5 %) as Asian-American. Four (19%) were aged between 18-29, sev en (33%) were aged 30-39, two (10%) were aged 40-49, five (24%) were aged 50-59, and th ree (14%) were aged 60-64. For selfreported overall health, tw o (10%) rated themselves poor, 12 (57%) reported themselves fair, four (19%) rated themselves as good, none (0%) reported themselves as very good, and three (14%) rat ed themselves as excellent. In terms of household income, 15 (71%) reported a household income less than $20,000 a year, five (24%) reported a household income between $20,001 and $40,000 a year, and one (5%) reported a household in come greater than $60,000 a year (this was a college student who was considering the income of hi s household with his parents, even though he was living on his own while attending school). Insurance status was a little more difficult to determine; many people fluctuated in and out of coverage, something that will be discussed at length in the next chapter. At the time that they reported CRN, four (19%) repor ted having insurance. Five (24%) had insurance coverage right before their CRN ep isode, but lost coverage temporarily and thus were not covered at the time of t he CRN. The remaining 12 (57%) did not have any insurance coverage. The 21 study participants listed a total of 33 medical conditions for which they reported CRN (after asking about the most rec ent episode, the investigator asked if there were other situations that they wanted to talk about). Twenty-one of the 33 could be considered chronic conditions; the rema ining 12 could be considered conditions 62

PAGE 63

where one would expect that t hey would resolve themselves in a short period of time. The most common chronic condition was high blood pressure (five instances), followed by chronic pain (three instances), choleste rol, depression, asthma, and allergies (two instances each) and bipolar, anxiety, bronchiti s, glaucoma, and hair loss (one instance each. The most common short-term conditions was a viral infection (5 instances), followed by stomach problems and short-term pain (two instances each), and high blood pressure during pregnancy, short-term hair problems, and short-te rm nail/skin problems (one instance each). As the conditions above indicate, CRN was reported for a wide variety of conditions. Sample Size as Determined by Saturation of Major Concepts As mentioned in Chapter 3, there was no a priori sample size calculation of how many people were to be samp led, but that sampling woul d be continued until saturation of the main factors or them es emerging from the interviews was reached, defined by Glaser and Strauss (1967, p.61) as that no additional data are being found whereby the sociologist can develop properties of t he category. Using the grounded theory approach of continually coding, analyzing, and sa mpling, it seemed as though all of the major factors had been developed by around 17-18 interviews, although it is difficult to pinpoint an exact time. In the last 3-4 interviews, it seemed as though no new concepts were being discussed, that their descriptions of their CRN experiences were fitting into the major categories that had been developed. Therefore, it was decided that saturation of the categories to be discussed below occurred after 21 interviews. Acquiring Medication This research was intended to discover w hat potential factors, if any, people took into consideration when not acquiring a pre scription, reportedly because of cost. 63

PAGE 64

Accordingly, the people recruited for this study were those that reported that they had not acquired a prescription because of cost, and they were asked about their experiences of the last time that this occurred. While they initially discussed what factors were involved when they did not acqui re the prescription, they also spoke about how they responded to this initial setback, and what they ended up doing with regards to ameliorating their medical situation that brought on the need for the prescription. Surprisingly, 9/21 actually eventually ended up acquiring the medication. This suggests that asking someone about what seems like a one-time event does not capture what is really occurring. Accordingl y, factors that have an effect on whether or not someone purchases medication do not just occur at one time point, but rather can occur over a period of time, as different actions are taken by the person wishing to acquire the medication and by other relevant parties. The resulti ng framework therefore attempts to capture this dynamic by not just examining factors that affect a decision at one point in time, but rather what happens du ring the whole time from which a person initially attempts to acquire a medication until the time that they either acquire it or do something else. Overview of th e New Framework Using the grounded theory approach of conti nual coding, analyzing, and constant comparison, major factors em erged that were the most important in determining why someone was not able to initially acquire a prescription, and additional factors that affected a persons decisions and actions if they were not offered immediate help in obtaining the prescription. The framework is shown in Figure 4-1. There were two distinct phases in the process, which begins when the person gets a prescription that they would like to get filled. The first phase relates to the initial 64

PAGE 65

attempt to acquire the prescrip tion. Two sets of factors were identified that affected whether or not a person was able to acquire the prescription: help factors and monetary factors. The research was origi nally intended just to look at the factors that affected this initial attempt to acquire the prescription, but the study participants provided additional insight into what they did after they were not able to acquire the prescription, leading to some of them eventually obtaining the prescription. The second phase involved whether or not people sought out help or additional resources to acquire their prescription, which depended upon their re-evaluated perceived need for the prescription and their knowledge / beliefs that help was available. At the end of the process, they either eventually acquired the prescription, or lived with an alternative therapy or simply went wit hout any therapy. Each of these phases will be discussed below. Attempt to Acquire Desired Prescription Whether or not they initially acqu ired the desired prescription depended on two types of factors: help factors and monetary factors. At this stage, there were two types of help factors: other persons provision of he lp to a passive patient, and the patients willingness to access known source s of help. Similarly, t here were two main monetary factors: the out-of-pocket (OOP ) prescription cost and the amount they were willing to pay. Those who either received help provided by others, were willing to access known sources of help, or were willing to pay the OOP prescription cost acquired their prescription. Those who did not have help provided, were not willing to access known sources of help, or were not willing to pay the OOP prescription co st did not initially acquire the prescription. Each of thes e factors will be described in detail below. 65

PAGE 66

Help Factors Other persons provision of help to passive patient Some people were able to get imm ediate assistance from a health care professional, where it seem ed that the help was offered to them, as opposed to the patient actively seeking it out and finding it on their own. These were instances where it seemed that the health care professional acti vely tried to assist someone once they were made aware of the problem the patient was having in acquiring their medication. One example was from a woman in her early 60s who had just been diagnosed with glaucoma, and could not acquire the prescription eye drops. I only had so much money, I was just ge tting ready to go on Social Security and it was very hard because some of the medications were very, very high, and I couldnt spend $100 per shot. So I just couldnt get things until I got on to a project called Project Access where I could get certain medication and certain doctors at a much lower cost I got it through here (the clinic where the interviews were taking place) Its supposed to be low income, you know, and what treat ment I was getting, Dr. X referred me to a certain doctor. It seemed like the doctor at the clinic made t he initial positive st ep of referring her to another doctor, which led to her getting into a program that offered medicine and other care for a much lower price. Another example came from someone who had just had her insurance cancelled. She came to the clinic to get new prescriptions for high blood pressure: Yes, yes, she told me that today I need it, and I dont have some money so she gave me some samples. Again, in this instance, the help came from the clinic doct or, who was aware of her situation and knew what to do to help her So, for some people, right at the time where acquiring medication was going to be a problem, a health care professional actively did something to help the pat ient acquire their medication. 66

PAGE 67

Patient willingness to access known sources of help Some people had pre-existing knowledge of the kind of assistance available for those who cannot acquire a medication becau se of cost, which directly impacted whether or not they eventually acquired the desired pre scription. For those who knew of an easily accessible location or program t hat assists people in acquiring medication, they went and utilized these services and acquired the prescription quickly. Some people had knowledge of a particular assistance program or place to go to get assistance from previous experience. One young man knew of a program that a hospital ran to help anyone acquire medications, based on his using it previously: (You) go to someplace like Holy Cross or Montgomery Gener al and just say you dont have insurance, or money to afford it. When you go to the hospital, they give you a form to fill out, and you take that to the hospital and then you have to pay 20 dollars every visit. Since this person knew about this program from a previous experience, he was able to use this information the next time t hat he was faced with attempting to acquire a prescription that was more than he was willing to pay. Another example comes from a woman who immigrated to the United States, and who suffered from allergie s. For a long time she did not acquire any medication because of cost, and just lived with her allergie s. She did not know that help for her to acquire this medication existed unt il a random visit to her manicurist: And then I was talking with the girl, and she told me about this place (the clinic). So we came together (referri ng to family), and I applied, so they helping me now. From that point on, she was able to acquire her medications for any future medical problems that arose. 67

PAGE 68

A similar example came from another immigrant woman who got sick fairly soon after she arrived in the United States. At the time, she did not know of any places to get assistance, and thus had to suffer through her flu and throat infection without any prescription medication. However, she l earned about the community clinic that would later on be able to help her: You know, because I came to the country new, so I dont know nothing about itFirst thing I learned is bec ause I came here in this community clinic and they have the WIC program for babies, I come with my cousin, she have a baby and now I start looking at the papers, they have it on the wall And then I learn. So now t hat Ive got my kids I still come here So many different program now. Its good, but I dont know about it, maybe they have them before, but I dont know about it. Once she learned about the clinic that c ould help her, she started going there for all of her medical needs. All of the three examples above indi cate that if a person knew about a particular clinic or progr am, then they would utilize th at knowledge in the future when they came to a situation in which they wanted to acquire a prescription but could not afford the medication. Monetary Factors Out-of-pocket prescription cost The cost of the prescription was most ly governed by ones insurance status, or rather, the phenomenon of havi ng the out-of-pocket (OOP) cost reduced by being covered by a third-party payer. For the ma jority that did not have any third-party coverage, the out-of-pocket cost was simply the market price of the drug without any kind of reduction in price that can be levera ged by larger institutions negotiating with drug companies. However, having some kind of third-par ty coverage did not guarantee that the OOP cost was low enough that people were willing to pay for it. In all of the examples 68

PAGE 69

below, people believed that t hey had adequate third party payer coverage to assist them in acquiring their prescriptions. Prescription not covered by third-party payer One person was covered by private insurance through his parents, yet th e insurance company did not cover a drug for hair loss that he wanted to acquire: I tried to get it through insurance, but insurance doesn t cover it. Because its not patented yet, so I thin k insurance companies dont want to risk it or something like that. I dont know. My doctor gave me an explanation of why it wa snt insured. In this case, having private insurance o ffered no help in reducing the price of the prescription, putting him in the same position as those without any third-party coverage. A gap in coverage from the third-party payer Four people experienced a gap in coverage, defined as having some kind of third-party coverage and then having it stopped for some reason, yet each person was confident that coverage would continue once the issue was resolved. Two people ex perienced a gap in coverage because of moving: (W)ell I was in between insurance with gapping because I had moved, so I left a place where I had insurance when I came down to Ma ryland they said my insurance is over, even though I have called the insurance company and they said the insurance is fine. Well, there was a mix-up, there was a mix-up in, I had moved and because my medication, my pain medication is a nar cotic, I cant, they cant call it in they had another pharmacy in DC that did deliveries, and they delivered the medication to my old address, and I nev er got it, so I couldnt afford to pay for it, because the insurance wont cover two prescriptions written in the same month, Both of these people had coverage, and we re fully expecting the coverage to continue after the gap, but while in the gap and thus without coverage, the OOP cost of their prescription was much larger than usual. 69

PAGE 70

Another person experienced a gap in coverage when beginn ing a new job: I mean, just like, the other day, my insurance ended, so I had to pick up some prescriptions. I tried to get there before it ended, but I didnt. I think one medication was $232. I was ge tting insurance through the state, and, because Im working now, they dropped me, so, and I didnt pick it up from here, so Im gonna have to (pick up coverage through her new employer), cause I need my blood pressure medicine. This person was transitioning from not working, and having coverage through the state, to working but not having coverage yet in her new workplace. So, while this transition could theoretically be seamle ss, people can get caught in the gap and not have coverage for a while. Finally, one person lost coverage because she did not send in her renewal forms in time: And it got my application inside to r enew it, and Im sending that off today, they just got, as long as they get it by the end of the month and Ill be cool, so Ill be able to afford it This patient experienced a gap in cove rage that caused t he prices of her prescriptions to become much higher: Its like 2.50 with it (insurance), and wit hout it its like 40 some odd dollars. In all of these situations, people experienced a gap in coverage that they knew was only going to be temporary, but still impacted their ability to purchase their prescriptions. Simply having some ki nd of coverage is not enough for these cases where a gap in coverage can occu r for a variety of reasons. Limit on number of prescriptions One persons third-party payer imposed a limit on the number of prescr iptions that could be covered at one time, leading any additional prescriptions to be acqui red only by paying market rates: 70

PAGE 71

my insurance company down there onl y allows you 5 prescriptions per month, and I had already had to get a c ouple of others, and thats when it happened and I couldnt afford it He felt that he needed this prescription, and even though he had insurance, this prescription (over the company-i mposed limit) was not covered. Insurance company delaying payment/coverage One persons insurance company stated that they would not pay for a prescription until a later date, leading to a gap in coverage: Well, this morning I went in and my insurance company decided well, were not gonna get, were not gonna pay for a prescription till the 14th(7 days away) so if I wanted it I had to pay 24 dollars, and I couldnt, I couldnt pay the 24 dollars. I get all my prescriptions at the same time, but the insurance company doesnt want to pay fo r them all at once. So, they just started breaking it down where you can t get it yet cause we dont want to pay for it yet. So, even with coverage for the prescription, this person still experienced a gap in coverage because of the reimbursement policy of the insu rance company. Third-party payer stopping coverage One person had their coverage stopped for no reason that she could perceive: Then last month, I got a letter from t hem telling me that they discontinue my medical because I make too much m oney. I work part-time, I dont work full-time. I do care-giving job. I make $241 every two weeks. It was not clear whether or not she was work ing prior to this, or if the third-party payer was a private insurance company or government run, but regardless, she had some kind of coverage which stopped unexpectedly. Another person had their coverage stop becaus e of an administrative mistake, but at the time he did not know it was just going to be a temporary gap in coverage: Well, the insurance actually dropped on me one time. They actually closed up and said no, and cut me off, and, because it lapsed, and I couldnt figure out why it lapsed, and it had everything to do with the mail. The mail didnt 71

PAGE 72

get me the paper in time. When I got the paper it was already two weeks late, so the insurance companys I guess he aint doing this anymore. They dropped me, so I had to go through, it took me three months to get my insurance back. This person then went through the three month stretch of time where he had problems obtaining prescriptions because of lack of coverage. These examples seem to indicate that going from having some sort of third-party coverage to not having any coverage can happen quickly, often with the person not knowing ab out it until they try to obtain their next prescription. Other / responsible third-party payer delaying payment One person experienced an accident (tile falling on his head) that was not his fault, and was told that the buildings insurance would cover his prescriptions: (I have) No medical insurance right now, but they send me a form, I call them, I dont get anything, anybody there to afford the bills. I was under the building, they sent me that they cover the medication, but when I call them, I dont get anybody. I just got a lawyer to represent me now. So even though this person technically has insurance coverage through the building liability insurance, it does hi m no good in terms of having OOP costs for prescriptions being reduced because of his coverage. Cost still too high with insurance One person had private insurance, and was able to enjoy the benefit of the OOP cost being reduced bec ause of the insurance, but the cost was still too high: Medicines that I take, theyre a littl e bit expensive. Its (the cost compared to previous months) about t he same. But its just that everything is going up, and up, and up, and up, and it gets to the point where you cant afford it anymore. 72

PAGE 73

She was able to get the benefit of the rela tively lower prescription cost due to having private insurance, but with other cost s, the lower price was still more than she was willing to pay. In summary, the OOP cost that people were presented with when trying to acquire a prescription was heavily depe ndent on whether or not they had some kind of thirdparty coverage to help bring the cost down. The majority did not have coverage, leading to higher out of pocke t costs. Having some kind of coverage, however, did not necessarily lead to easier acquisition of t he prescription, because of various reasons that led to the coverage not helping to bring down the OOP cost. Willingness to pay Once the OOP cost of the prescription was established, each individual seemed to consider how much they would be willing to pay for it. In the initial interviews, people were asked how much they were willing to pay for the prescription. Some of the respondents either did not understand the questi on, or were having a hard time coming up with an answer, so they were prompted with a question such as would you have paid 5 (or 10, or 20) dollars for it? after which they responded. After a certain number of interviews, it became clear that ever yone had some price at which they would be willing to pay for the medication. This price, however, was less than the OOP cost, leading to them initially not acquire the prescription. The respondents who were explicitly asked how much they were willing to pay had some of the following responses: $50 or less. I think probably maybe $6 is w hat I had to pay (for it). $20 would probably have been my limit. 73

PAGE 74

Probably $20. For those who were asked if they would be willing to purchase t he prescription at a lower price (usually $5 or $10 was given as the lower price), the following responses were given: Of course, of course. Oh yeah I would! (laughte r) Of course I would. If I had the money, yeah, of course. Oh yeah. Yeah, I probably could have afforded it, yeah. Yes, yes, of course, yes, I am sure I will pay that. One person even talked about how his situation was better now that he had a better job with a higher income: Now I have a better job anyway. I can get it. If it was like 200 I could get it. You dont want to feel the pain every day anyways. These responses indicate that all were willing to pay a certain amount for their prescriptions, but were not willing to pay any more at this particular time. Many people responded that they were not willing to pay mo re for the prescription because they had too many other expenses: And sometimes I just dont have the money, so I dont get it because it costs too much. Well, it was just a cost-benefit analysis and prioritizing what I need. Like surviving so I have to come back and get some help here, or go and ask some medmobile, because right now I am in a shel ter, I cant even afford my rent until I get a full time job. bills are keep on coming and coming and coming, Im about to be homeless, you know. I dont have m oney to pay for my bills now The prices, everything, is going up, but my check doesnt go up, you know, it stays the same, so I have to make wa ys to pay all the bills to try to make 74

PAGE 75

a living. But its ju st that everything is go ing up, and up, and up, and up, and it gets to the point where you cant afford it anymore. Given the sample with their relatively lo w household incomes (a s reported earlier), it is not surprising that they are not willi ng to spend more money on their prescriptions with all of the other items that they have to pay for. To check if any other factors besides ju st the cost had influenced their decisionmaking process about purchasing the prescription, they were asked directly what other factors were involved: No, it was just the cost. The cost. Thats (cost is) the sole reason. No, it was just the cost. No, I didnt buy it because it was too much for me. Just the cost. It might seem obvious that if one is asked about a time that they did not acquire a prescription because of cost, that a pers on would answer that cost is the only reason, as shown from the quotes above. One of t he research questions was to determine if other factors were involved besides just co st (with people still r eporting that it was because of cost), but for the ones who were not offered help or who knew about where to go to get help, it seems that there were no other factors involved. The next section will discuss other factors that were asked about that were shown not to have an impact on whether or not someone pur chased their prescription. Factors not Shown to Imp act Prescription Acquisition The factors above that impac ted whether or not som eone acquired a prescription centered mostly on monetary issues, specifica lly the price of the prescription and the 75

PAGE 76

amount that people were willing to pay. Stud y participants were asked about various other factors, which had been thought to possi bly have an effect on whether or not a person acquires a prescription. As shown below, none of these other factors seemed to have an impact on this acquisition. General beliefs about medication All respondents were asked about their general belief in medications, mainly because the adherence/compliance literature s uggested that some people did not take their medication (which had already been ac quired) because of their beliefs about medication. It was assumed that if people did not want to take their medication based on their beliefs, then these same people would probably not want to initially buy the medication, yet still report that they had not pur chased the prescription because of cost. However, for this population, everyone r eported that they were interested in purchasing the medication (as described above), if the cost was afford able to them. Consequently, when asked about their general beliefs about medication, many people were fairly positive, as the following examples demonstrate: I mean, Im a medication person, I mean, Im not like my mom who dont believe in medicine Medications are helping because Im ta king, right now Im taking 14 medications. I like my medication, it keeps me, it keeps me going, it really does. What was interesting, however, was that many people professed not to like taking medications, but did anyway because they had been found to be effective in the past. Therefore, they recognized that even if they did not like taking medication, the medication they had taken worked and they wanted to keep on taking these 76

PAGE 77

medications to help with their medical conditions, as the following examples demonstrate: I dont like medicine, but if I have to, I have to. I take medicine as a last I try everything else. Yeah, even if I have a headache, I wait. I dont like just go and take, I just wait until other stuff. Then I go and take. I dont like to take it but I aint got no other choice. Y ou either take it, live or die, or back and fo rth to the hospital. No, I dont like medicine in general! No No, I just have to. Oh, the (blood) pressure (medication) works. I have to. Ive been taking three particular medications for, wow, a long time, since early, late nineties ever y day, Im taking meds every day. Pretty much, you know, the only thing I think its just a shame that a man has to take that much medication to stay alive. I mean, if I didnt have to have it, it would be nice, but, you know, I just dont see me without it. No, I dont like the medicine at all, but if its going to make me feel better, of course Im going to drink it. Im not a medicine person. Id rather get better These statements indicate that alt hough they might have a generally negative belief about medication in general, they are willing to take their medication because it has shown to provide a benefit. The followi ng examples indicate how some of the respondents felt about the effectiveness of their medication: And it (blood pressure) still aint down to where its supposed to be but its down very good compared to what it was. Thats a result of taking the medicine! Oh it works. Y ou know, I mean the medicine works. They worked good, I feel better. I c an breathe more better, blow my nose more cause at first, like, your nose is like real stuffy, you cant smell nothing, somebody cooking food you can t smell anything, but you use that on the occasions, it cleans your sinuses out real good, so Flonase works really good. If it wor ks, youre going to use it. Ive been on it for years, so I guess it works well because, everything else would be detrimental 77

PAGE 78

It works well, as long as you take it when you are supposed to take it. For these individuals, their previous ex periences with the m edication had led them to continue to use it. Another interesting aspect of peoples beliefs toward medication effectiveness was that people generally believed that, for new conditions, the medication would be effective. This is not surprising, giv en that this population had knowledge of their medical condition and sought out help for it. It is logical that someone who follows these steps would continue to believe that whatever therapeutic plan the physician came up with (usually a prescription) would be effective. This attitude is shown in the following examples: Yeah, it will be much, much better, because the doctor told me, I have headache, this can take care of all of my pain arm pa in, and everything, so, but I cannot afford it. I will listen to the doctor first and try to fo llow them, but if its like, its not working, I will take it out (and try) something else, you know. This last quote sums up their attitudes on effectiveness. They believe their medication will be effective, and continue taki ng it as long as it is perceived to b effective. If it stops being effective, howev er, they indicated that they will probably stop taking it. Other utilization-related factors There were other factors that have been sh own to be related to adherence (Pound et al. 2005) that were not specifically in quired about, including the following (taken from the list presented in Chapter 1): They weigh the benefits vs. t he adverse effects of the drug Taking the medicine does not fit in their daily routine They are nave scientists they try it or stop it, and see what works best A negative self-identity co mes from taking the drug 78

PAGE 79

They are worried about addiction Surprisingly, there was no mention of any of these factors when people were prompted with open-ended questions regarding what happened when they did not acquire a prescription because of cost. As obvious as it sounds, as noted above, the decision did seem to be just about the cost. In summary, it seemed as though peoples general belief about medicine, peoples concerns about the effectiveness of the medicine, and other utilization-related factors such as potential medication side effects or worries about addiction to the medication did not seem to influence whethe r or not someone acquired a prescription. As detailed above, the prescription OOP cost and what people were willing to pay were the main factors involved. Response to Not Acquiring Prescription The previous section dealt with the factors involved in the people not initially acquiring their prescription. Of course, the need for acquiring a prescription did not go away; there was still some kind of medical condi tion that had to be dealt with. This next section examines the factors that impacted whether or not someone actively sought out getting help in acquiring their medication. T here were two major factors that influenced whether or not someone sought out help: their knowledge / beliefs about what help is available, and their perceived le vel of need for the prescription. Knowledge / Belief that Help is Available There were two people who expressed t he belief that there was no place they could go for assistance for acquiring their pr escription, so they did not even bother looking. One woman who was experiencing a gap in coverage was asked if she did anything to try to reduce the price of the prescription of her blood pressure medicine: 79

PAGE 80

What was there to do? Consequently, she did not pursue any ki nd of assistance, and waited for her coverage gap to close. In another case, a woman was pregnant and had problemswith blood pressure that could ultimately affect the fetus. She was advised by her doctor to get a prescription, but she did not have insuranc e and thus the OOP cost was too much for her to pay. When asked if she did anythi ng to try to acquire the medication, she responded: No, they dont do things like that. They dont. Trust me Id be looking, they dont do things like that. Even in this case where the stakes are high, she believed that no assistance existed for her, and consequently she did not look for or receive any. Instead she modified her lifestyle, and ulti mately had a healthy baby boy. The above examples illustrate how pre-ex isting beliefs affected whether or not someone sought out assistance for acquiring thei r initially too-expensive prescriptions. Those who believed that no assistance was available, independent of perceived need, did not seek out assistance, and ultimately had to use an alternative or live without. Perceived Level of Need fo r the Prescribed Medication For those who believed that assistance was available, the key factor that determined what they did in response to not acquiring a prescription because of cost was how much of a perceived need they had fo r their prescription. Those who had a higher perceived need tended to seek out assi stance or additional resources so they could acquire their prescrip tion. Those who had a lower perceived need tended to not 80

PAGE 81

seek assistance, and consequently used an alternative and/or lived without the prescription. A persons perceived need depended primarily on three factors: 1. The perceived consequences of not taking the prescription 2. The time period for which they desired the prescription 3. The perceived availability of alternatives Each of these factors will be examined in detail below. Perceived consequences of not taking the prescription People were asked directly about how severe they believed the consequences would be if they did not ta ke their medication. Giv en their previous and current experiences with their medical condition, most had an idea of how severe the consequences would be if they did not take their medication. Si mply put, the more severe the consequences, the greater t he perceived need for the prescription. Many people echoed a sentim ent that using their medication was absolutely essential to their improved health and quality of life, as the following examples show: I probably would have lost my ey e sight! Sure! (Glaucoma) Its gonna be bad. Yeah, if I st opped my medication. (Hair / skin) If I didnt take them I would be miserable. Id probably just be in the house, dont want to come outside and play, just dont want to be bothered. (Allergies) Im bipolar, schizophrenic, so, you know, I need these medicines. On one hand, these attitudes are positive to ward medication, reflecting that they recognize and value the use of medication to improve their medical conditions. On the other hand, it is difficult to imagine what it would be like to feel that you needed your medication, but could not afford it because it cost too much. Having these feelings about severe consequences from not taki ng the medication had a big impact on their perceived need for the medication. 81

PAGE 82

Several of the study participants had high blood pressure, and explained the severity of the disease as follows: But I couldnt go off my blood pressure medicine cause I know I would die, so I had to bite the bullet and pay for that one. I wouldnt be here, seriously. Blood pressure medication is nothing to play with you know, stroke, heart attack, other things you can develop. But, see the thing is with high blood pre ssure a lot of peopl e dont dont want to take their medicine. There are a whole lot of times, you know, I guess this medicine aint doin me righ t, but if you get high blood pressure, youre gonna have to carefully check with their doctor, to see if the medicine is working. If that medicine aint wo rking, then they put you on some other kind of blood pressure medicine, to see if thats working. It aint, it aint goin to happen in one day, you know, ju st takin one-time medicine, you know, my time I took, to get my pr essure down, I dont know how many medications I took for my high blood pressure. The five people who had chronic hypertension not only seemed to believe that it was vital for them to take m edication for survival, but also recognized the importance of regular monitoring by a physician (as echoed by the third quote above) to modify their regimens for optimal results. Given the population that was sampled (those who know about their condition and seek out care), this result is probably not surprising. These attitudes about the consequences of not taking high blood pressure medication were a major factor in their percei ved need for the medication. How do people form these opinions about disease severity and consequences of not taking medication? While this question could not be answered from the interviews, it was interesting to hear about the messages that people heard from their doctors when getting their medication: If a doctor tells me I need it, I need it, to me. Theyre smarter than me, and they went to school for th is, so, I listen to doctors. Because she said thats more im portant (choosing blood pressure medication over cholesterol). 82

PAGE 83

You have to because we need it, the doctor says so, we have to buy it, even expensive In each of these cases, the doctor told patients that they need a particular medication, or need this one instead of that one. These messages, along with experience from having the medical condition, probably contributed to a persons perception of the consequences of not taking the medication. In summary, people had beliefs on how severe the consequences would be if they did not take their medication. Those who perceived that severity was high led to a higher perceived need for the medication. The time period for which they desired the prescription Knowing that one will be without medication for a specific period of time as opposed an indefinite period of time affected the perceived need the people felt to acquire the medication. Generally, knowi ng that one will only need a prescription for a short period of time reduced the perceived need for accessing help or acquiring the medication. The most common scenario in which this was applicable was for when there was a known gap in third-party payer coverage for obta ining ones prescription. In many of the cases mentioned earlier, the people knew th at they would be without coverage for a known period of time, usually on the order of a month or less. On the other hand, there were examples of other people who had t heir coverage dropped for an indeterminate amount of time, such as the person w ho had her coverage stopped suddenly and needed her high blood pressure medicine: I dont know how long it will take, so I have to come back and get some help here, or go and ask some med-mobile 83

PAGE 84

This person, since she does not know how long it is going to be before her coverage will resume, has a higher perceived need for the medication than someone who knows that their medica tion will become accessible agai n in a known period of time. In this case, the woman with hi gh blood pressure who needs medicine for an unknown time period has a higher perce ived need for her medication than other individuals with high blood pressure who know that they will only be without coverage for a known, short period of time. Another way the time period came into account was for people who had a medical condition that typically resolves in a short period of time, such as an infection that is likely to be viral. It makes some sense that if one belie ves that a condition will last only a certain period of time, this might affe ct how much one perceives the need for the medication. There were several people w ho reported having an infection and desiring antibiotics. One example follows: I come really sick, I go t the flu, Im coughing, I have throat infection and everything so I just go in the hospital. So, they prescribe some medicine for me, I go in the pharmacy, I cant buy it, so expensive, so I dont buy the medicine. I go back home, take a tea, this, and I dont feel good. A week later, I go back in the hospital, they give me a lot of medicine, and I cant buy also, so I just, so sick one is $79. They give me three medicines, one is 52 and the other one is 47. I reme mber, I never for got that. When I go back a week later, medicine come in more expensive, because I come in more sick, so they change it, diffe rent medicines. This one I dont remember, but it costs more so I dont buy it also. This person was never explicitly asked whet her or not she felt that her condition (the flu, a throat infection ) was one that would be expected to resolve itself over a short period of time. In fact, there were a few people with perceived severe infections who did seek out help to purchase their medication. This case (and one other who had an infection) were the only two in which the p eople felt that their conditions were severe, 84

PAGE 85

yet did not have a high enough of a perceived need to seek out help. The one factor that these have in common is th at they were for probably vi ral infections (which would be expected to resolve itself in a short period of time), as opposed to a chronic disease such as high blood pressure or diabetes. Consequently, it seems as though the time period of need for the medication could have an effect on the perceived need for the medication in situations where the medical condition is expected to resolve itself in a short period of time. In summary, the time period for which the prescription is desired seems to have an effect on a persons perceived need for the m edication. The longer the time period for which one desires the prescription, the higher ones perceived need. The perceived availability of alternatives The final major factor that affected ones perceived need for the medication was what alternatives were available for people to use instead of their prescribed medication. Sometimes these alternatives were pharmacologic in nature, simply a cheaper drug that could hopefully solve the m edical problem to some extent. Other times, people tried to use some kind of natur al medicine or lifestyle behavior change to achieve the desired effect. The perceived presence, or lack thereof, of these alternatives in any case had an effe ct on their decision-making process. A common alternative that was used wa s to administer an over-the-counter medication to try to treat an acute condition, such as pain. In one example, a brick fell on a person, who had raised his arm up to block his head, and suffered a gash on his arm. He was able to get stit ches, but nothing for the pain: maybe like 150 or 200 dollars, the pain medication, the pain killer medication, so I cannot afford it. y eah, I did afford to get Motrin 800 milligrams, yeah, but t he expensive one, I didnt afford to get it. 85

PAGE 86

In another example, a young wo man is experiencing stomach pain, and is told that the medication is going to be over $100. She ended up buying an OTC antacid alternative and currently uses it, to moderate e ffect. When asked if sh e is able to afford medication for other conditions, she replies: I mostly use Advil. (laughter) Thats the best one, thats the main thing in our family, Advil, whatever pain you ve got, Advil it works a lot. So, for acute conditions such as these, havin g alternatives available, even if they are viewed as suboptimal, affects the decision that they will make regarding whether or not to purchase the medication. Using an alternative drug for a chronic condition was only mentioned once by the people interviewed. A person was going to be without his high blood pressure medication for a short period of time, so he opted to use water pills which he already had acquired to help tide him over. Using water pills by themselves as opposed to with his other anti-hypertensive medication woul d probably not be as effective, but he believed it would still be better th an not taking anything. Obviously, there are many different types of non-pharmacologic alternatives that are available for people. The most common th ing that people did was to try to modify their lifestyle in place of taking the needed drugs. Some examples are: so I just tried to de-stre ss my own self by exerci sing more, and trying to relax. But besides that, there was nothing I could do. (Being treated for depression) Yeah, mostly just try to eat healthy, drink water, lots of water, walk a lot, you know, try to stay calm, not to raise it up a little bit, try to stay calm, drink a lot of water, eat a lot of fruits and v egetables. Mostly, just simple things, eating healthy, walking. (High blood pressure) In all of these cases, people just tried to do something that they felt would improve their health in general, that did not cost as much as the unaffordable medication. 86

PAGE 87

Some people had extra medication availabl e from filling a previous prescription, and tried to stretch out whatever supply they had, figuring t hat taking part of a prescription was better than not taking anything. but I try, instead of taki ng it every day, I try to take it like once a week. (Allergy medicine) So I had to go off of it, and I knew that I would have to go off of it, so I tried to like medicate myself by slowly taking half the dose I was supposed to take and I was off that medication fo r a pretty long ti me. (Depression) These two had some leftover supply, so they spaced it out as much as they could when they realized that they would not be abl e to acquire additional medication. Still, knowing that one has some extra in reserv e could affect ones perceived need for the medication. One person mentioned borrowing medication fr om her sister in order to help her condition: so I had some of my sisters medicine, which was different, so I started to get some of hers (High blood pressure) It is unclear how much this would help, espec ially since it was a different medicine. Regardless, she seemed to believe t hat it was better than doing nothing. Some people also mentioned natural remedies or foods that could potentially help their medical condition One example was: from our country we have some herb or some stuff like that we use that a lot, and then here, they dont have some thing like that (General medical condition) These people seemed to realize that these remedies were not likely to be very effective, but still it may be better than not doing anything. In summary, a variety of alternatives we re perceived to be available to peoples prescriptions, with varying degrees of potent ial effectiveness. Accordingly, knowledge 87

PAGE 88

of these alternatives seemed to have an e ffect on the perceived need for getting the prescription. If one feels that an alternative can be moderatel y effective for a fraction of the price of the prescription, this could decrease the perceived need. Case studies demonstrating perceived need A persons perceived need for their prescr iption therefore was based on the above three factors, perceived severity of the c onsequences of not taking the medication, the time period for which they desired the prescr iption, and the available alternatives. To help illustrate how these three factors combined together, situations were examined in which a choice had to be made between at l east two different medications and the person could not initially afford either one. Case study #1 High blood pressure vs. anxiety This person experienced a gap in coverage, and had to prioritize between getting her high blood pressure medicine and her anxiety medicine. I mean it, I would rather be on it than not be on it ( anxiety medication), but I know the difference of, I could really hurt my heart if I go off my blood pressure medicine, while I may be i rritable and cranky to people but they can just deal with it. It seems as though her perceived consequenc es of not taking her blood pressure medication were more severe than her per ceived consequences of not taking her anxiety medication. In addition, it is import ant to note that she was very confident that her insurance coverage would resume in a known, short period of time: Well, sure, but the insurance company is going to take care of it. So, Im not really worried about it. Thus, she knows that she only has to worry about a short period of time to be without her desired prescriptions. She also k nows that there are alternatives for her 88

PAGE 89

existing anxiety medication, mentioning stretching it out and modifying her lifestyle. Therefore, she decides: But I couldnt go off my blood pressure medicine cause I know I would die, so I had to bite the bullet and pay for that one. In this case, the two most important factors that cause her perceived need of taking her blood pressure medication to be hi gher than that of her anxiety medication were the relative severity of the consequenc es of not taking the medication, and the availability of alternatives for her anxiety medication. Case study #2 High blood pressure vs. pain This person experienced a loss of coverage for going over the amount of co vered prescriptions in a one-month time frame. As a result, he had to make a decision regarding whether to acquire his medication for high blood pressure or fo r pain for the one-month period. As he described it: Yeah, I had to choose my pain medica tion over my blood pressure medication for like 15 days becau se of the 5 prescriptions He did not really discuss the consequences of not taking one m edication over not taking the other one, because t he other two factor s, the time period of need, and the perceived availability of alter natives were so dominant. In the end, his perceived need for the pain medication was higher than that of his high blood pressure medication, so he chose to acquire and take the pain medica tion and use alternatives for his high blood pressure: Yeah, yeah, actually I had a couple of extra blood pressure pills and some water pills, so that would have helped keep it down some, so I didnt have to be really pressed about it So, in this case, consequences of not taking the medication were not really considered, given the relative impo rtance of the two other factors. 89

PAGE 90

Case study #3 High blood pressure vs. cholesterol This is an interesting situation because it compares two chr onic conditions. This person had just unexpectedly lost her insurance coverage, and she did not know when she would be able to get coverage again. Regarding the consequences of not taking her high blood pressure medication, she said the following: Yes, it (blood pressure medication) will stop my headache. It will help me to go to work. Yeah, it will affect my job, because I feel the headache. I feel the heaviness, I feel the sometimes dizziness Regarding the availability of alternatives, she states: Cholesterol I can cut down on food but I cant cut down on my medication In addition, she stated the following regar ding the consequences of not taking her cholesterol medication: Well, if I get it OK. If not, its optional. Based on the relative consequences of not taking the medication, and the available alternatives, her perceived need for the blood pressure medication was higher than her perceived need for her cholesterol medication. These three case studies were shown to give one an idea of how, in situations where a person had to choose between two diffe rent medications, they calculated the perceived need of each one. In these cases, it was the combination of at least two of the three identified factors contributed to th is perceived need; note that not all three factors have to be involved for all decisions. It just shows that people take into account more than one factor when determining t he perceived need of a medication. People Seeking Help or Additional Resources Those who had a high enough perceived need actively went and sought out help. The most common form of help that was sought out was getting money from friends, 90

PAGE 91

family members, or their local church. Other people sought out prog rams that would get them coverage, and one person who asked for help was told where to go to get cheaper medications. Each of these exampl es will be illustrated in detail below. Money from Friends/Family Several people went and asked friends and family members for monetary assistance to help pay for their prescrip tions. One person was trying to get a prescription for a special shampoo: But I like, put money together, I try my sister and go get it. With help from her sister, she was able to acquire the shampoo. Another example is from someone who had high blood pressu re, and needed additional money to pay for his medication: Well, I borrowed money to get it, till I got some insurance, I got a dollar here, 2 bucks here, you know what Im sa ying. Borrowed till I get what I want. I still got it, but hey, it was costly. I had to scrounge up money to get my medicine. As he states, this lasted for a while until he was able to get some third-party payer coverage. Another person got unexpect edly dropped by his insurance company, and had to scramble to get the medicati on that he perceived that he needed: I borrowed money from friends tryi ng to get my medicine, my pain medicines $119 a bottle. Yeah, I, if it wasnt for my friends I would have, I would have, I would have sank like a big ship. In all of these examples, money from fri ends or family was able to help them get their medications. Money from Church Some people were able to go to their lo cal Church to get assistance. In one example, a man was newly diagnosed with bronchitis, and was prescribed an inhaler. 91

PAGE 92

He had insurance, but he had already used up hi s allotted five prescriptions per month, so he decided to ask his Church for help: It cost like 240 bucks, it was an inhaler, and I had to wait a few days. Finally, I went to my church and they helped me get it. In another example, a woman needed to get her anti-anxiety medicine. Although it was relatively cheap, she still did not have any money to pay for it, so she went to her Church: Yeah, but actually I still had to go to my parish and ask them to help me get it, so, you know. My church helped me get it. They gave me a 20 dollar card, plus they paid me cash, and gav e me like 10 dollars, you know I mean, I mean Ive had people come, y ou know, here and there to help me out a little bit In both of these cases, the people active ly sought out help for purchasing their prescriptions, and they were able to get it from their respective Churches. Location of Where to Go One person who was looking for help to pay for her prescriptions was told by a clinic worker where to go to get cheapest medication: when I ask them which like pharmacy is cheaper, they tell me Walmart, and I go to Walmart to get it. Yeah, the last time it was fine, because it was costin me, I think it was 5 or 10 dollars. While the health care professional was not able to directly assist this person, at least they were able to point this person somewhere else where they could get help. Working More to Get Extra Cash This section is titled seeking help or additional resources, and this was one example of someone getting additional resources on their own. One person felt a high perceived need for getting medication for his child ren, so he worked extra hours in order to obtain it: 92

PAGE 93

And then I had 6 children, and 5 of them are grown, but I can remember times, I mean things I had to do to get t heir prescriptions filled. It was just ridiculous I had to work extra hours every day for a week sometimes, yeah, when they got sick. So, while this might not be thought of as s eeking help in the traditional sense, he did do something pro-active to be able to purc hase a prescription that he felt he had a high need for. Looking for a Program / General Assistance Another way to seek help for obtaining a prescription (especially for a chronic condition) is to try and get into a program that will provide co verage for the long-term. This example is from the man above w ho had to scrounge up money to pay for prescriptions, and the doctors that he was goi ng to were not providing any additional assistance. Then he came to this clinic wh ere the doctors helped with his therapy and got him enrolled in a program so t hat he could acquire his medications. Ive only had two doctors, Dr. X and Dr. Y. These were the only two that got my pressure down to a point, well I thank God for Dr. Y and Dr. X, because if it wasnt for them, Dr. Y, if it wasnt for them I probably wouldnt be here today, you know, like they was patient, giving me, you know, switching different medicines to fi nd out which one is working right, and these medicines there, theyre working right. Yeah, I can afford them (now), I can get them. There is a special program (at the clinic), yeah, so I can get them. Once he got into the program, he wa s able to acquire his prescriptions. Adequate Provision of Help In all of these examples, the perceived need was high enough for people to actively seek out some kind of assistance, or to gather additional resources to try to be able to acquire the desired prescription. A ll but one of the people who actively sought out help were able to eventually acquire thei r prescription. The one person who was not able to was the man who had his insuranc e coverage stopped unexpectedly. He was 93

PAGE 94

able to get money from family and friends to get some of his medications, but not all of them. I had to do without my anger medicine cause its like a thousand dollars (laughter) so I couldnt afford that, I mean, I was doing everything I could possibly do to try to stay alive (laughter). Summary In summary, this section concerned itse lf with the type of help or additional resources that people sought out. Most of the help that was sought was purely monetary, of scroung ing up enough money to be able to purchase the prescription. There was also seeking out a program fo r long-term help, working extra to gather additional resources, and asking someone where to go to get a cheaper prescription. Only one person who sought out help did not re ceive it, because the amount of help that was needed was simply too much. Access Result As reported earlier, some of the people who initially reported CRN actually ended up with the desired prescription. In order to get the prescription, they all had to have some kind of help, which they either knew about, sought after, or was offered to them. For the people who did not acquire their pr escription, they either had to use an alternative or live without it. The alternatives that were used have been previously discussed, but are summarized below: Using an OTC medication Using an alternative drug for a chronic condition Using a similar drug borrowed from someone else Stretching out the existing supply of medication Using a natural medicine / remedy Modifying ones lifestyle 94

PAGE 95

There were several who, by not seeking he lp, ultimately chose to live without the prescription or any of the abov e-mentioned alternatives. Most of the disease states in which people lived without the prescription we re those that either resolved themselves in a relatively short period of time, or were not that severe, both of these factors leading to a lower perceived need by the individual. And sometimes I just dont have the money, so I dont get it because it costs too much. I dont even have a credit card yet, I only have debit, so I just suffer the consequences. I just had to suffer through it. Well, I went to school one day and I was playing gym and, cause I have asthma so, I needed an inhaler and I didn t have money to purchase it, so I went to CVS and they told me how much I had to pay for it, and thats when I didnt have enough to get it, so I just dealt with it. I couldnt afford it and I just couldnt take it, so I didnt ta ke it. And I just, I guess I dont know what happened. Everything went just fine after that but, I think I should have took it. I think I should have. In all of these cases, it seemed that these individuals would rather have taken their medication than not take it, even if the c onsequences they had to suffer through were not as severe as with other conditions. Summary In summary, there were two main factor s that influenced whet her or not a person initially acquired a desired medication. 1. Help factors either help was provided by someone else, or a person was willing to access known help 2. Monetary factors a per sons willingness to pay was greater than the OOP cost of the prescription 95

PAGE 96

Some of the people in this study were o ffered help or were willing to access known help, but most were not willing to pay the OOP cost of the prescription, so they then had to determine if they were going to seek out help. This depended on two factors: 1. Knowledge / belief that help is available 2. The perceived level of need of the prescription, which in turn depended on three factors: a. Disease severity or severity of co nsequences of not taking the medication b. Time period of need c. Available alternatives Those with the belief that hel p did not exist did not seek out help to pay for their prescription, independent of their perceived need. Those with a higher perceived need sought out help, while those with a lowe r perceived need did not seek out help. All but one of the people that sought out help eventually acquired their prescription, meaning that people who event ually acquired their prescriptions ended up falling into one of three categories: 1. Those who were provided help by someone else 2. Those who were willing to access known sources of help 3. Those whose perceived need of the pre scription was high enough such that they sought out help or additional re sources, and who received enough help People who did not eventually acquire their prescription also fell into one of three categories: 1. Those who did not believe that help existed 2. Those whose perceived need of the pre scription was not high enough such that they sought out help 3. Those whose perceived need of the pre scription was high enough such that they sought out help or additional resources, but did not receive enough help This new framework was intended to illustra te all of the factor s mentioned above. 96

PAGE 97

97 Perceived need for the prescribed medication Perceived consequences of not taking Rx Time period of need Perceived availability of alternatives Knowledge / belief that help is available Response to not acquiring prescription Help factors Patient willingness to access known sources of help Other persons provision of help to passive patient Out of pocket prescription cost Third party coverage or not Issues with existing coverage Willingness to pay Resources available Competing needs Monetary factors Attempt to acquire desired prescription YES; If help is provided or accessed, or willingness to pay > out of pocket Rx cost Rx acquired? Rx is acquired NO; If help is not provided or not accessed, and willingness to pay < out of pocket Rx cost Actively seeks help or additional resources? Patient uses alternatives or lives without Is adequate help provided? Population Inclusion criteria Have reported not acquiring a prescription because of cost in last year Have purchased a prescription in last 5 years 18 < Age < 65 General characteristics Have knowledge of a medical condition Have sought or are seeking treatment for the condition at this clinic (serving the under insured) Generally have the belief that taking medication will alleviate their condition NO NO YES YES Figure 4-1. Framework showing factor s affecting medication acquisition

PAGE 98

CHAPTER 5 DISCUSSION This study was undertaken to better understand the circumstances surrounding a person not getting a prescription filled becaus e of cost and to determine what other factors besides the cost came into consider ation. While national-level survey research had identified various demographic characteristics and behaviors of those who reported CRN, they did not capture which factors we re the most important for those making these decisions. As the new framework illust rated in the previous chapter, factors were identified that affected the in itial reporting of CRN, and additio nal factors were identified that affected the decisions that people m ade afterward in the hope of acquiring the desired medication. Sample One important caveat on the results is that the convenience sample was from a population that was different from the general population by key features, all of which would likely have an effect on the results. Firs t of all, this population knew that they had some kind of medical condition. While this may seem obvious to most, consider all of the public health campaigns that exist sole ly to convince people to get screened to see if they have a condition. The population for this study, however, had knowledge of whatever condition they had when they r eported CRN. This k nowledge in itself separated the population of this study from the at large population. Secondly, with the knowledge of having a medical condition, this population actively sought out treatment of some kind, which does not necessarily happen with the population at large. Simila r to what was mentioned above, another goal of many public health campaigns and/or health education in terventions is simply to get people to 98

PAGE 99

actively seek out treatment for a known medi cal condition. The popul ation for this study obviously sought out treatment for their condi tion, or else they could not have ended up in a situation where they would have to r eport CRN. In addition, this population sought out treatment at a clinic that catered to the under-insured; diffe rent experiences may have been described if recruitment took place at a clinic that served a population with a higher income and/or bette r insurance coverage. Finally, this population had a belief that medication could be used to improve whatever condition they had. For thos e beginning a new therapy when presented with a prescription (on which they would late r report CRN), none of them immediately rejected the idea that the medication could help solve their problem. For others trying to continue using a certain medication, they perceived that taking their medications had been effective with previous usage, so they wanted to continue to take them. This differentiates them from the at large population, as there are many who reportedly discount medication outright based on cultural or other beliefs. Therefore, it seems as though my sample is not repr esentative of the population at large, which is the population commonly studied when doing na tional-level survey research. This clarification is essential, because given the non-r epresentativeness of qualitative research in general, one of the most important factor s is to explicitly define the sample population from which you recruit, so that the results can be translated accordingly, and that other, possibly confi rmatory, research can be carried out with a similar population. Dynamic Nature of CRN All of the people in this sample reported CRN, yet some people eventually were able to acquire the medications. While this s eems contradictory, it is still true in the 99

PAGE 100

sense that there was an initia l attempt made to acquire t he medication which they were not able to do because of cost (leading them to report CRN), but then some were able to seek out and receive help, ena bling them to acquire their m edication. Similarly, some people reported an episode where they were not able to acquire the medication, but then learned where to go to receive help, so t hat the next time they needed to acquire a prescription, they knew where to go and could acquire it. The point of these examples is to show that whether or no t someone reports CRN can change over time, and that asking a di chotomous question about whether or not CRN has been reported within a certain time period may not be adequately representing what people are going through in terms of accessing medication. A person reporting CRN could: 1. Not have any more access problems because their coverage gap has ended 2. Have borrowed money to acquire th is prescription, but be uncertain about acquiring ones in the future 3. Have learned about a clinic to rece ive help, so will probably not have access problems in the future 4. Have not acquired this prescription, and because they do not believe that help is available, will probabl y not be able to acquire prescriptions in the future. Each of these alternatives occurred in the 21 people that were interviewed, and each represents a different kind of access probl em that needs to be solved. While they all reported CRN, it does not seem useful or hel pful to simply consider that they belong in the same bucket (those who have reported CRN), since their experiences and needs are so varied. 100

PAGE 101

Comparison with Other Frameworks As a starting-off point for this study, tw o different frameworks were considered that would be applicable to trying to describe this process, the framework of factors influencing rates of cost-re lated medication underuse a nd other problems due to medication costs by Piette et al. (2006a), and the Behavioral Model of Health Services Use by Andersen & Aday (Andersen & Davidson, 2007). Bodies of literature in two other areas, medication adherence, and heal th behavior and education, were also considered as they were hypothesized to help explain what fact ors people would take into account that would lead t hem to obtain and use medication. Help Factors In the new framework, people were able to ac quire their desired prescription if they were willing to access a known source of hel p, or if help was provided to them by someone else (a healthcare professional). Ne ither the Piette et al. (2006a) nor the Andersen & Aday (Andersen & Davidson, 20 07) frameworks mention anything that is similar to the concept of being willing to a ccess a known source of help. While both frameworks mention effects of the larger heal th system in general, neither seem to take into account sources of assistance that woul d be known to patients that exist for those mostly without third-party coverage, such as a local clinic or a hospital outpatient pharmacy program, two sources m entioned in this study. K nowledge of these programs and willingness to access them had a pronounced effect on whether or not someone was able to get their prescription, so th is factor should probably be included in any framework regarding medication acce ss for this kind of population. The other help factor was that of help being provided by someone else, in this case the prescribing physician, who either offered samples or referred people to a 101

PAGE 102

program. Piette et al (2006a) allude to this as a clinician factor, whereby the clinician is able to help someone acquire their prescrip tion. Having a clinici an provide help is not explicitly mentioned in the Andersen & Aday (Andersen & Davidson, 2007) framework, but would probably be considered an enablin g factor, thus enabling people to access their medication. Monetary Factors The other important set of factors that affected whether or not someone initially reported CRN was monetary factor s. Unsurprisingly, simila r factors exist in both the Piette et al. (2006a) and Andersen & Aday (Andersen & Davidson, 2007) frameworks. Piette et al. (2006a) re fer to these as financial pressures, which include the factors of income, Rx coverage, OOP Rx costs, and other health costs. These factors are very similar to the ones that were developed in this framework; the OOP Rx cost depended on Rx coverage and issues with coverage, and the willingness to pay depended on income and other costs in general (not ju st health costs). The Andersen & Aday (Andersen & Davidson, 2007) framework includes income as a predisposing demographic factor, and Rx coverage as an enabli ng financial factor. The biggest aspect of monetary factors that is different within the new framework is the issue of problems with existing Rx co verage. People in this study had various issues with their existing Rx coverage that directly affected their ability to acquire their prescriptions. These issues included: A gap in coverage for moving, switch ing jobs, or administrative problem A limit in the number of prescriptions covered Delayed reimbursement for covered prescriptions A stop of coverage for no apparent reason A co-pay that was still too high even with coverage 102

PAGE 103

Perceived Need In the new framework, after individuals are in itially not able to acquire their desired prescription, the extent of their perceived need influences whether or not they will seek help for acquiring the prescription. In a s ense this perceived need is re-evaluated; in other words, they had some amount of perceived need when obtaining the original prescription, but once they initially cannot acquire it, they evaluate how much they really need it. Everybody had a high enough init ial perceived need to acquire the paper prescription in the first place, but this perceived need potentially changed once they could not initially acquire the medication. Perceived need is mentioned in both the Piette et al (2006a) and Andersen & Aday (Andersen & Davidson, 2007) frameworks as having an effect on accessing medication. However, it is unclear whether these fr ameworks are referring to a persons perceived need when obtaining their paper prescription, or their re-evaluated perceived need after they could not acquire the medication. Therefore, the perceiv ed need used in this framework (one that is re-eva luated after the medication has initially not been obtained) may be different from the perceiv ed need in these other frameworks. In the new framework, perceiv ed need depended on three major factors: Perceived consequences of not taking Rx Time period of need Perceived availabilit y of alternatives The idea of people taking into account the se verity of consequences of not taking their medication is a common one in the framew orks and bodies of lit erature examining medication use in patients. In the Piette et al. framework (2006a), constructs that refer to a similar concept include perceived need, effect on current HRQL, and effect on 103

PAGE 104

life expectancy. In the A ndersen & Aday framework (Andersen & Davidson, 2007), this is captured by the construc ts of perceived need and per ceived health. The Health Belief Model (Rosenstock, 1966; Janz & Becker, 1984) has a construct labeled perceived severity. Social Cognitive Theory (Bandura, 1986) includes outcome expectations and knowledge of health risks. The Theory of Reasoned Action / Theory of Planned Behavior (Fishbein & Ajzen 1975) has the concept belief about the outcome of the behavior. In essence, this fa ctor is not new, and is comparable to many constructs that already exist. Given this, it is not surprising that people took severity of consequences of not taking the medicati on into account when determining their perceived need for the medication. For the individuals of this sample who had chronic high blood pressure, the perceived severity of the consequences of not taking their blood pressure medication was extremely high, leading to a higher perce ived need for the medication than many of those for more acute conditions. Piette et al. (2006b) report ed a similar finding in that, when faced with cost pressures, people ar e more likely to fo rgo symptom-relief medication, as opposed to preventive medicati on for chronic diseases. This was also shown in general adherence lit erature (Jin, Sklar, Min S en Oh, & Chuen Li, 2008) that people who believe that consequences of not taking medication are severe are more likely to be adherent, thus w ould exhibit a higher perceived need for the medication. So, it seems as though the belief in the seve rity of the consequences for not taking medication is an essential fact or when people are re-evaluating their perceived need for the medication after not being ab le to initially acquire it. 104

PAGE 105

The two other factors that influenced perceived need, time period of need and perceived availability of alter natives, have not been mentioned in either the Piette et al. (2006) or the Andersen & A day (Andersen & Davidson, 2007) frameworks as having an influence as to whether or not someone r eported CRN. This represents yet another way in which the new framework is different from these others that attempt to explain CRN. The time period of need referred to the am ount of time for wh ich medication was needed. This was manifested in two differ ent situations: a person had a gap in Rx coverage for a known period of time, or a person had a condition that would be expected to be resolved in a short amount of time. Those experiencing a gap in Rx coverage for a known period of time usua lly had chronic conditions where they had been taking the medication for a while before experiencing this gap. Since they knew that they may only be without medication for a finite period, less than 30 days for the people in this study, their perceived need for t he medication was different than if they knew that they were going to be without thei r chronic medication for a long time. So, even if they had a condition for which they believed the consequences of not taking the medication would be severe, their perceived need for the medication was not as high since they knew that Rx coverage that would help pay for thei r medication would be available in a known period of time. Some people had a condition that would be ex pected to resolve in a short amount of time, such as an infection. All of these people seemed to believe that the consequences of not taking medication would be severe at least in the short term, yet some of them did not have a high enough of a perceived need to seek out help to 105

PAGE 106

purchase their medication. On the other hand, all people who had chronic conditions where they believed that the consequences of not taking the medication would be severe had a high enough perceived need that they ended up seeking out help. Although the people who did not s eek out help were not asked explicitly if their time period of need affected their perceived need, the time period of need was the only apparent factor that would seemed to account for their lower perce ived need. Future research should attempt to determine if peopl e with acute conditions that are expected to resolve in a short amount of time take this time period of need into account when determining perceived need. Another major factor, that of the alternatives that are available, has not been explored in depth in either the Piette et al. (2006a) or Andersen & Aday (Andersen & Davidson, 2007) frameworks geared toward attempting to explai n CRN., However, availability of alternatives has been mentioned in the adher ence literature (Pound et al., 2005) in a similar way as defined in this study. The alternatives that people mentioned were not intended to substitute for the medication; no one seemed nave enough to believe this. Nor were they necessarily thoug ht of as minimally acceptable alternatives, because in some cases they were not seen to be acceptable. Rather, it was more of the idea that if a person has a choice of using an alternative, this could affect their level of perceived need of a more expensive medi cation. Pain was one example of a condition where people thought that there were relatively cheap, although sub-optimal, alternatives. A person is in pain, and is presented with a prescription that they cannot afford. There is however, an OTC product th at can possibly reduce pain, but probably not as effectively. Knowing that this alte rnative is available therefore reduces their 106

PAGE 107

perceived need for the prescription medication, so they do not seek help to acquire it, and they use the OTC product instead. Another example was for chronic conditions in which the condition could possibly be improved by lifestyle modifications, such as high cholesterol. Instead of trying to pay fo r an expensive medication, people figured that they could get some benefit from improving their diet or exercising more, so their perceived need for medication was lower. Again, people bel ieved that these alternatives were sub-optimal; however, kno wing that the alternatives were available seemed to decrease their level of perce ived need for the prescription medication. Knowledge / Belief that Help is Available Neither the Piette et al. (2006a) nor the Andersen & Aday (Andersen & Davidson, 2007) framework suggest that the knowledge or belief that help is available play any part in whether or not CRN is reported. T he Piette et al. (2006a) framework mentions sociocultural influences, but this is geared more towards general medication beliefs. Similarly, individual predisposing beliefs in the Andersen & Aday (Andersen & Davidson, 2007) framework also seem to refer to general medication beliefs. No beliefs about the availability of assistance to acquire m edication seem to be mentioned in either framework. Seeking Help or Additional Resources In the new framework, seeking help to acquire a medication is a response to initially not being able to ac quire a medication. The two major frameworks with which this new framework is being compared do not have this response step, so it is not surprising that seeking help to acquire a medication does not appear in them. In essence, the Piette et al. (2006a) and t he Andersen & Aday (Andersen & Davidson, 2007) frameworks only examine the factors that lead up to t he initial reporting of CRN. 107

PAGE 108

In the new framework, seeking help was show n to be the key factor of whether or not someone was able to eventually acquire thei r medication. But, since it only happens after the prescription is initia lly not acquired, there is no real comparison that can be made with the other frameworks. Factors Not Having an Eff ect in the New Framework It was postulated that some of the factor s that affect adherence to medication (that has already been acquired) would affect whether or not a person reported CRN, even though CRN is a condition that affects access to medication and not whether or not it is utilized given that there is access. For ex ample, a person is taking a medication for a chronic disease and needs to pay $10 to get a refill. This person does not like the side effects, so he or she does not purchase t he refill. When asked if he or she has experienced CRN, this person could repor t yes, even though the cost of the medication was of secondary importance in this persons decision making process. Several factors that affect adherence to medi cation as mentioned by Pound et al. (2005) include: They weigh the benefits vs. t he adverse effects of the drug Taking the medicine does not fit in their daily routine They are nave scientists they try it or stop it, and see what works best Their condition is improving They cant tell the difference They substitute with something else They dont like medicine in general A negative self-identity co mes from taking the drug They are worried about addiction Only one of these factors, substituting with something else, was similar to a factor in the new framework, perceived availability of alternatives. This factor only came into play when the person responded to not being able to initially acquire their medication, 108

PAGE 109

thus did not impact whether or not they initially acquired it (where CRN would be reported). In addition, there was no mention of side effects, or of not liking an effective drug. Many were believers in using medi cation in general to solve their problems. There were some who were not; while they ac knowledged that they did not like to take medication in general, they r ealized that they needed to continue taking a specific medication in order to mainta in their quality of life. Other therapy-related factor s that affect adherence are mentioned both by Piette et al. (2006a) and by Jin, Sklar et al (2008) which include the following: Route of administration Treatment complexity Duration of the treatment period Taste of the medication Requirements for drug storage None of these factors were mentioned by any of the people interviewed. In general, once people acquired thei r medication, they reported being happy to take it. So, in summary, it did not seem as though m any of these factors that affect whether someone utilizes a medication affects whether or not someone accesses it. One of the major objectives of this research study wa s to determine if any of the above mentioned utilization-related factors, in combination with monetary factors, led people to not acquire their medication and report that the medication was not acquired because of the cost. The results of this research, how ever, tended to show that, unless help was provided or if someone was willing to acce ss known help to get their medication, whether or not someone acquire d their medication initially depended solely on monetary factors. 109

PAGE 110

Summary While there are some similarities with t he Piette et al. (2006a) and the Andersen & Aday (Andersen & Davidson, 2007) frameworks, the new framework does contain many notable differences. The most obvious one is that CRN is seen as a more dynamic process; conditions change over time, so it should not be thought of as just a one-time event. Also, many factors related to the ut ilization of drugs, such as side effects or concerns about drug effectiveness, do not appear to be a factor in whether or not someone acquires a prescription in situations when they report CRN. In addition, the key factor in whether or not someone acqu ired their prescription was whether or not they received help. Finally, ones perceived need for the prescription not only took into account the perceived severity of not taking it, but also the time period for which the prescription was needed and the per ceived availability of alter natives for this period of time. This new framework gives a clearer understanding of the factors that affect whether CRN is reported in the first place, a nd the factors that af fect whether or not people eventually acquire their medication. Implications There are implications from this research regarding how physicians and pharmacists practice could change to help mo re people acquire their prescriptions, how health care policy could be shaped to help th is population, and how future research in this general area could be improved. Each of these is discussed in detail below. Practice Implications The key factor of whether or not a person with a potential problem of not acquiring medication because of cost ulti mately acquired their medication is whether or not they 110

PAGE 111

received help. There were three specific ways in which people is this study received help: 1. Help was offered by a physician based on being aware of t he patients financial situation at the time the medication was needed. Specific questions were not asked regarding how the physician becam e aware of this, but they offered assistance on the spot to help the patient acquire their medication. This was described as the physician taking a more active role in offering help. 2. The patient actively sought out help (or acquired additional resources) in response to not being able to initially acquire the medication. They sought out help from a variety of sources, including health care professionals. 3. At some point, the patient learned about a clinic or a hospital outpatient program where they could obtain medication bec ause of a more reas onable OOP cost. The people interviewed in this st udy learned about thes e programs based on previous experiences or by learning from others Health care professionals can play an im portant role in each of the different scenarios. In the first, the health care professional is acti vely trying to help the patient acquire their medication, based on their knowle dge of the situation. In the second, a patient is asking the health care profession al for help, and the health care professional may provide it. In the thir d, a patient can learn about a program from a health care professional. Since health care professionals can play such an important role in whether or not a patient receives help to acquire their m edication, it makes sense to examine how different types of health care professionals can help, and what the best help is that they can provide. Underlying this discussion is the realization t hat the current health care system in the United States is disjointed, with no one entity being responsible for an individuals care. Therefore, any assistance that could be provided by health care professionals would have to occur with the init iative of the individual, or possibly on the level of the clinic, practice, or pharmacy. In addition, heal th care professionals may not 111

PAGE 112

have the necessary amount of information, re sources, or incentives to provide the necessary assistance. In spite of all this, the roles of the two types of health care professionals that are integral to the process of receiv ing medication and would seem to have the best opportunities to provide help in acquiring medication, physicians and pharmacists, will be examined. Physicians Physicians are the ones who come in contact with the patient, diagnose the medical problem, and develop the therapeutic plan to alleviate it, often with the involvement of medication. This puts the physician in place to be able to assist the patient right away; unfortunately, this role of ensuring that patients are able to acquire their medication is one that most physicians are not used to and have never really been trained to perform, even though the AMA (2009) code of ethics, policy H-110.997 (Cost of Prescription Drugs) states t hat all patients must have acce ss to all prescription drugs necessary to treat their illnesse s. First of all, physicians are not very good at identifying which patients might need some kind of financia l help. Alexander, Casalino, & Meltzer (2003) report that even though both physicians (65%) and patients (63%) want to talk about medication costs, it rarely happens (85% of patients reporte d that it never did). Similarly, as Heisler, Wagner, & Piette ( 2004c) point out, physicians do not know which patients have problems paying for their medications. However, maybe it is not just the physicians failure to pro-actively ask pat ients about financial issues; Piette et al. (2004d) reported that only 33% of patients who had problems paying for their medications spoke with their doctors about costs. So, it seems that physicians are simply unaware that paying for medication is a problem for some of their patients, so that talking to their patients about cost iss ues and figuring out the be st way to help them 112

PAGE 113

acquire the desired medication is not a role that many physicians currently know how to fulfill. One of the reasons behind th is may be that physicians do not know what they can do to help people with financial problems, (Heisl er et al., 2004c). This was the finding reported by Reichert, Simon, & Halm (2000), in which 88% of physicians thought one should take cost into account when prescr ibing medication, but 80% did not know how much drugs cost, thus they did not feel equipped to help patients with cost problems. In reality, however, there are some fairly easy steps that physicians can take to help make medication more affordable fo r patients. As summarized by Alexander & Tseng (2004), some basic help can be given as follows: Switch to a generic, if available Stop non-essential medications Give samples to provide access for a brief period of time Prescribe a larger dose and teac h the patient to split pills Refer the patient to an assistance program Encourage the patient to shop around for the best price In this study, physicians who pro-actively offered help to patients in acquiring their medication used two of the strategies abov e: providing samples and referring to a program. Since some physicians are able to provide assistance to patients in order to acquire their medication, maybe it would be a good idea to train all physicians to screen for problems related to inabilit y to afford prescribed therapy and to provide assistance and referrals when problems are identified. Unfo rtunately, this is not part of the current training of physicians, and as Korn, Reichert, Simon, & Halm (200 3) report, it is challenging to get physicians to change their pr escribing patterns to make it easier for patients to acquire their m edication. After an educational intervention that was designed to teach doctors about medicati on costs and how insurance coverage 113

PAGE 114

impacts patient cost, the number of physician s who asked patients about cost only went from 22% to 27%. Training physicians to prov ide this kind of help, therefore, might not be the most feasible solution unless it is i ngrained in physicians during their training that this is a necessary component of working up a patient and devising a therapeutic plan. In summary, physicians are not used to providing help to patients in terms of the patients being able to acquire their medicati on. While there are some relatively straightforward strategies that can be im plemented, this has not been a role that physicians are trained for, and training them to change their behavior can be difficult, so it would be difficult to rely on physicians to assist people in ma king sure they can acquire their medication. W hatever help they can provide will obviously be beneficial, but in the short term one s hould not expect physicians to offer this kind of help on a regular basis. Pharmacists Pharmacists are obviously an integral par t of the drug deliver y system, as they have to endorse every prescription that is f illed, and also are typically present when a prescription is dispensed to the patient (with the exception of mail-order or online pharmacies). Even though the traditional ro le of the pharmacist does not include providing assistance such that people can acqui re their prescriptions (as evidenced by none of the study participants r eporting that they received help from a pharmacist), pharmacists, given their unique skill set and f unction in delivery system, seem poised to provide assistance to those who are not able to acquire their prescriptions because of cost in several key ways. First of all, pharmacists can more easily s pot the problem as it occurs. Since the pharmacist is present when a person is purc hasing their prescription, if the person 114

PAGE 115

cannot acquire the prescription because of co st, the pharmacist can be cognizant of the situation immediately. This is an obvious but important step; whereas physicians mostly prescribe and then go on to the next pati ent, the pharmacist is present when the purchase is attempted, thus is able to learn immediately of any problems. Next, given their greater knowledge of drugs and appropriate cheaper alternatives, pharmacists have the potential to act as better cost agents for patients, and can use this knowledge to make appropriate dispensing decisions direct ly to the patient (Mott & Cline, 2002; Suh, 1999). The most obvious knowledge that pharmacists have is of cheaper generic drugs, something that most physicians do not have general knowledge of (Banahan 3rd & Kolassa, 1997, Hellerstein 1998), especially considering that physicians are bombarded with information and gi fts from the pharmaceutical industry which promote their high money-making brand name drugs (Wazana 2000). Besides knowing about generic alternatives, pharmacist s also have the knowledge and skills to communicate either with physicians or insur ance providers about swit ching to a cheaper alternative drug. Therefore, especially co mpared with physicians, pharmacists are more able to recommend a cheaper alternative. Pharmacists who are willing to put in more effort are able to provide other ways to help patients acquire their medication. Many interventions have been attempted by pharmacists to increase access to medicati ons for the medically indigent, usually through assisting with manufacturer assist ance programs (Hotchkiss, Pearson, & Lisitano, 1998; Mounts, Ringenberg, Rhees, & Partridge, 2005; Prutting, Cerveny, MacFarlane, & Wiley, 1998), setting up a 340B pharmacy (through the US Public Health Service) to provide cheaper drugs (Dent, St ratton, & Cochran, 2002), or to provide 115

PAGE 116

general counseling to help the pa tients in the most expedient direction to cheaper drugs (Brahm, Palmer, Williams, & Clancy, 2007; Weiner, Dischler, & Horvitz, 2001). Admittedly, these types of interventions would require more effort, but this is something that pharmacists are able to do to hel p patients acquire their medications. Finally, the simplest thing that pharmaci sts (and any appropriate people that interact with prospective medication buyer s at a pharmacy) can do is to learn about what local programs are available, and to pass on a name, phone number, or website for the person to get more information. In ma ny cases, this can be as simple as telling the person about something as straightforwar d as dialing 211, the relatively new national phone number for general social services, which can then guide people to appropriate services. At this point, the pharma cist (or relevant person) does not have to help people directly, but should at least be able to pass along a phone number to an organization that can provide help. It was dishearteni ng to hear people tell about when they went to a pharmacy to tr y to acquire a medication that was more than they were willing to pay and to be told something along the lines of it is what it is or theres nothing I can do. The very least that pharma cies should do for these people is to give them a phone number. In summary, it seems that pharmacist s (and the profession of pharmacy in general) can do a lot to provide help to t hose who are not acquiring their medications because of cost. Pharmacists are present fo r every purchase in a community outlet, so should be able to easily determine if there is a problem. Pharmacists have a unique skill set that enables them to act as better cost agents for patients, namely they know about cheaper alternatives and costs in general. Finally, any time that people are not 116

PAGE 117

acquiring a prescription because of cost, the relevant person at the pharmacy can do at least the minimum of forwardi ng information about sources of help. Until the disjointed health care system changes and assistance to acquire medication can be given via more system-level interventions, it seem s as though pharmacists have the potential to provide assistance to these people who are not acquiring their prescriptions because of cost. Policy Implications One of the factors that influenced a persons perceiv ed need for the medication after they were not able to initially acqui re it was the time period of need for the medication. Specifically, those who had a shorter time period of need, because of either a known, finite coverage gap or bec ause they had a medical condition that would be expected to resolve in a short period of time, seemed to have a lower perceived need for their medication. From a policy perspective, it was interesti ng to observe that this time period of need was rather short for many of the people to whom this applied, measured in weeks or at most a month. These coverage gaps seem sma ller than those typically reported by those not having insurance for a certai n period of time. One Kaiser Family Foundation (KFF) general report on the uninsur ed (2008) stated that only 17% of the uninsured say that they have been uninsured for <6 months. Another KFF report concentrating on the length of coverage gaps (Haley & Zuckerman, 2003) stated that only 20% of those reporting being uninsured at some point in the last 12 months were uninsured 5 months or less, and thus concludes the uninsured are not largely composed of people experiencing short-term gaps in coverage. It seems as though the length of the coverage gap may be another wa y in which this sample population is 117

PAGE 118

different from the population at large. On t he other hand, it could al so be that a certain percent of those without coverage have no m edical conditions and so may not be as motivated to obtain insurance fo r prescription medications. Regardless, one major policy implication is that providing short-term assistance ( 30 days) would have helped many of the people in this study, in many cases eliminating the situation in which CRN was reported. Assist ing people with a subsidized 30 day supply of their medication would have bridged the coverage gap for those who reported one. In addition, this subsidized 30 day supply would have also helped those who did not acquire medication for their viral infections, which usually manifest themselves in a shorter period of time. For those who need additional coverage beyond 30 days, this first 30 day period could give people a buffer zone for finding help, by any of the means listed previously. Based on this study, it seems as though providing a subsidized 30 day supply of the needed medi cation could have a huge and positive impact for those reporting CRN. Research Implications For those performing general health policy research, several implications resulted from this study. Each of them is discussed below. Insurance does not necessarily equal access As described above, people in this study had various issues with their existing Rx coverage that directly affected their ability to acqui re their prescriptions. These issues included: A gap in coverage for moving, switch ing jobs, or administrative problem A limit in the number of prescriptions covered Delayed reimbursement for covered prescriptions A stop in coverage for no apparent reason 118

PAGE 119

A co-pay still too high with coverage People experiencing these issues would all report that they have existing Rx coverage, yet still also report CRN. Ther efore, simply having Rx coverage is not enough to enable access to prescriptions; one mu st have issue-free access. As mentioned in Chapter 2, many studies (Smith & Kirking, 1999; Stockwell et al., 1994; Fillenbaum et al., 1993; Sleath et al., 2004) using the Andersen & Aday (Andersen & Davidson, 2007) framework used having insu rance as a proxy for having access to medication. This research suggests that this may be a mistaken assumption, since many of these people had insurance coverage yet still could not acquire their prescriptions. Access related factors and othe r utilization related factors As mentioned previously, it did not seem as though many of the factors that affect whether someone actually utilizes a m edication affects whether or not someone accesses it. For this reason, it may ma ke sense in future research dealing with medication use to separate out the stages of a ccess and actual utilization. Clearly, in order for a medication to be utilized properly, it first must be accessed, then utilized. However, it seems as though there are differ ent factors that affe ct whether or not someone is able to access medication and then, once accessed, whether or not someone utilizes the medication. For exampl e, monetary factors influence whether or not a medication is accessible, but may not affect how someone utilizes the medication after it has been acquired. Similarly, worri es about side effects do not seem to affect whether or not the medication is accessed, but do seem to affect whether someone utilizes a medication. Combining both a ccess-related and other utilization-related factors in one framework, as both the Pi ette et al. (2006a) and the Andersen & Aday 119

PAGE 120

(Andersen & Davidson, 2007) frameworks do, therefore makes it harder to gain a complete understanding of each of these two different stages of the medication use process. This new framewor k will hopefully provide the foundation for future research that will concern itself solely with medi cation access; new, sepa rate frameworks can then be developed that concern themselves with medication ut ilization that assumes the medication has already been accessed. CRN is not a one-time, dichotomous event As mentioned earlier, all of the study s ubjects reported CRN. However, it does not seem useful to consider that they bel ong in the same bucket (those who have reported CRN), since their experiences and needs ar e so varied. For this reason, future survey-level research should probably avoid only asking a single yes/no question about whether or not CRN occurred, and design follow-up questions that can better tease out the dynamic characteristic of CRN. For ex ample, additional questions such as the ones below (after being revised to be proven valid and reliable) would help policy makers learn more about this population in order to better shape policies: 1. How many different prescription medicati ons did you not get filled in the last 3 months because you could not afford them? 2. The last time you experienced not getti ng a prescription filled because of cost, did you eventually acquire the medication? 3. If so, how long were you without your medication? (code responses in ordinal categories from one day to over a month) 4. Will you have problems acquiri ng prescriptions in the future? The first question would give policy makers an idea of how frequently CRN episodes are occurring for those who are repor ting at least one episode (from the initial yes/no question). This would give insight as to whether or not a sample person keeps 120

PAGE 121

experiencing CRN over and over again, or if it for just a one-time event. Different policies would be designed based on these different situations. As this study showed, many people who reported CRN eventually acquired their prescription. This second question would give some insight as to what percentage of people reporting CRN are eventually acquiring their prescripti ons; this percentage would influence the type of policy cr eated to help this population. For those who eventually acquired the pr escription, the third question regarding a persons time period of need would give poli cy-makers an idea on the extent of shortterm assistance necessary to help people bridge the gap until they can acquire their prescriptions. As mentioned earlier, many people reporti ng CRN only needed assistance for a relatively short period of time ( 30 days), due to having a condition that would be expected to resolve itself in a short per iod of time (a viral infection or cute pain episode) or experiencing a known gap in cover age. Answers to the third question would give policy makers a better idea whether or not the results from this study ( 30 days) are possibly generalizable fo r a larger population. Some of the people who reported CRN for th is study would not be expected to have problems acquiring prescriptions in the future, because their coverage gap had ended, or because they had found a clinic or program that they would access when needing prescriptions. Answers to the fourth question would give information regarding what percentage of the people who originally report CRN would likely not report CRN in future situations, giving a clearer picture of the true extent of the problem of CRN. The list of questions above is just a pre liminary attempt and is not intended to be exhaustive, but the answers to these questions (in addition to the initial yes/no question 121

PAGE 122

regarding CRN that is current ly used) can be used to gain some insight on this population reporting CRN, and ho pefully affect policy. Limitations There are many limitations that apply to any kind of qualitative research, namely that the sample is necessarily not represent ative, the data collected not reproducible, and the coding and analysis dependent on the researcher. While nothing can be done regarding the sample and the data collection, another researcher was brought in to independently code and analyze t he data with issues resolved by a third researcher, which improves the validity and reliability of the results. Having independent researchers brought in also helps to ensur e that the process of grounded theory was followed, such that the theory formed must be grounded in the data that were collected. Another limitation is that the invest igator, as the only recruiter of study participants and the only interviewer, may have been biased in who he attempted to recruit, and how he directed the interviews. The age and racial distributions of the participants that were recr uited are diverse, but he still could have shown some unconscious bias by taking a glance at the people in the waiting room and then choosing when to go to recruit patients. R egarding the interviews, whatever deviations he did take were at least known to t he independent researcher that coded and analyzed the data. It may also have been desirable to recr uit more people with higher incomes and/or better insurance. The clinic was chos en as a place to recruit because it seemed like a location where people who would be li kely to report CRN would show up on a consistent basis. However, other studies have shown that those with higher incomes 122

PAGE 123

also report CRN. There did not seem to be a convenient and efficient way to target these people, so more focus was placed on the population of this particular clinic. Summary of Conclusions and Implications The following points highlight what was learned in this study and include potential avenues for further research: Conclusions Dynamic nature of CRN In the current literature, CRN is usually treated as a one-time event, determined by a yes/no question. This research showed that some who reported not acquiring a prescription because of cost actually ended up acquiring the prescription. Some were able to get into a program or discovered a clinic where they could get lower-priced prescriptions, such that CRN will be unlikel y to be reported in the future. The new framework better describes the phenomenon, and should hopefully help guide future research in this field. Getting help (or additional resources) was the difference between getting and not getting the prescription The people who eventually acquired their prescriptions received help in one of the following three ways: 1. They were provided help by someone else 2. They were willing to access known sources of help 3. They had a perceived need of the prescr iption that was high enough such that they sought out help or additional resources, and received enough help If no help is provided or willingly accessed, initial prescription acquisition depends on monetary factors without help, only monetary factor s influenced whether or not the prescription was acquired (i.e. the willingness to pay was greater than the OOP cost). 123

PAGE 124

The OOP cost was mostly dependent on w hether or not one had third party drug coverage, and whether there were issues wit h the coverage that provided barriers to access. People reported being willing to pay something (on the order of $5-$20) for their prescription even in situat ions where they reported CRN. Access not affected by other utilization-related factors Many factors have been shown in the adherenc e literature as having an effect on utilization; in this study, none of these prev iously mentioned factors, such as general medication beliefs, affected whether or not one acquired the prescription. Perceived need was b ased on three factors In this study, level of perceived n eed for the medication was shown to be an important factor in determini ng whether or not someone sought out help to pay for their prescription. Perceived need was dependent on the combination of three factors: 1. Perceived consequences of not taking the medication 2. Time period of need 3. Perceived availabi lity of alternatives One of these factor s, the time period of need, has not been previously shown to affect ones perceived level of need for the medi cation. People doing research using a similar construct as perceived need should take these three factors into account. Implications Future research should distinguish betw een access and other utilization related factors The drug use process consists of first accessing the drug, and then utilizing the drug once it has been accessed. There are some factors that have been shown to affect access (income, third par ty coverage, etc.) that do not affect utilization, and there are other factors (see list above) that have been shown to affect utilization, but do not 124

PAGE 125

affect access. Futu re research that examines the drug use process should consider drug access and drug utilization as two separate and sequential stages, and clarify which factors affect each stage. Insurance does not necessarily mean access In previous health care utilization resear ch, especially research that uses the Behavioral Model of Health Services Use by Andersen & Aday (Andersen & Davidson 2007) as a conceptual framework, simply having insurance is equated with having access. This study showed that many people who reported having insurance had issues with their insurance that affect ed their access, including having a gap in coverage, the coverage stopping unexpectedly, the drug not being covered, delayed reimbursement, there being a limit on how many drugs will be covered, and the OOP still being more than ones willin gness to pay even with coverage. Care should be taken to ensure that having insurance r eally does equate to having access. Short-term solutions Many of the people who reported not acquiri ng a prescription because of cost only needed medication for a short period of time (less than or equal to a month), either because of a short gap in coverage, or because they had a medical condition which would be expected to resolve itself in a short period of time (such as a viral infection or a one-time, acute pain episode). Any poli cy or program designed to help someone acquire just a month-long supply of medica tion would seemingly help many people. Physicians can offer help, but should not be relied upon to do so In this study, two people reported that physicians helped them acquire their medication, one by offering samples, another by referral to a progra m. While it was good that these physicians offered to help, many physicians do not have the knowledge 125

PAGE 126

126 about drug prices, and generally do not consi der price (or other access issues) when developing a therapeutic plan for the patient, so it would be difficult to rely on physicians to try and improve peoples access to medication. Pharmacists are in a unique position a nd have a unique skill set to help people acquire their prescriptions Pharmacists have the potential to help people acquire their medications. First of all, they are usually present when a person initially does not acquire their prescription, so they are aware of the problem as it happens Next, given their greater knowledge of drugs and appropriate cheaper alternatives, pharmacists have the potential to act as better cost agents for patients. They are in a position to communicate directly with a physician or insurance company to come up with the most appropriate and costeffective solution. Finally, since they are present when the problem occurs, they can at least provide some information (such as a phone number or website) so that the person has somewhere to go to attempt to get help. Until the disjointed health care system changes and assistance to acquire medica tion can be given via more system-level interventions, it seems as though pharmaci sts have both the unique position and skill set to provide assistance to these people who are not acquiring their prescriptions because of cost. Summary This research was undertaken to gain a better understanding of situations in which people report not acquiring a pr escription because of cost. It is hoped that the new framework, coupled along with other results, will enable fu ture researchers to do a better job in identifying the problems that need to be solved, and coming up with appropriate solutions to eventually increase access to medications for this population.

PAGE 127

APPENDIX INTERVIEW GUIDE Section 1 General experie nces with and without CRN Experience with underusing medication because of cost 1. Please describe the situation where you di d not get a prescription filled (or space it out, or not refill, as appropriate. 2. Why did you not have it filled (or spac e it out, or not ref ill, as appropriate)? 3. Were there any other reasons, in addition to cost, that affe cted your decision? Price / Value 1. How much did the prescription cost?? 2. At what price would you have purchased it? 3. What did you do to try to reduce the price? 4. What assistance did you get to help you pay for it? 5. What did anybody else (doctor, pharmacist, social worker, friend, family, etc.) do to try to help you afford the medication? 6. When you talk about cost, what does that mean to you? Experience with purchasing a medication/using as directed 1. Now please tell me about the last time y ou purchased a prescripti on. What was it for? (drug and/or disease state) 2. What was different about this experience from the situation where you did not fill the prescription? Section 2 Directed questions to explore specific constructs (to be used as appropriate) 1. What did you think would happen to your health if you did no t take/underused your medication? (Perceived susceptibility, perceived benefits, attitude behavioral belief, expectations) 2. What actually happened? 3. How serious did you think the risks were by not using the medication? (Perceived severity) 127

PAGE 128

128 4. Generally, how do you feel about taking medications? 5. How much do you value the supposed impr oved health you would get from taking your medications as suggested? Section 3 Demographics (tape record ing turned off and answers to be recorded by hand) Age Roughly how old are you? Sex (This will not be asked but recorded by observation in notes recorded after very interview) Self-described race/ethnicity How would you describe your race/ethnicity? Self-rated health (1-5) How would you rate your overall health: poor, fair, good, very good, or excellent? Household income What is your approximate household income: less than $40 thousand, or greater than or equal to $40 thousand? o (If less than $40 thousand) Is it le ss than $20 thousand, or between $20 and $40 thousand? o (If greater than or equal to $40 thousand) Is it between $40 and $60 thousand, or greater than $60 thousand? # in household How many people are in your household? Approximate monthly househo ld OOP costs for prescrip tions Approximately how much money is paid out-of-pocket for prescriptions every month for your entire household? Insurance status Generally, what kind of insurance do you have? Drug coverage What kind of coverage do you have for purchasing prescriptions?

PAGE 129

LIST OF REFERENCES Alexander, G., Casalino, L., & Meltzer, D. (2003). Patientphysician communication about out-of-pocket costs. Journal of the American Medical Association, 290 953-958. Alexander G., & Tseng C. (2004). Six strategies to identify and assist patients burdened by out-of-pocket prescription costs. Cleveland Clinic Journal of Medicine, 71 (5), 433-437. American Medical Association. (2009). Health and ethics policies of the AMA House of Delegates Retrieved September 7, 2009, from http://www.ama-assn.org/adcom/polfind/Hlth-Ethics.pdf Andersen, R.M. & Davidson P. (2007). Improving access to care in America: individual and contextual indicators. In: Andersen R, Rice T, Kominski J, eds. Changing the U.S. Health Care System: Key Issues in Health Services Policy and Management (pp. 3-31). San Francisco, CA: Jossey-Bass. Bajcar, J. (2006). Task analysis of patients medication-taking practice and the role of making sense: a grounded theory study. Research in Social and Administrative Pharmacy, 2 (1), 59-82. Banahan 3rd, B. & Kolassa, E. (1997). A physician survey on generic drugs and substitution of critical dose medications. Archives of Internal Medicine 157 (18), 2080-2088. Bandura, A. (1986). Social Foundations of Thought an d Action: A Social Cognitive Theory Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (2004). Health promotion by social cognitive means. Health Education & Behavior 31(2), 143-164. Blumer, H. (1969). Symbolic Interactionism: Perspective and Method Berkeley: University of California Press. Brahm, N., Palmer, T., Williams T., & Clancy, G. (2007). Bedlam Community Health Clinic: a collaborative interdisciplinary health care service for the medically indigent. Journal of the American Phar maceutical Association, 47 398-403. Bower, K., Frail, D., Twohig, P., & Putnam, W. (2006). What influences seniors choice of medications for osteoarthritis? A qualitative inquiry. Canadian Family Physician, 52 (3), 343. Carder, P., Vuckovic, N., & Green, C. (2003). Negotiating medications: patient perceptions of long-term medication use. Journal of Clinical Pharmacy and Therapeutics, 28 409-417. 129

PAGE 130

Dent, L., Stratton, T. & Cochran, G. ( 2002). Establishing an on-site pharmacy in a community health center to help indige nt patients access medications and to improve care. Journal of the American Phar maceutical Association, 42 497-507. Dolovich, L., Nair, K., Sellors, C., Lohfeld, L ., Lee, A., & Levine, M. (2008). Do patients expectations influence thei r use of medications? Canadian Family Physician, 54(3), 384-393. Dowell, J., & Hudson, H. ( 1997). A qualitative study of m edication-taking behavior in primary care. Family Practice, 14 369-375. Felland, L. & Reschovsky, J. (2009). More non-elderly Americans face problems affording prescription drugs Center for Health Syst em Change tracking report No. 22. Retrieved March 20, 2009 from http://hschange.org/CONTEN T/1039/?topic=topic01 Fillenbaum, G., Hanlon J., Corder, E., Ziqubu-Page, T., Wall Jr., W., & Brock, D. (1993). Prescription and non-prescription dr ug use among black and white communityresiding elderly. American Journal of Public Health, 83 (11), 1577-1582. Fishbein, M. & Ajzen, I. (1975). Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research Oxford, England: Pergamon Press. Glanz, K., Lewis, F. M. & Rimer, B. K. eds. (1997). Health behavior and health education: theory, research, and practice 2nd ed San Francisco: Jossey-Bass. Glaser, B. & Str auss, A. (1967). Discovery of grounded theory Chicago: Aldine. Glaser, B. G. (1978). Theoretical sensitivity: Advances in the methodology of grounded theory. Mill Valley, CA: Sociology Press. Guest, G., Bunce, A., & Johnson, L. (2006) How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18 (1), 59-82. Haley, J., and Zuckerman, S. (2003). Is lack of coverage a shortor long-term condition? Menlo Park, CA: Kaiser Family Foundation. Heisler, M., Wagner, T., & Piette, J. (2004a) Patient strategies to cope with high prescription medication costs: who is cutting back on necessities, increasing debt, or underusing medications? Journal of Behavioral Medicine, 28 (1), 43-51. Heisler, M., Langa, K., Eby, E., Fendrick, A., Kabeto, M., & Piette, J. (2004b). The health effects of restricting medication use because of cost. Medical Care, 42, 626-634. 130

PAGE 131

Heisler, M., Wagner, T., & Piette, J. (2004c). Clinician identification of chronically ill patients who have problems paying for prescription medications. American Journal of Medicine, 116 753-758. Hellerstein, J. (1998). The im portance of the physician in the generic versus trade-name prescription decision. The Rand Journal of Economics 29(1), 108-136. Hotchkiss, B., Pearson, C., & Lisitano, R. (1998). Pharmacy coordinat ion of an indigent care program in a psychiatric facility. American Journal of Health-System Pharmacy, 55 1293-1296. Janz, N. K., & Becker, M. H. (1984). T he Health Belief Model: a decade later. Health Education Quarterly, 11 1-47. Jin, J., Sklar, G., Min Sen Oh, V., & Chuen Li, S. (2008). Factors affecting therapeutic compliance: a review from the patients perspective. Therapeutics and Clinical Risk Management, 4 (1), 269-286. Kaiser Family Foundation (2008). The uninsured: a primer Retrieved August 27, 2009 from http://www.kff.org/unins ured/upload/7451-04.pdf Kennedy, J. & Erb, C. (2002) Prescription noncompliance due to cost among adults with disabilities in the United States. American Journal of Public Health 92, 1120-1124. Kennedy, J., Coyne, J., Sclar, D. ( 2004). Drug affordability and prescription noncompliance in the United States: 1997-2002. Clinical Therapeutics, 26 (4), 607-614. Kennedy, J., & Morgan, S. (2006). A cross-national stud y of prescription nonadherence due to cost: data from the Joint Canada-United States Survey of Health. Clinical Therapeutics, 28 (8), 1217-1224. Korn, L., Reichert S., Simon, T., & Halm, E. (2003). Improving physicians knowledge of the costs of common medications and willingness to consider costs when prescribing. Journal of General Internal Medicine, 18 31-37. Lincoln, Y. & Guba, E. (1985). Naturalistic Inquiry New York: Sage. McCorry, D., Marson, T. & Jacoby, A. (2009) Understanding routine antiepileptic drug decisions: a qualitative analysis of patients accounts of hospital consultations. Epilepsy & Behavior, 14(1), 210-214. Morse, J. & Richards, L. (2002). Read me first for a users guide to qualitative methods Thousand Oaks, CA: Sage. 131

PAGE 132

Mott, D. & Cline, R. (2002). Ex ploring generic drug use behavior. Medical Care, 40, 662674. Mounts, V., Ringenberg, D., Rhees, K., & Partridge, C. (2005). Implementation of a patient medication assistance program in a community pharmacy setting. Journal of the American Pharmaceutical Association, 45, 76-81. Munro, S., Lewin, S., Swart, T., & Volmink, J. (2007). A review of health behavior theories: how useful are these for devel oping interventions to promote long-term medication adherence for TB and HIV/AIDS. BioMed Central Public Health, 7, 104. Piette, J., Heisler, M., & Wagner, T. (2004a). Cost-related medication underuse among chronically ill adults: the treatments people fo rgo, how often, and who is at risk. American Journal of Public Health, 94 (10), 1782-1787. Piette, J., Heisler, M., & Wagner, T. (2004b). Pr oblems paying out-of-pocket medication costs among older adu lts with diabetes. Diabetes Care, 27 384-391. Piette, J. & Heisler, M. (2004c). Probl ems due to medication costs among VA and nonVA patients with chronic illnesses. The American Journal of Managed Care, 10(part 2) 861-868. Piette, J., Heisler, M., & Wagner, T. (2004d). Cost-relat ed medication underuse: do patients with chronic illnesses tell their doctors? Archives of Internal Medicine, 164 1749-1755. Piette, J., Heisler, M., Horne, R., & Alexander, G. (2006a). A conceptually based approach to understanding chronically ill pati ents responses to medication cost pressures. Social Science and Medicine, 62 846-857 Piette, J., Heisler, M., & Wagner, T. (2006b). Medication c haracteristics beyond cost alone influence decisions to underuse pharmacotherapy in response to financial pressures. Journal of Clinical Epidemiology, 59 739-746. Pound, P., Britten, N., Morgan, M., Yardley, L., Pope, C., Daker-White, G., & Campbell, R. (2005). Resisting medici nes: a synthesis of qualitative studies of medicine taking. Social Science & Medicine, 61 133-155. Prutting, S., Cerveny, J., MacFar lane, L., & Wiley, M. (1998). An interdisciplinary effort to help patients with limited prescripti on drug benefits afford their medication. Southern Medical Journal, 91 815-820. Reed, M. & Hargraves, L. (2003). Prescription drug access among working-age Americans Center for Studying Health System Change, Washington, DC. Retrieved July 14, 2006, from http://hschange.org/CONTENT/637/ 132

PAGE 133

Reichert, S., Simon, T., & Halm, E. (2000) Physicians attitude about prescribing and knowledge of the costs of common medication. Archives of Internal Medicine, 160 2799-2803. Rosenstock, I. M. (1966). Why people us health services. Milbank Memorial Fund Quarterly, 44 94-124. Safran, D., Neuman, P., Schoen, C., Montgomery, J., Li, W ., Wilson, I., Kitchman, M., Bowen, A., & Rogers, W. (2002). Prescription drug co verage and seniors: how well are states closing the gap? Health Affairs, W253-268. Sleath, B., Thorpe, J., Landerman, L., Doyle, M., & Clipp, E. ( 2004). Medication use among black and white caregivers of ol der male veterans with dementia. The American Journal of Geriatric Pharmacotherapy, 2 (2), 133-140. Smith, S. & Kirking, G. (1999). Access and use of medications in HIV disease. Health Services Research, 34 (1), 123-144. Soumerai S., Ross-Degnan, D., Avorn, J., McLaughlin, T., & Choodnovskiy, I. (1991). Effects of Medicaid drug-payment limit s on admission to hospitals and nursing homes. New England Journal of Medicine, 325 1072-1077. Soumerai, S., Pierre-Jacques, M., Zhang, F., Ross-Degnan, D., Adams, A., Gurwitz, J., Adler, G., Safran, D. (2006). Cost -related medication nonadherence among elderly and disabled Medicare beneficiaries. Archives of Internal Medicine, 166 1829-1835. Stack, R., Elliott, R., Noyce, P., & Bundy, C. (2008). A qualitat ive exploration of multiple medicines beliefs in co-morbid di abetes and cardiovascular disease. Diabetic Medicine, 25 1204-1210. Steinman, M., Sands, L., & Covi nsky, K. (2001). Self-restriction of medications due to cost in seniors without prescription drug coverage. Journal of General Internal Medicine, 16 793-799. Stockwell, D., Madhavan, S., Cohen, H., Gi bson, G., & Alderman, M. (1994). The determinants of hypertension awareness, treatment, and control in an insured population. American Journal of Public Health, 84 1768-1774. Strauss, A. L. (1987). Qualitative analysis for social scientists New York: Cambridge University Press. Strauss, A. L. & Corbin, J. (1990). Basics of qualitative research( 1st ed .). Thousand Oaks, CA: Sage. 133

PAGE 134

134 Strauss, A. L. & Corbin, J. (1998). Basics of qualitative re search: Techniques and procedures for developing grounded theory (2nd ed.). Thousand Oaks, CA: Sage. Suh, D. (1999). Trends of generic substitution in community pharmacies. Pharmacy World & Science 21(6), 260-265. Tamblyn, R., Laprise, R., Hanley, J., Abraham owicz, M., Scott, S. Mayo, N., Grad, R., Latimer, E., Perrault, R., McLeod, P., H uang, A., Larochelle, P. & Mallet, L. (2001). Adverse events associated with prescription drug cost-sharing among poor and elderly persons. Journal of the American Medical Association, 285 421-429. Viswanathan, H., & Lambert, B. (2005). An inquiry into medication meanings, illness, medication use, and the transformative pot ential of chronic il lness among African Americans with hypertension. Research in Social and Administrative Pharmacy, 1 (1), 21-39. Wazana, A. (2000) Gifts to physicians from the pharmaceutical industry. Journal of the American Medical Association 283 (20), 2655-2658. Weiner, S., Dischler, J., & Horvitz, C. (2001). Beyond pharmaceutical manufacturer assistance: broadening the scope of an indigent drug program. American Journal of Health-System Pharmacy, 58, 146-150.

PAGE 135

BIOGRAPHICAL SKETCH Prior to pursuing doctoral education in the Department of Pharmaceutical Outcomes and Policy at the University of Flor ida, Mark A. Allen attended the University of California at Berkeley, ear ning a bachelors degree in mat hematics. After a year of teaching math at a public junior high sc hool, Mark joined a start-up engineering company as their original technical specialis t; during his 11 year tenure, he established himself as a leading expert in the field of gear design and was awar ded three patents. Mark then returned to graduat e school and earned a masters degree in sociology at the University of Florida. Marks career goals and research interests include healthcare policy and the design, implementation, and evaluation of programs aimed at increasing access to medication in general. 135