«Exploring risk factors of non-adherence to immunosuppressive medication in kidney transplant recipients: improving methodology & reorienting research ...»
Exploring risk factors of non-adherence
to immunosuppressive medication in
kidney transplant recipients: improving
methodology & reorienting research goals
Erlangung der Würde eines Doktors der Philosophie
der Universität Basel
aus Waregem, Belgien
Faculty responsible: Prof. Dr. M. Tanner, Swiss Tropical Institute, Basel Thesis advisor: Prof. Dr. S. De Geest, Institute of Nursing Science, University of Basel Co-referee: Prof. Dr. M. T. Nolen, School of Nursing, Johns Hopkins University External expert: Prof. Dr. K. Dracup, School of Nursing, University of California at San Fransisco Genehmigt von der Philosophisch-Naturwissenschaftlichen Fakultät auf Antrag von Prof. Dr. M. Tanner, Prof. Dr. S. De Geest, Prof. Dr. M. T. Nolen, und Prof. Dr. K.
Basel, den 04.07.2006 Prof. Dr. Wirz Table of contents
TABLE OF CONTENTS
1. INTRODUCTION 7
1.1. ADHERENCE 7
1.2. ADHERENCE IN TRANSPLANTATION 7
1.3. MEASUREMENT OF NON-ADHERENCE 8
1.4. OUTLINE OF THE RESEARCH PROGRAM 9
2. PREVALENCE, CONSEQUENCES, AND DETERMINANTS OF NON-ADHERENCE IN ADULT
RENAL TRANSPLANT PATIENTS, A LITERATURE REVIEW 13
2.1. INTRODUCTION 14
2.2. THE BEHAVIORAL DIMENSION OF KIDNEY TRANSPLANTATION 14
2.3. PREVALENCE OF NON-ADHERENCE 15
2.4. CONSEQUENCES OF NON-ADHERENCE 15 2.4.1. Clinical consequences 18 2.4.2. Economic consequences 25
2.5. DETERMINANTS OF NON-ADHERENCE
ACKNOWLEDGEMENTSI would like to express my gratitude to the many people who supported me in completing this dissertation. My gratitude goes in the first place to my supervisor, Prof.
Dr. Sabina De Geest. I could not have imagined a better advisor and mentor for my PhD. We already fruitfully collaborated since my days as a Master’s student at the K.U.Leuven in Belgium. This collaboration reaches a provisional height with the finalization of this dissertation. I also thank Prof. Dr. Marcel Tanner of the Swiss Tropical Institute for accepting me as a doctoral student in the PhD program of epidemiology. I am also indebted to Prof. Dr. Marie Nolan of Johns Hopkins University (Baltimore, USA) and Prof. Dr. Kathy Dracup of UCSF (San Francisco, USA) for their willingness to be part of my dissertation committee. This work would not have been possible without the most valued collaboration with the Renal Transplant Program of the University Hospital of Basel and the Kantonsspital of Aarau. I therefore thank wholeheartedly Prof Dr. J. Steiger, Prof. Dr. Andreas Bock, Dr. Michael Dickenmann, Nicole Thannberger, Suzanne Köfer, and Dr. Stefan Schaub and other team members for their support during my dissertation. Thanks also to the Swiss National Science Foundation who funded the SMART study of which this dissertation is a part.
Special expressions of appreciation go to the members of our Leuven-Basel Compliance Research Group led by Prof. Dr. S. De Geest. Sincere thanks to Petra Schäfer-Keller, Dr. Fabienne Dobbels, Ariane Desmyttere, and Gerda Drent, for the stimulating discussions of my research findings and their peer support over the past years. Sincere gratitude also goes to Dr. Jim Young of the BICE who advised me on statistical issues while he was in Basel or afterwards from New Zealand. I am also indebted to Prof. Dr.
I. Abraham of Matrix45 for his help in editing the final sections of my dissertation.
Ein besonderer Dank gilt den vielen Kollegen und Kolleginnen des Instituts für Pflegewissenschaft der Universität Basel und der Abteilung Klinische Pflegewissenschaft des Universitätsspitals Basel. Die Zusammenarbeit mit ihnen hat mir viel Freude bereitet. Ich hoffe dass meinen Umzug von Leuven nach Basel dem Institut von Nutzen gewesen ist und in die Zukunft noch immer sein wird. Denn die Entwicklung der Pflegewissenschaft an der Universität Basel liegt mir am Herzen.
Ik wens ook mijn vroegere collega’s aan de universiteit van Leuven met wie ik nauw heb samengewerkt en samenwerk welgemeend te bedanken: Prof. Dr. Bernadette Dierckx de Casterlé, Prof. Dr. Koen Milisen, Prof. Dr. Philip Moons, Els Steeman, Prof.
Dr. Geert Verbeke, Prof. Dr. Emmanuel Lesaffre, Stephen Fieuws, Dr. Irina Cleemput en Prof. Dr. Frank Buntinx. De kansen, ondersteuning en inspiratie die ik van de Leuvens collega’s heb gekregen als assistent aan de KU-Leuven hebben me een stabiele basis gegeven om aan dit doctoraat te beginnen. Dank ook aan de overige Leuvense collega’s waarmee ik nauwe collegiale banden heb die ik steeds zeer op prijs heb gesteld.
‘k Zoe uak nog willen min liefke (x) en min familie bedanken vaneigen, plus min zuster omda ze mij g’holpen ee mee den ipmok van deesten bouk. ’t En zal veur de achterblijvers wel nie altijd gemakkelijk geweest zin da ‘k ip nen blauwe moandag al min ne kiër min teetebose gepakt ee en ‘t angezet benne noar ’t land van melk en ziëm (of koas en chokla, gelijk da je ‘t wilt). M’ en zin nog nie van alkoar vervrend en me ‘n goan dat uak nie loaten gebeuren.
“In general, the benefits of proven medical therapies are available only to patients who actually use them.” Richard Kravitz Introduction
1.1. Adherence Chronic diseases, defined as diseases with a long, indefinite duration and little prospect of immediate change, are considered to be a major future health care challenge 1.
Having a chronic disease often implies life-long therapy in order to prevent or delay progression of the disease. Obviously, the effectiveness of therapies depends on patients’ level of adherence. Adherence to the therapy is therefore a cornerstone requirement of successful chronic illness management 2.
Adherence (also called compliance or concordance) can be defined as « the extent to which a person’s behaviour – taking medication, following a diet, and/or executing lifestyle changes, corresponds with agreed recommendations from a health care provider » 3. It « is a behavioral process, strongly influenced by the environment in which the patient lives, including the health care practices and systems. Adherence assumes that a patient has the knowledge, motivation, skills and resources required to follow the recommendations of a health care professional.» 4
1.2. Adherence in transplantation Solid organ transplant recipients are chronically ill patients, because the transplantation did not fully eliminate the need for medical treatment. Recipients have to adhere to a life long medication regimen that prevents the immune system from rejecting the transplant and manages emerging co-morbidities 5 6 7. Reduceing the risk for rejection requires a high degree of adherence to the prescribed immunosuppressive. However, despite the dangers related to imperfect adherence, previous research has shown that a substantial proportion of solid organ recipients fail to take their immunosuppressives as prescribed. An estimated 20 to 25% of the adult heart, liver and renal transplant patients are non-adherent to their immunosuppressive therapy 7. Non-adherence is expected to cause 20% to 90% of late acute rejections and 16% to 23% of graft losses in solid organ transplant recipients 7. Furthermore, the non-adherence problem is expected to be one of the causes for yet unexplained but observed stagnation in long-term kidney graft survival 8. Economic consequences are a higher cost per quality adjusted life year 9.
Improving outcomes in solid organ transplantation is considered to be one of the main goals in transplant research and clinical management of transplant patients for the upcoming years 10. A possible strategy to reach that goal is to enhance adherence in patients. A prerequisite to enhancing adherence is that patients at risk for nonadherence can be identified, which implies that studies are needed that unveil risk factors of adherence behavior. Risk factors for non-adherence are manifold. In 2003, the WHO published a taxonomy of risk factors of non-adherence (see figure) 3. Risk factors in this model can be socio-economic, therapy-, patient-, condition-, and health care system- or health care worker-related. Adherence research has primarily focused on socio-economic, patient-, condition- and treatment-related factors. The WHO framework aims at directing attention to risk factors from all adherence-determining dimensions including health care related factors 3. As the main aim of this research program is to investigate risk factors for non-adherence to immunosuppressive drugs Introduction in kidney transplant patients, the WHO model will be used as a guiding framework in this dissertation.
Fig. 3. WHO framework: the five dimensions of non-adherence
1.3. Measurement of non-adherence A prerequisite for quantitatively investigating adherence behavior is that adherence can be assessed in a valid way. Unbiased methods are needed that capture non-adherence in its sub-clinical stage (i.e. before recurrent rejections or graft loss suggest nonadherent behavior). Several measurement methods exist, recently concisely discussed in a review article published in the New England Journal of Medicine 11. Measurement methods relevant for solid organ transplantation can be divided into two groups: direct or indirect measures.
Direct measures, i.e. observation of medication intake and biological assay of drug levels or drug metabolites in the blood or urine, permit examining actual drug ingestion. Observation is a very labor-intensive method, not feasible in non-clinical settings, and thus not very common in transplantation research. Blood assay on the other hand is used more frequently in routine clinical practice to assess levels of immunosuppressive. A disadvantage of assay is that it does not capture intake dynamics and that “white coat-compliance”, referring to patients taking their medication before a clinic visit in which a blood sample will be taken, may bias the results.
Indirect measures do not prove actual intake of the medication; rather, they estimate how much patients could have ingested based on information coming from patient selfreports or diaries, collateral reports from family members or clinicians, rates of prescription refills, pill counts, or electronic monitoring (EM). The advantage of most of these indirect methods (except for EM) is that they are relatively easy to use, often at the expense of a lower sensitivity or validity 11 12. This, however, does not apply to EM. EM is a technologically advanced method, which relies on microprocessor equipped pill packages or bottles that register their opening times to assess adherence behavior. EM is today’s most sensitive adherence measure 15. It measures adherence Introduction with great resolution, and unveils not only the taking dimension but also the temporal dynamics of medication taking. For research purposes, EM became the assessment method of choice over the last years. This does not mean, however, that EM is the perfect method. Despite its nice diagnostic properties, many researchers are reluctant to declare EM the gold standard method in adherence research, as EM’s alleged superiority has not really been thoroughly substantiated in research yet 11. In an recent editorial contribution elaborating on the validity of non-adherence measurement methods, DiMatteo, an authority in the area of adherence, therefore included EM among the methods that need to be scrutinized in future research: “To obtain a firmer grip on the best measurement methodologies, it will be necessary to compile and assess the value, appropriateness, reliability, and validity of a wide range of possible measurement strategies. […] Our findings about the prevalence, correlates, and consequences of nonadherence will always be tied to the methods we use, so we must understand them better.” “Research is needed that carefully examines measurement and methodology issues, including numerous adherence measurement strategies such as […] electronic monitoring.” 16 Because the main study of this dissertation will use EM as a measurement method for non-adherence, we will comply with DiMatteo’s call by including a section on validity of EM measurement.
1.4. Outline of the research program This dissertation consists of four chapters, of which the first two prepare the reader for the two main risk factor studies described in the last two chapters.
The literature review presented in chapter one summarizes the evidence on the prevalence, determinants, clinical and economic consequences of non-adherence with immunosuppressive drugs in renal transplant patients. A literature search, which yielded 38 articles was used to calculate 1) weighted mean prevalences of nonadherence and 2) weighted mean prevalences of rejection episodes or graft losses that can be attributed to non-adherence. In addition, economic consequences of nonadherence will be reviewed. Investigated risk factors of non-adherence will be summarized using the WHO framework, and suggestions are made for further research.
Chapter two summarizes existing knowledge about the validity of electronic medication monitoring. A framework systematizing sources of bias in EM assessment is presented.
The framework discerns internal and external assumptions underlying unbiased EM measurement. Internal validity assumptions presuppose (1) correct functioning of the EM equipment, (2) correspondance between EM-bottle openings and actual intake of the prescribed doses, and (3) absense of influence of EM on a patient’s normal adherence behavior. External validity refers to EM biasing the representativeness of the sample. The four validity assumptions were tested using data from the Supporting Medication Adherence in Renal Transplantation (SMART) study, which included 250 adult renal transplant patients whose adherence to immunosuppressive drugs was measured over a 3-month period by EM 17.