CANCER PATIENT DECISION-MAKING AND RELATIONAL AUTONOMY RELATED TO CLINICAL TRIAL PARTICIPATION by JENNIFER ANN HARRISON BELL Hon. B.A., University of Toronto, 2004 M.A., McGill University, 2007 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Interdisciplinary Studies) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2014 © Jennifer Ann Harrison Bell, 2014 ii Abstract Clinical trials (CTs) in cancer care play an essential role in advancing knowledge and improving patient care. Low CT enrolment, however, threatens this field of science and may prevent people with cancer from benefiting from the latest therapeutic interventions. The aim of this research was to explore cancer patients’ CT decision-making process and how they exercise relational autonomy within this process. Relational autonomy acknowledges patients are situated within a larger relational and socio-political context that is characterized by inherent power differentials and social inequities that may influence CT decisions. The objectives of this research were to: a) understand cancer patients’ decision-making process related to CT participation and how they exercise relational autonomy within this process; b) examine the personal, social, and system influences on cancer patients’ decision-making process related to CT participation and their ability to exercise their relational autonomy; and c) identify the best practices utilized by CT personnel that support cancer patients in exercising their relational autonomy in the CT decision-making process. Two different yet complementary methodologies, interpretive description and grounded theory, guided in-depth interviews with 12 CT personnel, 40 breast and prostate cancer patients, and 11 support persons to address the objectives. Three major themes were uncovered that impact cancer patients’ decision-making process and ability to exercise their relational autonomy: (1) power differentials between patients and physicians, (2) therapeutic misconception, and (3) inequities in access to CTs. The overarching construct ‘no wo/man is an island’ captured patients’ CT decision-making process and experiences of autonomy, including the relational complexity of CT decisions and the key influences on this process. Results from this research highlight how CT iii decision-making is a complex endeavor comprised of key phases and processes that are not only personally but also socially and structurally located. Practice implications of this research include targeted education for CT personnel and patients to equalize power relationships within CT recruitment. In addition, standardization of cancer drug approval, better monitoring and follow-up cancer care, and a more accessible and quality healthcare system can address structural barriers in order to support patients’ relational autonomy within the context of CTs. iv Preface This dissertation is based on a national collaboration that received funding from the Canadian Institutes of Health Research (CIHR) open operating grant competition (2011-2013). I initially conceived the research project design with significant support from Dr. Lynda Balneaves. Dr. Balneaves is listed as the principal investigator of the project since CIHR policies prevent students from being listed as the principal investigator. Dr. Anita Ho, co-applicant, was involved in the data analysis, drawing on her expertise in trust and autonomy in medical decision making. Dr. Kim Chi, Medical Director of the BC Cancer Agency (BCCA) Clinical Trials Department, and Dr. Karen Gelmon, Scientist at the BCCA, were co-applicants. They assisted with recruitment of the interview participants and offered clinical insights into the final theory. Dr. Harriet Richardson, collaborator, is a Project Coordinator with the NCIC Clinical Trials Group (CTG) and an Assistant Professor at Queen’s University. She is an important liaison with the NCIC CTG, which will facilitate future dissemination activities based on this research. Two graduate research assistants, Ms. Genevieve Breau and Ms. Maryam Matean, assisted in the recruitment of participants, data collection, and data coding. I initially conceived the study design with the support of Dr. Balneaves. I took a leadership role in collecting the data, performing the analysis of the data, interpreting the findings from data analysis, and writing the dissertation chapters. Dr. Balneaves was instrumental in advising on the writing of the research document. She was the initial reviewer for the drafts of each chapter of this dissertation and provided significant guidance and feedback regarding the data interpretation and analysis. My doctoral committee members, Drs. Anita Ho and Patricia Rodney, provided ongoing guidance and input throughout the v research process. They also provided significant feedback on each of the chapters included in this dissertation. Sarah Brumwell provided editorial service in accordance with the Editors’ Association of Canada Guidelines for Ethical Editing of Theses/Dissertations. I composed Chapter Two, the systematic literature review, with significant edits and feedback from Dr. Balneaves in order to prepare the document for submission to the CIHR operating grant competition. A version of this chapter has been submitted to a peer-reviewed journal. Dr. Lynda Balneaves was instrumental in shaping this chapter for journal submission and is included as a co-author on the submitted manuscript. I composed Chapter Three, methodology, with significant edits and feedback from Dr. Balneaves in order to prepare the document for submission to the CIHR operating grant competition. Grant members, Dr. Kim Chi and Dr. Karen Gelmon provided guidance regarding participant recruitment methods. Approval from the University of British Columbia (UBC) Behavioural Research Ethics Board was obtained for the research chapters, Chapters Four and Five, involving interviews with human subjects (certificate number: H10-02904). vi Table of Contents Abstract .................................................................................................................................... ii!Preface ..................................................................................................................................... iv!Table of Contents ................................................................................................................... vi!List of Tables .......................................................................................................................... xi!List of Figures ........................................................................................................................ xii!Acknowledgements .............................................................................................................. xiii!Dedication .............................................................................................................................. xv!Introduction ............................................................................................................................. 1!Background: The Canadian Healthcare System ........................................................................... 1!Cancer Incidence ............................................................................................................................ 2!Cancer Clinical Trials .................................................................................................................... 4!Critique and Need for Current Research ....................................................................................... 7!Structure of the Dissertation ......................................................................................................... 10!Chapter One: Relational Autonomy as Theoretical Lens ................................................. 12!Traditional Accounts of Autonomy .............................................................................................. 12!Feminist Theories of Autonomy .................................................................................................... 14!Relational Autonomy and Gender ............................................................................................... 17!Spectrum of Relational Autonomy Theories ................................................................................ 18!Procedural Relational Autonomy ................................................................................................. 18!Substantive Relational Autonomy ............................................................................................... 20!Hybrid Account of Relational Autonomy .................................................................................... 21!Relational Autonomy as Guiding Theoretical Framework .......................................................... 23! vii Chapter Two: Integrative Literature Review of Cancer Patient Decision Making About Clinical Trials with Implications for Patients' Relational Autonomy ............................. 27!Methods ........................................................................................................................................... 28!Results .............................................................................................................................................. 29!Summary of Systematic Reviews ................................................................................................ 29!Summary of Quantitative Literature ............................................................................................ 31!Summary of Qualitative Literature .............................................................................................. 42!Discussion ........................................................................................................................................ 49!Relational Autonomy and Clinical Trial Recruitment ................................................................. 49!Conclusions ..................................................................................................................................... 50!Chapter Three: Methodology .............................................................................................. 73!Study Design ................................................................................................................................... 73!Interpretive Description ............................................................................................................... 74!Grounded Theory ......................................................................................................................... 75!Theoretical Framework and Use of the Literature ..................................................................... 76!Study Population ............................................................................................................................ 78!Recruitment Strategies ................................................................................................................... 79!Sampling Strategies ........................................................................................................................ 81!Interviews ........................................................................................................................................ 82!Data Analysis .................................................................................................................................. 85!Validity and Reliability .................................................................................................................. 91!Ethical Considerations ................................................................................................................... 95!Chapter Four: Cancer Patient Accrual to Clinical Trials: A Qualitative Analysis of Clinical Trial Personnel Perspectives .................................................................................. 98!Findings ........................................................................................................................................... 99! viii Personal Influences on Patients’ Decision-making ...................................................................... 99!Social Influences on Patients’ Decision-making ....................................................................... 106!Structural Influences on Patients’ Decision-making .................................................................. 116!Recommendations ........................................................................................................................ 122!Provision of Clinical Trial Decision-making Support ............................................................... 123!Mobilization of Social Resources .............................................................................................. 124!Attention to Structural Barriers .................................................................................................. 125!Summary of Findings ................................................................................................................... 126!Power Differences Between Patients and Physicians ................................................................ 127!Equitable Access to Clinical Trials ............................................................................................ 130!Chapter Five: "No Wo/Man is an Island"—Cancer Patients' Experience of Autonomy Related to Their Decision-Making Process about Clinical Trial Participation ............ 132!Introduction to the Theory ......................................................................................................... 133!No Wo/Man is an Island ............................................................................................................ 133!Phase I: Prior to the Clinical Trial Offer - Predisposing Factors ............................................ 139!Phase II: Trial Offer ..................................................................................................................... 142!Phase III: Decision-making ......................................................................................................... 147!Seeking Information ................................................................................................................... 147!Establishing Credibility ............................................................................................................. 153!Weighing Pros and Cons ............................................................................................................ 157!Maintaining Social Relationships .............................................................................................. 173!Influencing Factors .................................................................................................................... 180!Risk Tolerance and Choice Perception ...................................................................................... 186!CT Decisions ................................................................................................................................. 188!Characteristics of Trial Acceptors .............................................................................................. 188! ix Characteristics of Trial Decliners .............................................................................................. 189!Trial Withdrawers ...................................................................................................................... 189!Recommendations ........................................................................................................................ 191!Conclusion ..................................................................................................................................... 192!Chapter Six: Discussion ...................................................................................................... 198!Relational Autonomy Analysis of Cancer Patients’ Decision-Making Process About Clinical Trials .............................................................................................................................................. 198!Revisiting Relational Autonomy Theory ................................................................................... 198!Theme A: Justice and Equitable Access .................................................................................... 199!Theme B. Therapeutic Misconception ....................................................................................... 204!Theme C. Social Influence ......................................................................................................... 212!Theme D: Gender and Decision Making ................................................................................... 223!Clinical Implications .................................................................................................................... 226!Education for CT Personnel ....................................................................................................... 228!Limitations ................................................................................................................................. 235!Conclusion ..................................................................................................................................... 237!Future Research ......................................................................................................................... 237!Bibliography ........................................................................................................................ 240!Appendix I: Recruitment Poster ........................................................................................ 259!Appendix II: Permission to Contact Patients ................................................................... 260!Appendix III: Patient Information Package ..................................................................... 261!Appendix IV: Letters of Explanation ................................................................................ 262!Appendix V: Consent Forms ............................................................................................. 265! x Appendix VI: Demographic Information (Patients) ........................................................ 271!Appendix VII: Interview Guides ....................................................................................... 273! xi List of Tables Table 1.1 – Clinical Trial Phases .............................................................................................. 5!Table 2.1 – Systematic Literature Reviews Reviewed ........................................................... 53 Table 2.2 – Quantitative Literature Reviewed…………………………………………….....56 Table 2.3 – Qualitative Literature Reviewed………………………………………………...71 Table 5.1 – Reported Patient Characteristics…………………………………………….....133 xii List of Figures Figure 5.1 – "No Wo/Man is an Island" ............................................................................... 138! xiii Acknowledgements This dissertation would not have been possible without the support and guidance from several people and organizations. First, I would like to thank my dissertation committee. In particular, I am sincerely grateful for and appreciative of my doctoral supervisor, Dr. Lynda Balneaves. Her devoted mentoring taught me the “ins and outs” of successful grant writing. Her thorough reviews of each chapter of my work and helpful discussions about the research kept me on track and helped me become a better writer and scholar. I hope that I may have the opportunity to be as fantastic a mentor some day to others as she has been to me. I would also like to thank my doctoral committee members, Drs. Paddy Rodney and Anita Ho. Both reviewed several iterations of each chapter and provided helpful comments and feedback. I am grateful for the enthusiasm they exhibited towards the project and for their insights that improved my work. Second, I am sincerely grateful for the support of co-investigators on this project, Drs. Kim Chi and Karen Gelmon. This research would not have been possible without their commitment to the project and help with participant recruitment. A special thank you also to all the clinical trial personnel, patients and support persons who shared their experiences with me. This research would not have been possible without their willingness to share their insights and perspectives. Thank you also to graduate research assistants Genny Breau and Maryam Matean, who helped with participant recruitment and the initial coding of interview transcripts. Third, thank you to my friends and family for supporting me through this venture. In particular, my Wednesday night potluck group who kept me sane over the course of my xiv program, my best girlfriends with whom I went rock climbing with most Fridays to burn off energy and rejuvenate, and several friends/colleagues who were also going through the dissertation process and provided practical “tips and tricks,” shared resources with me and provided emotional support. Thank you to: Ashley and Mauro Castellarin, Angela Hawk, Stefanie Goertzen, Chelsea Branch, Tanya Martens, Maria MacDougall, Bryn Lander, Kirsten Halverson, Nathalie Vigouroux, Mandy Catron, Carmen Merkel, Alicia Semaka, Deborah Tang, Leah Lambert, and Daniel Buchman. Finally, my deepest and most heartfelt thank you to my partner, James Anderson, for believing in me and continuing to encourage me to trust my instincts and develop my skills. Last, I have received generous funding to support my doctoral training from the following organizations: Canadian Institutes for Health Research, Psychosocial Oncology Research Training Program, Ethics of Health Research and Policy Training Program and the University of British Columbia. I am grateful for this financial support, which has allowed me to pursue this doctoral degree. xv Dedication To my family 1 Introduction The overarching problem that this dissertation addresses is the lack of knowledge concerning how patients with cancer experience the clinical trial (CT) decision-making process, with particular attention paid to the personal, social and structural influences on this process. In this introduction, I provide a brief overview of essential background information for this research, outline the research objectives, and introduce the overall structure of the dissertation. Background: The Canadian Healthcare System Canada’s healthcare system is provincially funded and delivered with support from the federal government in terms of cash and tax transfer payments. In order to receive federal payment for healthcare costs, provinces and territories must demonstrate that their health plans are aligned with the principles of the Canada Health Act (1985). The principles include public administration, comprehensiveness, universality, portability, and accessibility. Public administration requires that health plans be delivered on a not-for-profit basis. Comprehensiveness means all residents are covered for “medically necessary” health services. Universality requires public insurance plans to cover all residents according to the same terms and conditions. Portability allows residents to be treated by their public health plans wherever they are across Canada. Finally, accessibility requires access to insured services without direct or indirect financial charge or discrimination based on age, health status, or ability to pay (Council of Canadians 2010). These conditions are meant to ensure universal access to quality healthcare for all Canadians regardless of financial or other barriers. It also allows the federal government to have some control over the provincial administration of healthcare (Council of Canadians 2010). 2 In 2002, Commissioner Roy Romanow published a monumental report on the future of public healthcare in Canada (Commission on the Future of Health Care in Canada & Romanow 2002). In the report, Romanow identified serious, countrywide disparities in access to care that challenge the values of the Canada Health Act and the long-term sustainability of the system. He also identified a fragmented prescription drug system, which has resulted in disparities of coverage across the country. Even though Pharmacare (provincial insurance programs for prescription drugs) was not considered an insured healthcare service under the original provision of the Canada Health Act, Romanow argued that drugs are increasingly required for the successful management or treatment of illness and should be integrated within the healthcare system. Now, over a decade later, disparities based on income, education, sex and geography continue to challenge the Canadian healthcare system (Asada, Yoshida & Whipp 2013), including how care is provided to those living with cancer (Penner et al. 2012). Cancer Incidence As disparities in health care are increasing in Canada, so too is the incidence of cancer. According to the Canadian Cancer Society, cancer is the leading cause of death in Canada. In total, 29% of men and 24% of women will die from this disease, with the incidence of new cancer diagnoses expected to increase across the country due to the growing and aging population. It is expected that 45% of men and 40% of women will develop cancer during their lifetimes, with breast and prostate cancer predicted to be the most common new diagnoses (Canadian Cancer Society 2012). Cancer care delivery, like other healthcare services, is a provincial responsibility, with most provinces delivering services through regional operations (Sullivan et al. 2004). 3 Timely access to cancer treatment and prevention is necessary for improved health outcomes and quality of life and to control the high incidence of the disease (Berrino et al. 2007). However, there is variation in incidence rates of cancer across Canada and within provinces due to differences in the populations, risk factors, and early detection behaviour/activities (Canadian Cancer Society 2012). Death rates also vary between regions, arguably because of inequitable access to cancer screening and treatment (Canadian Cancer Society 2012). In a recent comprehensive review of the Canadian literature, Maddison and colleagues identified inequities in access to cancer screening, radiotherapy, and end-of-life care services based on income, age, and geographic location (Maddison et al. 2011). For example, cancer patients with lower income and those who live far from a cancer centre are less likely to access treatment services, such as systemic therapy (Bairati, Jobin, Fillion, Larochelle & Vincent 2007; Maddison et al. 2011; Saint-Jacques et al. 2008). Access to cancer drugs is vitally important for medical care as new therapeutic biotechnologies and pharmacogenetics offer unprecedented control of disease, yet Pharmacare is not yet covered by the insured services provision of the Canada Health Act. Instead, each province is responsible for evaluating and deciding which drugs will be covered under their own insurance plans, resulting in discrepancies between provinces regarding the drugs that are currently approved and available to cancer patients (Standing Senate Committee on Social Affairs 2012). As a consequence, inequities exist in relation to access to cancer drugs across Canada. A recent workshop report (Richter 2010) identified the following as barriers to cancer drug access; however, it is important to note that these barriers may also exist for drugs for many other chronic conditions: • Out of pocket cost for patients who do not have insurance for drugs administered out 4 of hospital; • Private insurance policies that no longer guarantee coverage;!• Variations in coverage among provincial plans – different provinces cover different treatments;!• Timely coverage of new drugs by provincial and territorial plans. Cancer Clinical Trials CTs are used to evaluate new cancer interventions and provide the foundation of evidence-based practice. Randomized controlled trials (RCTs) are generally recognized as the “gold standard” because they are designed to minimize the risk of study bias and provide objective evidence about an intervention. When the results of many RCTs are pooled, they can produce the best available evidence about an intervention or drugs at the population level (Rogers 2004). Increasingly, CTs are becoming part of the cancer care system as a way for patients to obtain the latest and expensive drug therapies. Although not all CTs involve drugs, they are often the primary focus of this type of research.1 CTs are conducted in phases that begin after basic biomedical research on a drug has been completed in a laboratory on cell and mice models. The following table outlines the various phases of clinical research. 1 CTs are an “investigation in respect of a drug for use in humans that involves human subjects and that is intended to discover or verify the clinical, pharmacological, or pharmacodynamic effects of the drug, identify 5 Table 1.1 – Clinical Trial Phases2 Trial Phase Description Phase 0 Trials that administer a very small dose of a chemical or biologic agent or other types of interventions to a small number of people to gather preliminary information about how the agent is processed by and affects the body. Phase I Trials that test an intervention for safety in a small group of participants. Phase II Trials that administer an intervention to a larger group of participants to test its effectiveness. These trials also continue to look at the safety of interventions. Phase III Trials that compare the effectiveness of a new intervention, or new use of an existing intervention, with the current standard of care. These trials also examine how the side effects of the new intervention compare with those of the usual treatment. These trials usually involve large groups of people who are randomized to one of two trial arms. Phase IV Trials that are conducted after a drug has been approved by a governing body (e.g., Health Canada) in order to evaluate its effectiveness and long-term safety. CTs may compare an experimental intervention against a placebo. They may include randomization, where participants are assigned to arms of the study by chance, which often occurs in Phase III trials. They might also be single- or double-blind controlled studies where either the participant or the participant and the research team are blinded to which medication is being administered. These procedures are used to reduce bias in the study findings. There are two ways cancer CTs are typically sponsored in Canada. Trials are either implemented by cooperative groups or by the pharmaceutical industry. Cooperative groups 2 This table has been adapted from the National Cancer Institute’s website. Retrieved December 30 2013, from http://www.cancer.gov/cancertopics/factsheet/clinicaltrials/clinical-trials. 6 include the National Cancer Institute of Canada (NCIC) clinical trials group (CTG), a national oncology group primarily made up of academic investigators and physicians, which receives core funding from the Canadian Cancer Society Research Institute and is supported by the Canadian Cancer Society. Both cooperative and pharmaceutical sponsors are responsible for ensuring local administrative costs of trials, illustrating the complexity of the CT system in Canada. These include costs related to trial operations, such as human resources for audit, monitoring, and safety reporting, and operational and technological infrastructure, such as meeting administration, sample sharing, databases and biobanking. Costs also include payment to institutions to support CT recruitment of human subjects and related laboratory testing that are increasingly required in studies (The Standing Senate Committee on Social Affairs 2012). Between 1980 and 2012, the NCIC CTG carried out 255 Phase III trials involving more than 65,000 participants and 188 Phase I or II studies involving more than 5,000 participants (Canadian Cancer Society 2012). Cancer CTs may be conducted in physician offices, cancer centres, academic healthcare organizations, clinics and community hospitals. Patients typically gain access to cancer CTs through their physicians. However, current accrual to cancer trials is extremely low. The national average for adult patient recruitment to Canadian cancer CTs in 2009 was 7% with the provinces ranging from 2% to 11% (Canadian Cancer Research Alliance 2011). Increasingly, cancer CTs are occurring internationally due to the pressures of patient accrual. In a recent analysis of trial accrual performance among adult oncology studies in the United States (US), Cheng et al (2011) reported a total of 81.5% of trials did not achieve accrual goals within the anticipated accrual period. This resulted in trials remaining open longer, thus resulting in additional costs, or closing earlier without achieving the minimum projected 7 accrual as defined within the study protocol. Low accrual compromises the success of CTs, wastes valuable resources, and squanders an opportunity for improving patient outcomes. As such, future research must explore opportunities for increasing patient accrual to CTs, while continuing to ensure human research subjects’ safety and wellbeing throughout the research process. Critique and Need for Current Research Since the Nuremberg trials in 1947, the protection of human research subjects has been an international moral imperative (Rothman 2003). Several codes and guidelines have been developed world-wide to support the ethical conduct of clinical research (World Medical Association 1964, 1975, 1983, 1989, 1996, 2000, 2002, 2004, 2008), including the Canadian Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans, 2nd edition (TCPS 2) (Canadian Institutes of Health Research 2010) and the Belmont Report (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research 1979). Three basic ethical principles are addressed within these research ethics guidelines: respect for persons, beneficence and justice. While these principles are all relevant to the conduct of CTs, the principle of respect for persons and its application through the informed consent process is particularly relevant to CT recruitment and accrual. Informed consent procedures entail providing autonomous individuals the right to choose to participate or not in research based on their own considered judgment (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research 1979). Autonomy typically refers to individuals’ ability to freely determine and act upon their values and/or preferences. The influence of others is often seen as detrimental to autonomy because it can prevent individuals from making their own choices. Previous 8 research on cancer patients’ CT decision-making has been dominated by an individualist approach to autonomy. The focus has been on individual patients’ decisions with significant attention to the kinds of information required by patients to ensure their informed consent to trial participation (see Albrecht, Penner & Ruckdeschel 2003; Albrecht, Ruckdeschel, Riddle, Blanchard, Penner, Coovert & Quinn 2003; Loh, Butow, Brown & Boyle 2002). For example, Brown, Butow, Boyle and Tattersall (2004) found that within the informed consent process, oncologists scored poorly on communicating essential information to patients, such as explaining randomization and alternative therapy options available if the patient were to decline the trial. Also, many oncologists were found to implicitly urge patients to consider either standard care or trial participation, thus potentially comprising the voluntariness of patients’ informed consent (Brown et al. 2004). However, the field of healthcare ethics is increasingly adopting the broader notion of “relational autonomy”, which recognizes the importance of social circumstances and significant relationships on individuals’ self-determination and informed consent (Beauchamp & Childress 2012; Sherwin 1998). This relational aspect is reflected in some studies that have broadened the examination of physician-patient communication to include the influence of family members on communication and patients’ informed consent. For example, Albrecht, Penner & Ruckdeschel (2008) developed a holistic model of communication by examining the complex dynamics between physicians, patients and significant others, and the bearing these relationships had on patients’ CT decisions. Relational autonomy (as will be described more fully in Chapter One) acknowledges the inescapable interplay of social and structural forces on individuals’ decision-making processes (Sherwin 1998). Respecting persons’ relational autonomy in the context of human 9 subjects’ research thus requires attending not only to the personal aspects of cancer patients’ CT decision-making process but also the broader social and structural dimensions that impact their autonomy within this process (Sherwin 1998). This requires attending to not only familial or significant relationships influential to cancer patients’ autonomy but also to the broader socio-political environment in which these relationships are developed and sustained. Thus, the theoretical lens of relational autonomy relates well to understanding and addressing larger equity challenges of health care delivery and the structural complexity of CTs on individual cancer patients’ decision making. Research Objectives The purpose of this dissertation is to explore cancer patients’ decision-making process about CT participation through the lens of relational autonomy in order to develop new insights and practice and research recommendations about cancer patients’ CT decision-making and trial recruitment procedures. Accordingly, the objectives of this study are to: • Understand cancer patients’ decision-making process related to CT participation and how they exercise relational autonomy within this process; • Examine the personal, social, and structural influences on cancer patients’ decision-making process related to CT participation and their ability to exercise their relational autonomy; • Identify the best practices utilized by CT personnel that support cancer patients in exercising their relational autonomy in the CT decision-making process.!The overall outcome of this dissertation research is an in-depth description and theory of the nature of CT decision-making through the lens of relational autonomy. This theory will 10 provide a foundation for the development of decision support strategies that will enhance how Canadians living with cancer are approached, recruited, and enrolled in cancer CTs. Structure of the Dissertation This dissertation is comprised of six chapters, plus this Introduction. Chapter One introduces relational autonomy as the guiding theoretical framework that informs multiple aspects of the dissertation research, including the primary study design, data collection and analysis. The purpose of Chapter One is to provide an overview of relational autonomy theory, which is contrasted with more traditional accounts of autonomy. Differing feminist theories of relational autonomy are discussed and Susan Sherwin’s account is argued as most applicable to this dissertation research. Chapter Two provides a literature review in order to situate the current study within the relevant field of research. An overview of studies examining cancer patient treatment decision making as well as CT decision making is presented, since there is often overlap as trials are discussed in the treatment context. Key gaps in the literature are identified with regards to the socio-political influences on cancer patients’ decision-making processes that are important for patients’ relational autonomy. This sets the stage for the empirical chapters that follow. Chapter Three presents the methodology of the dissertation research. This dissertation uses two different but complementary qualitative methodologies: interpretive description and grounded theory. Breast and prostate cancer patients, and their support persons, were interviewed for the grounded theory study. CT personnel were interviewed for the interpretive descriptive study. Grounded theory and critical feminist analysis combined data from both studies in order to obtain perspectives relevant to personal, social, and 11 structural influences on patients’ CT decision-making processes. An overview of the study designs, participant recruitment, interview conduct, and data analysis is provided in this chapter. Chapters Four and Five of this dissertation are based on data collected through open-ended interviews with cancer patients, support persons, and CT personnel in 2011 and 2012. Chapter Four uses interpretive description to explore CT personnel perspectives (e.g., oncologists, CT nurses) about cancer patients’ experience of autonomy within the CT decision-making process and best practices for supporting cancer patients’ autonomy. Chapter Five explores breast and prostate cancer patients’ decision-making process related to CT participation in order to develop a theory “grounded” in their actual experiences. Finally, Chapter Six uses relational autonomy theory and a critical feminist analysis to further explore and analyze the socio-political findings from Chapters Four and Five. Key themes that are discussed include issues of justice and equitable access to CTs, therapeutic misconception and the dual role of the oncologist-investigator, and social influences. Issues relating to gender that may place undue pressure on patients’ relational autonomy in the context of breast and prostate cancer patients’ CT decision-making processes are explored. Practice recommendations are proposed to support cancer patients’ relational autonomy within their CT decision-making processes. Finally, the limitations of the current research are considered and future research is proposed to advance the field. 12 Chapter One: Relational Autonomy as Theoretical Lens In this dissertation, the theoretical lens of relational autonomy was used to explore the ethical dimensions of clinical trial (CT) decision making from the cancer patient’s perspective. The purpose of this chapter is to outline an autonomy theory in light of feminist notions of relational autonomy. First, the concept of autonomy for healthcare and trial decision making is examined and placed in philosophical context. This is then followed by feminist critiques and an elaboration on the unique features of a “socially-situated” (Sherwin 1998) or relational conception of autonomy. It is argued that Sherwin’s relational autonomy theory is most applicable to this dissertation. Practical ways in which the theory and associated critical feminist lens guided the dissertation are outlined. Traditional Accounts of Autonomy Western bioethics presumes the importance of autonomy as a right for individuals to determine for themselves what shall be done to their bodies (Mappes & DeGrazia 2006). In the research context, patients exercise this right by choosing whether to participate in CTs based on their own considered judgment and providing informed consent to participate (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research 1979). Autonomy to provide informed consent requires patients exhibit: a) intentionality to perform the relevant action; b) understanding of the relevant information; c) freedom from internal constraints (e.g., depression, addiction); and d) freedom from external constraints (e.g., physical barriers, coercion) (Mappes & DeGrazia 2006). Autonomous patients are thus presumed to demonstrate capability to understand and appreciate the decision at hand and be free from coercion or external influence so that their respective 13 values and choices are not unduly influenced or biased by healthcare providers or others (Tomlinson 1986). The writings of two prominent philosophers, in particular, have heavily influenced the predominant interpretation of autonomy within the healthcare context: Immanuel Kant (1724-1804) and John Stuart Mill (1806-1873).3 According to Kant, persons are those beings who are intellectually capable of deriving for themselves and acting in accordance with the rational law. Autonomy, according to some interpretations of Kant, involves self-sufficiency and the freedom of will from external considerations that allow an individual to act on the basis of rational reasons alone.4 Rational reasons are those that appeal to only universal maxims to guide action instead of reasons based on self-interest or natural inclinations. Autonomous persons are therefore rational people who are expected to only act from reason (emotional aspects of decisions are non-reasoned). They should be afforded due respect and, as such, must not be treated as mere means to an end, but as ends themselves: “so act as to treat humanity, whether in your own person or in that of any other, never as a means, but always at the same time as an end” (Kant 1964, p.96). Whereas Kant’s writings may have influenced how rationality evolved as an important criterion for autonomy, the philosophical writings of John Stuart Mill (1806-1873) may have imparted the second influence on the concept of autonomy, namely the notion of 3 Mill’s writings could be said to offer more of a political view of the boundaries of personal decision-making, whereas the moral value of autonomy is often derived from the philosophy of Immanuel Kant (1724-1804). See Thomas May (2005) for a discussion of the distinction between moral and political requirements of autonomy. Moral autonomy requires adherence to rational law. Political autonomy requires respect for decisions within the political sphere. 4 However, whether Kant’s writings provide moral justification of autonomy as self-sufficiency and independence is debated in the literature. See Secker (1999). 14 individual independence. According to Mill, autonomy, or more accurately, liberty, refers to persons’ ability to determine for themselves what constitutes the good life. The emphasis, however, is on “negative liberty” or freedom from the interference of others in order for the individual to actively shape one’s own goals, pursue one’s own projects, and flourish in the personal development of one’s own capacities and character (Mill 1974). Above all, the individual is seen as best positioned to direct oneself in the manner one sees fit, provided that he or she does not cause harm or infringe on others’ expressions of freedom to do the same. Feminist Theories of Autonomy Feminists reject the view that the influence of others always impedes self-determination and that independent and rational decision-making are ideal characteristics of the autonomous person (Barclay 2000). Instead, feminists support a relational concept of the self and corresponding theory of autonomy.5 Although feminist accounts vary, relational autonomy theorists recognize the inherent intersubjectivity of persons and that important relationships, as well as social, historical, and political institutions, significantly influence self-determination. Complete freedom from interference in self-determination is seen as unrealistic and even undesirable as it negates, for 5 Increasingly, mainstream ethicists also support a more relational concept of autonomy. The thinking of Beachamp and Childress (2012), authors of a prominent bioethics text now in its seventh edition, has evolved from recognizing autonomy as an individualistic concept in their first edition of the textbook to acknowledging its relational attributes in later editions. Despite the evolution of the concept of autonomy within the field of bioethics, health care providers still misinterpret respect for autonomy to require non-interference with patients’ rational decision making (Ells, Hunt & Chambers-Evans 2011). 15 example, the importance of social relationships in fostering persons’ development of autonomy capacities. Furthermore, the socio-political environment influences patients’ ‘rational’ choices such that those patients with more social and economic resources are more likely to be deemed rational, and thus, autonomous to make decisions (Secker 1999b). Ascriptions of rationality may be contingent upon the social and economic circumstances of the decision maker. For example, the rationality and reasonableness of a preference for discharge home may be contingent upon the ability to pay for homecare services. Two otherwise identically situated patients, one who can pay for home care and one who cannot, may be judged quite differently by healthcare providers. When the patient who can pay for these services states a preference for discharge, her decision may be deemed reasonable. By contrast, when a patient who cannot pay for these services states this preference, her decision may be deemed unreasonable. Inequitable distributions of goods and resources are expressed in assymetrical ascriptions of rationality; patients who are disadvantaged are more likely to be judged irrational on these grounds alone. This is clearly problematic. An attractive feature of relational accounts of autonomy is that these accounts attune us to the ways in which social, political, economic and other contextual factors influence both patients’ exercise of autonomy and healthcare providers’ assessment of patients’ autonomy. In general, to be relationally autonomous, persons must at least possess adequate capacities for reflecting upon, identifying, re-affirming, or revising their identity-conferring values, goals and commitments (e.g., Meyers 1989; Sherwin 1998). Autonomy is a continuous process of developing and refining these capacities through experiences and opportunities for self-determination. This account is so far consistent with traditional theories 16 except that feminists recognize social influences and the socio-political context as central to the development and realization of self-determination. The broader socio-political context inevitably influences personal decision making, including persons’ sense of what is important, their guiding values, goals, and commitments, relationships with others, and senses of identity (Mackenzie & Stoljar 2000; Nedelsky 1989; Sherwin 1998). Socio-political contexts and relationships can facilitate autonomy by providing access to quality education, healthcare, and financial resources that allow persons to develop or exercise their autonomous capacities. Socioeconomic disparities (e.g., lack of education, income, social resources), asymmetries of power, status and knowledge, disparities in healthcare access, and the undue or coercive influence of family and friends, however, can impede persons’ autonomy. Mackenzie and Stoljar (2000) identify three distinct but inter-related levels in which socio-political context can influence autonomy. These include: • At the level of belief and desire formation; • In the development of capacities and competencies necessary for autonomy; • In limiting a person’s ability to act on autonomous desires or make autonomous choices (p. 21). !In order to redress oppression and unjust socioeconomic conditions that disempower some groups in exercising their autonomy, relational autonomy theories include a strong commitment to social justice. Social justice refers to the degree to which society supports the conditions necessary for all persons to exercise capacities, express experiences, and participate in determining actions while simultaneously respecting group differences without oppression (Young 1990). In healthcare, social justice may require improving the social and 17 material conditions for disadvantaged groups that influence their health status and ability to exercise capacities. Feminists employ a critical stance in order to identify instances of power differences and social inequities that may further undermine social justice and relational autonomy so that they may be addressed. Critical feminism recognizes that persons are embedded in and emerge from social relations that are organized by power dimensions rooted in inequalities experienced by racial, cultural, ethnic, class, and gender groups (Qin 2004). In particular, critical feminism allows the impact of gender socialization and its intersection with other inequalities on relational autonomy to be made visible and addressed. The following section details the relationship between relational autonomy and gender. Relational Autonomy and Gender Sherwin and other feminist theorists argue that the traditional theoretical account of autonomy is inherently masculine because it upholds rationality, independence, and self-reliance, as ideal characteristics of the autonomous person (Mackenzie & Stoljar 2000; Sherwin 1998). The result is to deny the autonomy of those who do not meet the ideal, especially persons who are dependent on others because of illness (Sherwin 1998). Gender and its interactions with other variables have important implications for autonomy from a relational standpoint. For example, Sherwin and other feminists (Dodds 2000; Secker 1999b), argue that the rationality of women and other oppressed groups is more likely to be questioned, and they are viewed as lacking autonomy because of sexism, ageism, and other stereotypes. Further, women are more likely to be viewed as embodying feminine characteristics, such as emotion and dependency, which are in direct opposition to traditional conceptions of autonomy. When gender intersects with class, women who have fewer 18 socioeconomic resources may be less likely to develop capacities for autonomy and more likely to be perceived as lacking competence (see, for example, Secker 1999a). In comparison, men who otherwise would be viewed as meeting the autonomy ideal might, as cancer patients, fall short of this ideal. As Arthur Frank contends, illnesses such as cancer can destabilize previous ways of knowing the self; it is “a loss of the ‘destination and map’ that had previously guided the ill person’s life” (Frank 1997, p.1). A relational autonomy lens acknowledges gender and associated variables as important contextual features that may constrain or promote patients’ development and enactment of personal autonomy when making treatment or trial decisions.6 Spectrum of Relational Autonomy Theories Relational theorists differ on which capacities are necessary for autonomy and whether meeting capacity conditions is sufficient to ensure an autonomous self. Broadly speaking, theories of relational autonomy can be located on a spectrum, with procedural accounts on one end and substantive accounts on the other end (Mackenzie & Stoljar 2000).7 Procedural Relational Autonomy At one end of the spectrum are procedural notions of relational autonomy. These accounts specify that a person need only exercise a range of capacities for autonomy to be enacted without regard for the resulting desires, values, and commitments. Importantly, capacities are recognized as themselves developed or undermined in socio-political contexts. 6 Several authors have explored the influence of masculinity and feminity on cancer patients’ self-identities. For a rich discussion, see Bush 2000; Gray, Fitch, Phillips, Labrecque and Fergus 2000; Oliffe 2005; Oliffe 2006. 7 According to Mackenzie and Stoljar, procedural conditions are necessary, although insufficient, for substantive accounts of relational autonomy (see 2000, p. 19). 19 Therefore, procedural accounts are also interested in the historical formation of capacities and whether the socio-political environment helped or hindered their development. For example, cancer patients who are from a paternalistic society might not have had opportunities to develop capacities for autonomy. Thus, their societal context has undermined their autonomy potential for making cancer treatment or trial decisions. Diana Meyers’ theory of autonomy competency is a highly influential procedural account of relational autonomy (Meyers 1989). She argues that autonomy is an activity that consists of persons exercising a range of skills for self-direction, such as introspective or reflective skills, communication skills, reasoning skills, and imagination skills. By exercising these skills people come to grasp “who they are, what matters to them, how they want to develop or change, what constraints limit them and how they can best give expression to their integral desires, beliefs, affections, and values” (Meyers 1992, p. 126). Autonomous persons possess degrees of proficiency or competency in exercising these skills and may be more or less competent in some areas of self-determination than in others. The socio-political environment is central to persons developing autonomy skills. In Meyers’s view, differential access to opportunities to practice autonomy skills may impede or promote autonomy. For example, larger social and cultural ideologies that exist in some societies that construct negative messages about women who are strong and self-interested may undermine some women’s ability to critically evaluate their own needs and preferences (for example, see Bottorff et al. 1998). Thus, a patient’s autonomy skills are limited by a socio-political landscape that creates subtle and powerful pressures on her self-determination. To the extent that the patient is limited by dominant and oppressive social conventions, she may require further opportunities to develop a more robust capacity for 20 self-reflection, which in turn requires social and cultural frameworks that support rather than undermine autonomy-promoting capacities. Substantive Relational Autonomy At the other end of the spectrum are substantive notions of relational autonomy. These perspectives consider not only a person’s capacity to enact autonomy (as in procedural theories) but also specify what sorts of desires, values, and commitments a person may legitimately hold. Paul Benson (1994) argues that autonomous persons must have the ability to tell the difference between right and wrong or between rational and false beliefs. Whether a desire is legitimate or not depends upon the kinds of socialization a person has been subject to and whether it results in a person not being able to distinguish true and false norms underlying their action. Persons may fail “normative competence” when, for example, masculine norms of competition and risk-taking encourage men with cancer to participate in clinical trials (CTs) because they want to be seen as ‘macho’ and receiving the leading edge of treatment. The men’s sense of self-worth is equated with risk-taking as a result of their effective masculine socialization. As such, substantive theorists would identify these men as lacking some autonomy to the extent they have allowed a false norm to govern their lives. Returning to the earlier discussion of the influence of social and cultural ideology on autonomy, even if a female patient with breast cancer was able to self-reflect and identify her own beliefs and values (thus satisfying the procedural account of autonomy), the patient’s ultimate conclusion that her needs are not important is problematic, according to substantive accounts, because it stems from the internalization of problematic social norms concerning women’s moral worth. Benson and other substantive theorists argue that a 21 person who holds personal desires that reflect oppressive socialization can never be fully autonomous, even if his/her reflective and critical capacities are intact (Benson 1994). This view seems to imply that others can override the person’s autonomy in the areas of their lives where the norm governs, and instead make decisions for them that are in their own best interests. Hybrid Account of Relational Autonomy In this dissertation I will rely upon Sherwin’s theory of relational autonomy (Sherwin 1998), which appears to be a hybrid of procedural and substantive notions of relational autonomy theories. As such, Sherwin’s theory possesses the strengths of both theories while avoiding many of their respective weaknesses. For example, a major weakness of substantive accounts is that they can exacerbate rather than ameliorate oppression (Mackenzie & Stoljar 2000). Members of already marginalized groups are deemed non-autonomous because they fail to meet rational or true standards set by strong substantive theorists. By contrast, the major weakness of procedural accounts is that they may not identify autonomy-undermining beliefs, thereby leaving persons vulnerable to interalized oppressive socialization. However, Sherwin (following Meyers’ theory) recognizes degrees of relational autonomy, whereby autonomy fluctuates and exists on a continuum, thus avoiding the entrenchment of oppression associated with strong substantive accounts. Sherwin also suggests further conditions for autonomy that limit the potential for internalized oppression. She argues autonomous persons must exercise self-trust and self-worth. That is, persons must believe in their ability to determine who they are and what is important to them, and trust the judgements they make about themselves (McLeod & Sherwin 2000, p. 262). This limits the sorts of desires an autonomous person may hold, thus 22 guarding against internalized oppression, without going so far as to specify the actual content of an autonomous preference (as in strong substantive accounts). Sherwin’s theory is categorized as weakly substantive because she argues that particular attitudes towards the self (e.g. self-worth and self-trust) are crucial for autonomy in addition to the capacity for critical reflection (McLeod & Sherwin 2000; Sherwin 1998). Similar to other feminist accounts, Sherwin argues that persons are, to a significant degree, socio-politically constructed. One’s attitudes towards oneself, like one’s capacities for engaging in activities that are constitutive of identity and autonomy, are fostered or hindered within interpersonal and political relationships of power and powerlessness (Sherwin 1998). Relational autonomy applied to the context of CTs allows providers to appreciate how the broader socio-political environment may inform or frame the patient-oncologist relationship and influence, for example, which patients with cancer are offered a CT based on providers’ assumptions about their competence or economic status. Supporting cancer patients’ autonomy in this context requires addressing potential biases of providers, enabling patients to exercise their capacities and engaging in activities to promote social justice (Baylis, Kenny & Sherwin 2008; McLeod & Sherwin 2000; Sherwin 1998). This includes addressing unequal background conditions, such as differences in education, working conditions, and access to other healthcare resources (e.g., supportive cancer care) that limit the spectrum of meaningful choice options available and hinder patients’ development of autonomy capacities. In this dissertation, I contend that relational autonomy is an ideal approach to understanding, exploring and, ultimately, supporting cancer patients’ decision-making process within the context of CT recruitment. As applied, the theory can guide the 23 development and maintenance of all persons’ autonomy skills over time, no matter the level of self-determination a person currently holds. All cancer patients, regardless of whether they appear independent or not, are inherently social beings whose healthcare decisions and capacities for autonomy are developed within and continually influenced by the socio-political context. Sherwin’s theory draws attention to how even those patients who appear autonomous from a traditional individualist perspective would have developed their autonomy skills with the assistance of a supportive socio-political context prior to the present healthcare encounter. At the same time, other individuals who appear to enact the traditional notion of autonomy may not have had a supportive social context, and thus, are not achieving the ideal of relational autonomy. These “fully autonomous” patients can still be supported through a relational autonomy lens in maintaining, rather than developing, their autonomy skills. For example, CT personnel play a key role in maintaining patients’ autonomy by engaging them in a shared decision-making process and providing them with meaningful information in an accessible format. Relational Autonomy as Guiding Theoretical Framework Sherwin’s theory is particularly suited to the present study for a number of reasons. First and foremost, she was the first feminist theorist known to apply relational autonomy specifically to the healthcare context (Sherwin 1998). Instead of focusing narrowly on informed consent as a way to respect patient autonomy (which is consistent with traditional autonomy theory), Sherwin uncovers how social context and power relationships inform and influences patients’ autonomy and healthcare decision-making processes more broadly. This requires uncovering the power differences that may exist between patients and healthcare 24 professionals due to differences in knowledge, status, and the fact that professionals play a gatekeeper role to health resources and CTs. Furthermore, addressing oppression and unjust social conditions experienced by marginalized populations within healthcare and the broader community will promote greater relational autonomy within these groups. For example, paying attention to the larger healthcare context, Sherwin’s theory (like all other relational autonomy theories) draws attention to the kinds of options that are made available for patient choice or informed decision-making. She draws attention to the structural conditions underlying the options made available to patients and the potential for limiting patients’ relational autonomy by the absence of meaningful options because of, for example, differences in socioeconomic resources. Patients’ relational autonomy may be greatly inhibited by certain opportunities not being placed on the table for consideration due to healthcare access issues and ability to pay. For example, cancer patients’ autonomy may be limited when they are unable to access a trial due to their geographical location or because their economic position prevents them from travelling to a study site. Thus, attention shifts away from the individual patient making healthcare choices to issues of social justice and access to basic goods and resources that support meaningful decision making. Finally, Sherwin’s theory recognizes self-determination as an ongoing process that may be developed and refined. Even though a person may exhibit autonomy at one point in time, social and structural forces as well as personal experiences can cause person’s capacities to wax and wane in response to illness or new power dynamics. This especially corresponds with the proposed study’s emphasis on the process of cancer patient’s CT 25 decision making and the ongoing role that family, healthcare providers and policy-makers can play in enabling patient autonomy. Sherwin’s theory is thus appropriate as it allows exploration of how to best understand and subsequently support patients’ capacities for autonomy and attitudes towards the self within the provider-patient relationship. It also allows for broader social justice recommendations to apply to a wider circle of relationships, such as policy-makers, national institutes or advocacy groups, through policy and practice change. Sherwin’s theory of relational autonomy (Sherwin 1998) was used to inform multiple aspects of the study, including the use of two methodologies to gain access to different social influences on patients’ CT decision-making process (i.e., interpretive description and grounded theory), and how and from whom the data was collected and subsequently analyzed. In keeping with relational autonomy theory, a critical feminist lens was applied to the analysis phases of the grounded theory research in order to explore gendered power differences in the context of CT decision-making process. Using a grounded theory analysis in conjunction with a critical feminist lens allows power dynamics within social processes to be made visible so that they can be addressed and potentially lead to the emancipation of vulnerable persons. A critical feminist lens was not applied to the analysis phase of the interpretive descriptive study because it was believed that the power hierarchy inherent in the patient-physician relationship and its influence on the CT decision-making process would be more visible from the perspective of patients rather than clinicians. Grounded theory is informed by the fundamental tenets of symbolic interactionalism, which stipulate that humans are dynamic and always exist in relation to others. Social interaction mediates human action by causing one to interpret, revise or reconsider one’s 26 intention or purpose (Blumer 1969). Symbolic interactionalism has been criticized, however, for not theorizing power or for not situating instances of human action within a larger socio-political picture (Kushner 2003). Therefore, in this study, the addition of a critical feminist lens to the grounded theory analysis allowed exploration of the background conditions and power differences that exist for the study population, and more specifically, within patients’ CT decision-making process. Important relationships also play a central role in relational autonomy theory. Therefore, the study population included not only breast and prostate cancer patients but also their support persons and CT personnel. CT personnel, in particular, were interviewed for the interpretive descriptive study to elaborate on larger, macro-level issues that would have bearing on patients’ CT decision-making processes and relational autonomy. Triangulation8 of grounded theory, interpretive description and critical feminist analysis relates well to relational autonomy and its recognition of the socio-political context on persons’ self-determination. By uncovering the socio-political dynamics (e.g., power relationships) and opportunities for enabling patients’ autonomy at the structural level, findings from this dissertation research will be able to inform the development of recruitment procedures and decision support interventions that will empower patients and create more autonomy-supportive CT practices. 8 Triangulation refers to using two or more approaches to answer the research question, thus allowing for a comprehensive picture of the studied experience (Knafl & Breitmayer 1989). For further discussion about how triangulation relates to relational autonomy and the objectives of this dissertation see Chapter Two. 27 Chapter Two: Integrative Literature Review of Cancer Patient Decision Making About Clinical Trials with Implications for Patients' Relational Autonomy The purpose of this chapter is to present an integrative review of the literature that summarizes the factors and contexts influencing cancer patient decision making related to treatment and clinical trial (CT) participation. The theoretical framework of relational autonomy is used to guide the organization of the research findings into personal, social, and structural influences. A secondary aim is to consider how socio-political influences impact patients’ relational autonomy in the context of CT decision-making. Both the treatment and CT decision-making literature is examined because CT discussions often arise as part of clinical care and are sometimes perceived by cancer patients and clinicians to offer an opportunity for state-of-the art or innovative therapy (American Society of Clinical Oncology 1997). Thus, literature describing cancer patients’ treatment decision-making processes may contribute to our understanding of patients’ CT decisions. This understanding is essential to the development of CT recruitment procedures that are better able to address the barriers to relational autonomy experienced by some cancer patients within their CT decision-making process. It also holds the potential to improve cancer patients’ overall cancer trajectory experience. Methodology I chose an integrative approach to guide my review of the literature for this dissertation. Whittemore and Knafl (2005) identify integrative reviews as most appropriate for collecting and synthesizing experimental and non-experimental research in order to add depth and breadth (e.g., rich and varied perspectives) to the phenomemon of interest. This approach is well suited to my dissertation where I am interested in ethical as well as 28 empirical inquiry, including both quantitative and qualitative research, to inform a comprehensive understanding of the current evidence of CT decision making in the context of cancer. Whittemore and Knafl’s (2005) integrative review process was selected due to its systematic approach to data collection, data extraction, and methods of analysis and synthesis, which enhanced the rigour of the review process and improved the accuracy of the conclusions. These strategies are described below. Methods The databases PubMed, CINAHL, and EMBASE were searched to locate relevant literature published between 1995 and 2012. Keywords entered into these searches included a combination of “cancer” or “neoplasm”, “decision-making”, “patient selection”, “patient participation”, and “clinical trial” or “randomized controlled trial.” Articles had to be written in English and limited to adult cancer populations. Studies were included in this review if they examined factors or contexts that may influence cancer patients’ decisions to participate in CTs. Research assessing intent or hypothetical decision-making was included, as were studies involving patients who had previously taken part in CTs and/or patients who had never been approached to participate in CTs. Systematic reviews examining CT participation in cancer populations were also identified. Research that examined decisions about participation in cancer screening trials, however, was excluded in order to narrow the scope of the review and because these patients may not be making decisions from the perspective of having a cancer diagnosis. To supplement results, a related literature search was conducted by replacing the initial keywords “clinical trial” or “randomized controlled trial” with the keyword “treatment.” Articles in the supplementary review also had to be written in English and limited to adult cancer populations. Only qualitative research focusing on cancer 29 patients’ decision-making process was included because the purpose was to shed further light on aspects of patients’ decision-making process that would be relevant to CT recruitment. My doctoral research supervisor, Dr. Lynda Balneaves, and I independently examined titles and abstracts of articles to decide if each met the inclusion criteria and study aims. Disagreement was resolved through discussion and, where necessary, the full text article was accessed and assessed. Once consensus was obtained, I summarized all relevant articles in a literature review matrix to assist with critical evaluation of article findings and quality review of the studies. I developed a 2-point scale (high or low) to assess the quality of each article. A high score was provided to articles that demonstrated methodological or theoretical rigour and data relevance (Moher, Liberati, Tetzlaff, Altman, & PRISMA Group 2010). A low score was provided to articles that I determined did not meet these standards. None of the research in the matrix was subsequently excluded based on a low data quality assessment. Dr. Balneaves independently reviewed the matrix. Key categories of the matrix included author, title, country, study purpose, sample size and type, data collection method and design, outcome measures, key study findings, and study limitations. Results Summary of Systematic Reviews Seven systematic reviews were identified in the initial literature search, 5 of which met PRISMA (Moher, Liberati, Tetzlaff, Altman, & PRISMA Group 2010) standards for reporting quality and were included in the present review (see Table 2.1.). One of these reviews included a meta-analysis of 33 original research articles addressing patient-identified barriers to participation in cancer CTs (Mills et al. 2006). The other systematic review included 9 primary research articles specifically addressing barriers to recruitment of older 30 patients into cancer CTs (Townsley, Selby, & Siu 2005). Another systematic review identified 33 articles that explored poor recruitment or non-participation in CTs within the field of cancer (Cox & McGarry 2003). Of the final two reviews, one evaluated 16 American studies that addressed factors influencing cancer patients’ CT decision-making (Biedrzycki 2010). The other review assessed barriers and facilitators to the participation of women and minorities in CTs in 22 studies (Schmotzer 2012). Limitations of systematic reviews. Although these systematic reviews summarize the cancer CT participation literature, they face several limitations related to generalizability of the findings. Three of the systematic literature reviews were significantly limited in scope and unable to provide a comprehensive analysis of studies conducted to date. For example, one review was limited to addressing barriers to recruitment of older patients into cancer CTs (Townsley et al. 2005). Although this work is important, it cannot provide a broader understanding of the barriers to all adult cancer patients’ participation in CTs. Another of the reviews was limited to original research conducted in the United States and within a restricted timeframe (2004-2010) (Biedrzycki 2010). This significantly limits our understanding of the numerous personal-, social-, and structural-related factors associated with CT participation that are present across different ethno-cultural groups and cancer populations. The systematic review and meta-analysis conducted by Mills et al. (2006) did provide a comprehensive overview of both quantitative and qualitative studies addressing patient-identified barriers to enrolment in CTs. Schmotzer’s review assessing barriers and facilitators to CT participation experienced by women and minorities is also relevant (Schmotzer 2012). The present literature review updates and expands this previous work by 31 highlighting the role of socio-political barriers to CT participation and the importance of interventions that support cancer patients’ relational autonomy in overcoming these barriers. Summary of Quantitative Literature In total, 36 quantitative studies were included in this review (see Table 2.2). None of these studies used experimental research methods. Twenty-four of the quantitative studies were descriptive surveys, three studies were retrospective chart reviews, two studies were case-control, six used multiple research methods, and one used a probability trade-off design. Of the quantitative studies, 20 were conducted in the United States, four in the United Kingdom, four in Canada, three in Australia, and one each in Italy, Korea, China, France and Taiwan. Two studies focused specifically on patient participation in Phase I CTs, three studies focused on Phase III trials, one study focused only on Phase II and III trials, one study focused on Phase I and II trials, and the remaining studies either focused on all forms of randomized trials or did not specify what type of CTs they focused on. All studies focused on cancer drug trials except for one study, which examined patient preferences related to a surgical cancer trial. Fifteen of the quantitative studies primarily assessed hypothetical participation or potential future willingness to participate in a CT. The remaining 21 studies reported on factors affecting accrual related to those patients who had been approached to participate in a CT, were currently enrolled in a CT, or had declined or withdrawn from a CT. Factors associated with CT participation: Findings from the quantitative literature. Personal, social, and structural factors influencing CT participation were obtained from the literature. Personal factors were identified as demographic influences on cancer CT accrual such as patients’ age, ethnicity, gender, and disease stage. Personal factors also included patient characteristics that could affect accrual, such as patients’ values and beliefs 32 or knowledge and awareness of CTs. Social factors referred to relationships, such as cancer patients’ relationships with family or friends, or the patient-physician relationship, or demographics that had social components that could influence accrual. Structural factors related to institutional policies/procedures or the political environment within which CTs are conducted. This includes access to resources such as healthcare, insurance coverage, and institutions hosting cancer CTs. Findings were separated into these three broad categories because relational autonomy theory recognizes the influence of social context and structural factors on personal decision-making. Highlighting all three factors in a systematic fashion thus served to elucidate the relevance of these influences on cancer patients’ relational autonomy. Although personal, social, and structural factors are presented as discrete concepts, I recognize there is fluidity between these categories and as such, overlap of influences can occur. For example, where patients live in relation to cancer centres hosting trials as well as their socioeconomic or financial resources to attend trials are interrelated and influence patients’ ability to participate in CTs. In fact, the intersectionality of these factors is itself socially and politically significant, since their configuration may marginalize certain populations in accessing CTs. This raises challenging justice considerations related to who gets the opportunity to participate in CTs and which populations benefit from research. Therefore, in the following sections some factors are discussed in one category and elaborated upon or corroborated in another. This is followed by a call for further consideration of the socio-political implications of these factors. Demographic factors influencing accrual. Most of the quantitative body of research found that several demographic factors affect cancer patient accrual to CTs. For example, 33 several studies found that younger age was a predictor of CT participation (Advani et al. 2003; Ellis, Butow, Tattersall, Dunn, & Houssami 2001; Go, Frisby, & Lee 2006; Gross, Filardo, Mayne, & Krumholz 2005; Sateren et al. 2002) with the exception of one study, which found that older age was significantly associated with CT participation (Biedrzycki 2011). Race or ethnicity was also associated with greater participation in CTs, with Caucasian cancer patients participating in studies more frequently than other ethnic groups (Advani et al. 2003; Agrawal et al. 2006; Avis, Smith, Link, Hortobagyi, & Rivera 2006; Gross et al. 2005; Sateren et al. 2002). Discrepancies between study findings were found to exist in relation to other personal demographics or characteristics. Some studies found more advanced disease stage (Kim et al. 2008; Klabunde, Springer, Butler, White, & Atkins 1999; Li & Jiang 2010) and fewer comorbidities (Kemeny et al. 2003) were associated with cancer patient participation in CTs. However, it seems reasonable to assume that patients with more advanced stage of disease would have more, not fewer, comorbidities. Contradictory findings were also apparent for the influence of gender on CT participation. One study found males have more favourable perceptions of CTs, which encourages their participation in cancer trials (Kim et al. 2008). Other studies found no statistically significant differences in cancer CT accrual based on gender (Klabunde et al. 1999; Sateren et al. 2002). Other personal factors that influence accrual. Cancer patients’ beliefs and attitudes have been identified as being associated with CT participation. These include a strong desire to help others, perceived personal benefit, and hope for a cure (Agrawal et al. 2006; Avis et al. 2006; Catania et al. 2008; Davison, So, Goldenberg, Berkowitz, & Gleave 2007; Jenkins & Fallowfield 2000; Jones et al. 2006; Kemeny et al. 2003; Lara et al. 2001; Schutta & 34 Burnett 2000; Wright et al. 2004). For example, Agrawal et al. (2006) found information about whether the experimental treatment killed cancer cells to be the most useful piece of information for 63% of patients deciding whether to participate in a Phase I oncology study. Similarly, other studies found cancer patients’ hope that the trial offered some medical benefit and belief that it was the best option available to be the most important reasons for CT participation (Catt, Langridge, Fallowfield, Talbot, & Jenkins 2011; Truong, Weeks, Cook, & Joffe 2011; Wang, Tsai, Chen, & Tsay 2011). Conversely, lack of knowledge about CTs, inconvenience, and perceived drawbacks or risks detracted from patient participation in CTs (Agrawal et al. 2006; Avis et al. 2006; Meropol et al. 2007). Negative attitudes or beliefs about CTs, including patients’ fear of randomization, concerns about the experimental nature of a CT, and lack of therapeutic benefit were some of the main reasons cancer patients declined participation in CTs (Ellis et al. 2001; Jones et al. 2006; Jenkins & Fallowfield 2000; Kim et al. 2008; Meropol et al. 2007; Weckstein et al. 2011). For example, most cancer patients opted not to participate in a CT because they feared their treatment would be delayed or they would receive a placebo if they were to accept CT participation (Ellis et al. 2001; Jones et al. 2006). Also, concern about side effects associated with an experimental treatment was a significant detractor from CT participation as was fear of being a “guinea pig” (Catt et al. 2011; Li, Yu, & Jiang 2010; Sabesan et al. 2011; Wang et al. 2011). Some of the quantitative studies reported confounding rather than statistically predictive variables that influence patient attitudes or willingness to engage in cancer CTs. For example, awareness and knowledge of a CT has been identified as a strong influence on cancer patients’ willingness to participate in CTs. But greater awareness of cancer CTs has 35 also been associated with younger age (Jones et al. 2006; Kim et al. 2008), race (e.g., Caucasian) (Meropol et al. 2007), higher education (Kim et al. 2008; Meropol et al. 2007), greater clinical stage (Davison et al. 2007), and higher economic status (Kim et al. 2008). With regards to education, for example, one study found in a sample of 100 cancer patients that only 23% of patients with elementary education had some general knowledge about the purpose of CTs as opposed to 98% of those with a university or college educations (Jones et al. 2006). Therefore, more research is required to understand the interplay between and among these variables, and to make the link between cancer patient willingness to participate in CTs and actual accrual. Limitations of literature assessing personal factors. What have been lacking in the literature exploring the personal factors influencing CT participation are how personal characteristics or attributes are informed by social and structural level influences, and vice versa. Such knowledge will allow investigators to explore and address some of the differences in CT participation found among different races, age groups, and levels of CT awareness. It might also allow some “assumptions” about who is willing to participate in CTs to be challenged, thus potentially addressing recruitment bias. It should be noted that personal factors are just the tip of the iceberg. The interplay with social and structural factors needs to be understood to avoid an oversimplification of the complex and interrelated array of influences on CT decision-making when addressing accrual challenges. Also, education and decision support related to CT participation could be tailored to address the most important issues, rather than surface-level differences among participants and non-participants. Finally, it should be noted that demographic factors might be just a proxy for underlying socio-political reasons for low accrual in certain populations with 36 cancer. Education and income level, for example, when explored further, may reveal differential access to higher education and better paying occupations related to their social class or cultural group. More in-depth research on cancer patients’ decision-making process with greater attention and sensitivity to the interplay between socio-political influences is required. Such attention is crucial for removing barriers and promoting patients’ relational autonomy in the CT context. Social factors influencing accrual. Socio-demographics such as higher education (Advani et al. 2003; Agrawal et al. 2006; Avis et al. 2006; Ellis et al. 2001) and greater household income (>$15,000/year) (Advani et al. 2003; Agrawal et al. 2006) have been found to be predictors or characteristics of cancer patients who participate in CTs. Other studies have reported poor CT participation from lower socioeconomic backgrounds and from marginalized groups (e.g. elderly, women, ethno-cultural) (Advani et al. 2003; Brown & Topcu 2003; Gross et al. 2005; Roberson 1994; Robinson, Ashley, & Haynes 1996; Townsley et al. 2005). These socio-demographic factors may overlap with personal factors identified in the previous section. For example, Lara and colleagues suggested lack of participation from marginalized groups might be a result of cancer patients’ lack of knowledge and awareness about CTs (Lara et al. 2005). Other studies found patients with low socioeconomic status have greater severity of comorbid conditions, which precludes them from CT eligibility (Go, Frisby, & Lee 2006; Gross et al. 2005; Kemeny et al. 2003; Kornblith et al. 2002). Greater understanding of the structural influences underlying lower socioeconomic status and the interplay of personal, social, and structural factors impacting CT accrual is required in order to more fully understand the barriers to relational autonomy that are disproportionately experienced by these populations. 37 With regards to relationships that were found to impact cancer patients’ participation in a CT, recommendations received from family members and friends were highly associated with CT participation (Avis et al. 2006; Blackhall, Murphy, Frank, Michel, & Azen 1995; Kim et al. 2008; LaVallie, Wolf, Jacobsen, & Buchwald 2008). Also, altruism or desire to help future cancer patients was among the top predictors of CT participation (Sabesan et al. 2011). Family support for attending clinical institutions hosting CTs also ranks high in importance, particularly for rural or remote cancer patients (Sabesan et al. 2011). Research has also shown that physicians and their relationships with patients are important factors in patients’ decisions about CT participation. In particular, the trust that patients place in their physicians has been found to be an important predictor of CT participation (Avis et al. 2006; Catania et al. 2008; Coyne, Demian-Popescu, & Brown 2004; Davison et al. 2007; Eng, Taylor, & Verhoef 2005; Jones et al. 2006; Jenkins & Fallowfield 2000; LaVallie et al. 2008; Meropol et al. 2007; Nurgat et al. 2005; Schutta & Burnett 2000). In one survey of 102 cancer patients, 76% reported trust in the physician-investigator as the main reason for taking part in a CT (Catania et al. 2008). Not surprisingly, direct recommendations from trusted physicians have been found by other researchers to be among the most important predictors of CT participation among cancer patients (Avis et al. 2006; Ling Rees, & Hardy 2000; Solomon, Pager, Young, Roberts, & Butow 2003; Baum 1993; Cox & McGarry 2003; Ellis 2001; Flessig, Jenkins, & Fallowfield 2001; Jenkins & Fallowfield 2000; Mancini et al. 2007; Mills et al. 2006; Roberts 2002). These recommendations are even more powerful when physicians have been described as being “overly enthusiastic” or emphasizing the social benefit of taking part in cancer 38 research (Avis et al. 2006; Flessig et al. 2001; Jenkins & Fallowfield 2000; Llewellyn-Thomas, McGreal, Thiel, Fine, & Erlichman 1991; Mancini et al. 2007). Further research is needed, however, to understand how patients develop trust, whether their trust applies to the treating physician, the physician-investigator, or a third-party researcher, how trust might blur therapeutic and research ethics boundaries, and how trust is maintained across a diverse range of patient populations and social contexts. In sum, a variety of interpersonal and social relationships help shape cancer patients’ relational autonomy within the context of CTs by providing patients with relationships with trusted individuals that help guide their behaviour, as well as perspectives, opinions, and recommendations that inform their decision making. These social relationships may also practically support rural patients by making CT participation a logistical possibility, thus broadening the scope of options that would otherwise be unavailable for some cancer patients. Limitations of literature assessing social factors. This literature is limited by a lack of appreciation and critical understanding of the interplay between personal, social, and structural factors that influence cancer patients’ decision-making about CT participation. For example, the importance that cancer patients have placed on their physicians’ recommendations as well as the perceived altruistic benefits of taking part in CTs raises the question of whether some patients may feel socially constrained or pressured to participate in cancer research. In particular, the inherently dependent nature of the relationship between patients and their physicians cannot be ignored. The possibility of such undue influence has been raised by Mills et al. (2006), who reported in their meta-analysis and systematic review that “…20% of 100 cancer patients surveyed felt coerced to participate in a CT” (p. 146), 39 although the authors did not define coercion in their article. Given the difficulties (or uncertainties) in both defining and measuring coercion, caution is required when interpreting this finding in light of how coercion is defined in mainstream ethics. For example, the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (2nd edition) identifies coercion as involving a threat of harm or punishment for failure to participate (CIHR 2010). At the very least, Mills et al.’s finding draws attention to the potential for undue influence that may arise when patients are recruited into CTs by individuals in a position of authority (i.e., physicians). This influence may be very subtle and unintended, stemming from misconceptions held by both physicians and patients. Physicians may believe that certain CTs are particularly beneficial to some patients, which they may convey indirectly in how enthusiastically they present CTs (Avis et al. 2006; Baum 1993; Joffe, Cook, Cleary, Clark, & Weeks 2001). Joffe et al. (2001) found that in their study, physicians who were also researchers focussed more on what CTs may do for patients as a treatment, rather than underscoring the purpose of research is to benefit future patients. From a patient perspective, some cancer patients have identified concern about the impact of their decision about CT participation on their physician-patient relationship and the care they receive (Avis et al. 2006; Davison et al. 2007). For example, Avis et al. (2006) identified the possible change in a patient’s relationship with a healthcare provider as a significant factor that differentiated CT accepters from decliners. In the absence of in-depth interviews, cancer patients’ interpretation of what constitutes undue influence and the role of social structures that empower or disempower patients in the decision-making process (e.g., low socioeconomic status, ethnicity) are not clear. Furthermore, a broader understanding of how macro level issues, such as the structure 40 and delivery of cancer care as well as availability of cancer CTs, affects the healthcare provider-physician relationship is required. Structural factors influencing accrual. Turning towards structural influences, a variety of factors associated with policies and procedures embedded within healthcare systems and institutions, as well as resource issues, have been identified as important influences on cancer patients’ participation in CTs. For example, uncertainty regarding coverage for costs related to CT participation (e.g., transportation costs, laboratory tests and medication associated with the trial) has been found to have a significant effect on patient accrual in the US (Avis et al. 2006; Klabunde et al. 1999; Lara et al. 2001; Meropol et al. 2007). American cancer patients with access to health insurance were more likely to participate in CTs than patients without health insurance (Agrawal et al. 2006; Avis et al. 2006; Sateren et al. 2002). However, study findings differed on whether patients who had access to government-funded health insurance versus private health insurance were more likely to enroll in cancer CTs (Gross et al. 2005; Klabunde et al. 1999; Lara et al. 2001). In Australia, one study found that cancer patients were motivated to accept CT participation because it would allow them access to otherwise unavailable treatments (Sabesan et al. 2011). However, whether this was because treatments were otherwise unavailable due to insurance or drug availability issues remains unexplored. In Canada, among the main reasons patients decided to participate in CTs was so they could access free drugs (Jones et al. 2006). However, this was the only Canadian study to uncover access issues. More research is required that examines the Canadian context, given its universal healthcare system and recent reports (Canadian Cancer Research Alliance 2011; 41 Dent & Yurichuk 2011) that have identified that inequitable access to expensive cancer drugs across the country may be influencing patients’ CT decisions (see Introduction). Healthcare organizational factors have also been shown to affect the climate within which CT participation is situated. Studies have shown patient accrual is lower in busier clinics and increasingly complex pharmaceutical company protocols have forced institutions to limit the CTs conducted because of resource and staff limitations (Go et al. 2006; Grunfeld, Zitzelberger, Coristine, & Aspelund 2002; Siminoff, Zhang, Colabianchi, Sturm, & Shen 2000; Wright et al. 2004). In addition, decreased access to care and to institutions hosting clinical research has been identified as a major barrier to cancer patients’ participation in CTs, particularly for rural or remote cancer patients (Sabesan et al. 2011). Further discussion of these findings has uncovered how these structural barriers relate back to social factors, such as inflexible work hours, transportation and child care issues, and communication difficulties (e.g. language and literacy) (Gross et al. 2005; Kornblith et al. 2002; Melisko et al. 2005). Applying a relational autonomy lens to this research will elucidate the complex interplay between CT policies and practices and social factors that contribute to some cancer patients’ inability to access CTs within these organizations. Overall limitations of quantitative literature on CT participation. Overall, this primarily quantitative body of research points to not only the personal nature of decisions about CT participation but also the important socio-political context of cancer patients’ decisions about CTs. While this research has identified important factors associated with CT participation, it has been primarily survey research informed by investigator-developed concepts. As such, it has provided limited understanding of the complex interplay of factors underlying patients’ decision-making process and the experience of making CT decisions 42 from the patient’s perspective. Without sufficient understanding of how personal, social, and structural factors interact and influence patients’ relational autonomy and decision-making, researchers will be unable to develop targeted interventions to support patients’ relational autonomy and ultimately support innovative evidence-based programs of cancer care. In the following section I introduce and discuss the contributions and limitations of the qualitative literature as it relates to cancer patients’ decision-making processes about CT participation. Summary of Qualitative Literature In total, 15 qualitative studies met the inclusion criteria for this review (Table 2.3). Of the 15 qualitative studies, six were conducted in the United States and two were conducted in Canada. Twelve of the qualitative studies used semi-structured interviews and three used focus groups to elicit information from cancer patients about clinical trial participation. CT decision-making processes: Findings from the qualitative literature. In marked contrast to the relatively well-developed body of quantitative literature on factors related to CT accrual, qualitative research exploring the CT decision-making processes of cancer patients has been minimal, with only 15 studies having been conducted to date (Cox & Avis 1996; Coyne, Demian-Popescu, & Brown 2004; Ellis & Butow 1998; Eng Taylor, & Verhoef 2005; Grunfeld, Zitzelberger, Coristine, & Aspelund 2002; Huizinga, Sleijfer, van de Wiel, & van der Graaf 1999; Kohara & Inoue 2010; Kvale, Woodby, & Williams 2010; Madsen, Holm, & Riis, 2007; Mills et al. 2011; Schaefer, Ladd, Gergits, & Gyauch 2001; Schutta & Burnett 2000; Shah et al. 2012; Shannon-Dorcy & Drevdahl 2011; Stevens & 43 Ahmedzai 2004). This research alludes to the complex interplay between personal and social factors within cancer patients’ CT decision-making. Most striking in this literature is the important role of family members and health professionals in patients’ CT decision-making processes. Physicians in particular are important because they often provide information about CTs and create a climate of trust in which patients felt comfortable agreeing to take part in research (Cox & Avis 1996; Coyne et al. 2004; Kohara & Inoue 2010; Madsen et al. 2007; Schaefer et al. 2001).9 Not all patients, however, felt fully informed about the purpose of a CT and its procedures (e.g., randomization), which resulted in feelings of anxiety, uncertainty and distrust among these patients (Ellis & Butow 1998; Madsen et al. 2007). In Huizinga et al.’s (1999) study of 14 Dutch cancer patients, this uncertainty included patients misinterpreting their CT participation as being part of their standard care. Similar findings were reported by Stevens (Stevens 2004), who explored the reasons why British cancer patients declined to participate in CTs. This body of qualitative research also highlights patients’ concerns about the potential lack of clinical equipoise10 associated with cancer CTs (Madsen et al. 2007) and 9 It is not clear in many of these studies whether physicians informing patients about CTs were treating physicians or physician-investigators. This difference may have important implications for patients’ relational autonomy. For example, the dual role of physician-investigators may impact the trust patients have in these physicians and the way in which information was provided, since physician-investigators could have secondary interests that conflict with the patient’s best interest. 10 Clinical equipoise means there is genuine uncertainty in the expert medical community about whether a treatment will be beneficial (Freedman 1987). 44 the perception that a CT was the only treatment option (Cox & Avis 1996; Coyne et al. 2004). In one study, interviews with female cancer patients revealed that many expressed a clear preference for the experimental treatment, perceiving it to be more effective than standard care (Madsen et al. 2007). As a result, the women felt “forced” to participate in the CT in order to obtain the new treatment. Participants also revealed that they perceived there to be a “moral imperative” for cancer patients to take part in CT research to help future patients and further science (Madsen et al. 2007). For some patients who declined CT participation, this sense of obligation was pervasive and led to feelings of guilt (Stevens & Ahmedzai 2004). These findings hint at the moral obligation to promote the social good, or social pressure that cancer patients may internalize as part of the CT decision-making process. Applying a relational autonomy lens to this body of research would help uncover these subtle norms that may undermine rather than support patients’ decision-making process, and that require attention within CT procedures. Several studies spoke of the role of family members and other support persons in cancer patients’ decisions about CT participation (Coyne et al. 2004; Grunfeld et al. 2002; Huizinga et al. 1999; Kohara & Inoue 2010; Schaefer et al. 2001). One study reported “family pressure” as among the main reasons cancer patients gave for participating in a CT (Cox & Avis 1996). The degree to which family members influenced patients’ choices for the patients’ own good (versus the family’s good), how they were involved in the decision-making process, and patients’ preferences with regards to the involvement of their family in CT decisions, however, remained largely unexplored. An exception was Schaefer et al. (2001), who found participants carefully considered their social and family responsibilities as part of their decision to take part in a CT. 45 Summary and limitations of qualitative literature describing cancer patients’ CT decision-making processes. In summary, this body of research highlights the important role of physicians, family members, and support persons on cancer patients’ decision-making about CT participation. However, these studies did not explicitly seek to explore the social context of CT decision making, nor did they uncover any structural influences on patients’ decision-making process. Therefore, more targeted research informed by the theoretical framework of relational autonomy is required that examines to what extent patients are able to exercise autonomy, particularly when embedded in social and political structures, and faced with subtle pressures stemming from both perceived moral and social obligations. In addition, further research is required that explores the actual decision-making process and how cancer patients make the decision to engage in a CT. Within this process, beliefs about CTs that may constrain patients’ relational autonomy and unduly influence their CT decision may be further explored and addressed. Furthermore, a relational autonomy perspective may encourage researchers to consider views from other individuals who are involved in the decision-making process and who are influential from the patient’s perspective. Treatment decision-making in cancer care: supplementary review of the literature. A related body of literature that informs our understanding of how cancer patients make decisions about CT participation is qualitative research that has focused on patients’ decisions about cancer care. This is particularly relevant given the potential for therapeutic misconception that has been identified by bioethicists (Appelbaum, Roth, Lidz, Benson, & Winslade 1987) in which cancer patients do not distinguish between treatment options offered as part of standard care or through clinical research. Cancer patients’ beliefs about the difference between CTs and treatment is immediately relevant to the previous section on 46 CT research and the importance of relational autonomy as a lens to guide future research, since this belief may also be influenced by underlying social and political contexts. Treatment decision-making processes. To date, there have been approximately 20 published qualitative studies that have explored the way in which cancer patients make decisions about their cancer care as well as the factors that influence their decisions (Balneaves, Truant, Kelly, Verhoef, & Davison 2007; Charles, Redko, Whelan, Gafni, & Reyno 1998; Davies, Rhodes, Grossman, Rosenberg, & Stevens 2010; Elit et al. 2003; Halkett, Scutter, & Borg 2007; Kelly-Powell 1997; Lam, Fielding, Chan, Chow, & Or 2005; Maliski, Heilemann, & McCorkle 2002; Markovic, Manderson, & Quinn 2006; O’Brien et al. 2008; Pierce 1993; Pieterse, Baas-Thijssen, Marijnen, & Stiggelbout 2008; Sainio, Eriksson, & Lauri 2001; Sanders & Skevington 2003; Watanabe, Takahashi, & Kai 2008; Wenzel & Shaha 2008; Zhang & Siminoff 2003; Ziebland, Evans, & McPherson 2006). This research has highlighted the iterative and complex nature of the decision-making process, with cancer patients moving back and forth between identifying information needs, seeking information, establishing trust with information sources and healthcare providers, making treatment decisions and evaluating the outcomes of those decisions. What is striking within these descriptions is the important role patients’ social network, including trusted health professionals, played in their decisions about treatment options. Often, patients reported discussing treatment alternatives with their family members, friends, and fellow cancer survivors as a way of weighing the potential risks and benefits of therapy options, including the possible effects on relationships, ability to carry out their social roles, and overall quality of life (Balneaves et al. 2007; Davies et al. 2010; Elit et al. 2003; Halkett et al. 2007; Kelly-Powell 1997; Lam et al. 2005; Maliski et al. 47 2002; Markovic et al. 2006; Pieterse et al. 2008; Sainio et al. 2001; Wenzel & Shaha 2008). For many cancer patients, minimizing life disruption for themselves and their families was an important consideration in their decisions, which were described as being “a family affair” (Davies et al. 2010; Kelly-Powell 1997; Lam et al. 2005). Social networks, however, may not always be supportive of cancer patients’ decision-making processes (Davies et al. 2010; Elit et al. 2003; Maliski et al. 2002; Zhang & Siminoff 2003). For example, in Zhang and Siminoff’s study (2003) of 37 lung cancer patients and 40 family members, some patients reported experiencing pressure from family members to select treatment options that were not congruent with their own preferences. Whether this pressure by family members can be considered “undue influence” and to threaten cancer patients’ relational autonomy in the decision-making process is not clear. Given the interdependent nature of families, as well as the potential emotional vulnerability of some patients at certain points within the cancer journey (Weisman & Worden 1976), making treatment choices that preserve relationships may be as important as selecting a treatment that will enhance survival. There has been limited research, however, that has explored treatment decision making from the perspective of family members and support persons (Maliski et al. 2002; Zhang & Siminoff 2003). The findings from future research may be applied to the context of CTs, and shed further light on how social networks undermine and/or support cancer patients’ relational autonomy in the decision-making process. Health professionals were also acknowledged by cancer patients as being an integral part of the treatment decision-making process (Charles 1998; Davies 2010; Elit 2003; Markovic 2006; Pieterse 2008; 2001; Sanders 2003; Watanabe 2008; Wenzel 2008; 48 Ziebland 2006). Cancer patients have been found to frequently express the desire to “share” their treatment decisions with their oncologist and other healthcare providers above and beyond the disclosure of information for patients’ informed consent (Charles et al. 1998; Pieterse et al. 2008; Ziebland et al. 2006).11 This preference has led to a strong movement in cancer care towards shared decision-making, which assumes cancer patients will take an active role in treatment decisions in consultation with their health professionals (Butow & Tattersall 2005; Charles, Gafni, & Whelan 1997; Charles 1999; Wenzel & Shaha 2008). This might mean encouraging health professionals to engage patients’ participation in the decision-making process to the extent that patients want to be involved or directive of their care. However, patients may experience physical, emotional, and relational challenges following a cancer diagnosis. Such challenges may undermine patients’ ability to think clearly about their treatment options and inhibit them from taking an active or participatory role in the decision-making process. Some individuals, at certain times in the cancer trajectory, may prefer a passive role that defers decisions to their physicians (Davies et al. 2010; Degner & Sloan1992; Degner et al. 1997; Wenzel & Shaha 2008). Other research has uncovered patients describing their medical consultations to be intimidating and “emotionally charged”, with some individuals reporting feeling “abandoned” and “scared” when asked to make a treatment decision, for fear of making a wrong choice (Pieterse et al. 2008; Sanders & Skevington 2003; Ziebland, Evans, & McPherson 2006). 11 For example, in Charles et al.’s study (1998), female patients with cancer expressed appreciation of physicians who explored patients’ understandings of treatment benefits and risks using a ‘decision board’, and helped them to interpret scientific information. 49 The power imbalances that are inherent in the patient-physician relationship with regards to knowledge and access to care (Say & Thomson 2003) may also place limits on patients’ autonomy in the treatment decision-making process. Contributing to these power dynamics are socio-political factors, such as socioeconomic status and ethno-cultural norms (Lam et al. 2005; Watanabe, Takahashi, & Kai 2008; Zhang & Siminoff 2003). For example, in Watanabe et al.’s (2008) study of Japanese cancer patients’ treatment decision-making preferences, paternalism is a widely pervasive cultural tradition. Japanese physicians are seen by the community to have the indisputable expertise and authority to make treatment decisions without seeking input from patients. Finally, resource constraints and limited insurance coverage may also inhibit meaningful patient participation in treatment decisions by restricting access to, and thus choice of, provider, treatment, and hospital (Degner 2002; Maliski, Heilemann, & McCorkle 2002). In order to overcome these socio-political barriers and promote patient participation, researchers need to examine the gaps and limitations at the structural level and how these influence patient decision-making (Degner 2002). Such recommendations hold relevance with regards to CT decision making and point to the importance of applying a relational autonomy lens to future decision-making research. Discussion Relational Autonomy and Clinical Trial Recruitment Guidelines for the ethical conduct of human subjects research demand that researchers are aware of the potential for individuals to be unduly influenced by others, especially by those in positions of authority, which could undermine their ability to autonomously consent to participate in research (CIHR 2010). Researchers have been slow, however, in adopting a 50 relational autonomy lens when exploring cancer patients’ CT decisions. This is despite growing recognition within the bioethics literature of the significance of social networks and socio-political influences on patients’ autonomy within general healthcare decision-making processes (Donchin 2001; Ho 2008; Mackenzie & Stoljar 2000; Sherwin 1998). Several studies have identified relational autonomy as being the most appropriate framework to capture the relational complexity of many forms of medical decision making, including advance care planning (Robinson 2011), disclosure of genetic information (Gilbar 2007), and decision-making at end of life (Candib 2002; Eliott & Olver 2008). It is timely to expand the theoretical application of relational autonomy to the unique decision-making processes associated with CT participation, particularly given the difficulties faced by the clinical research community related to accrual and the ethical imperative to ensure respect for research subjects’ autonomy within CT procedures. Future research informed by relational autonomy theory has the capacity to gain a broader sense of the social and structural influences on cancer patients’ CT decisions and holds the potential to inform policy and practice recommendations, particularly with regards to CT recruitment and informed consent procedures. For example, future study findings may suggest ways of involving support persons in the dialogue about CT participation or addressing socio-political barriers that may exist for some marginalized groups in clinical settings. Conclusions While an extensive body of inquiry has described patient attitudes towards, and barriers to, participation in CTs, it has been primarily survey research that has been driven by investigator-developed concepts, which are informed by traditional notions of patient 51 autonomy. This research is limited in its understanding of patients’ perspectives of the decision-making processes surrounding CT participation and the socio-political context in which these decisions are made. Future research is essential to the development of CT recruitment procedures that are better able to address the barriers to relational autonomy experienced by some patients within their CT decision making and for potentially improving CT enrolment. Finally, it is important to acknowledge the potential limitations of using a relational autonomy lens to inform research on cancer patient decision-making processes related to CT participation. Supporting cancer patients’ enactment of relational autonomy within the CT decision-making process may not necessarily lead to greater CT participation. In fact, it may lead to the opposite consequence with better-informed and more autonomous patients deciding not to participate in CTs. There is reason to believe, however, that patients who are fully supported in making decisions that reflect their beliefs and values, acknowledge their social contexts, and address the socio-political barriers that exist to CT participation, will feel more empowered to take part in a well-designed CT. This belief stems from research that suggests patients with a decreased sense of personal empowerment and who face socio-political barriers are less likely to participate in CTs (Ellis & Butow 1998; Ellis et al. 2001; Jenkins & Fallowfield 2000; Llewellyn-Thomas 1991). Moreover, there has been suggestion that upholding ethical standards that protect patients’ welfare and interests (such as respect for patient autonomy), and ensuring sound and socially valuable science may contribute to patients’ trust in the research process and their willingness to participate (Weijer & Miller 2007). However, whether or not increased enrolment in CTs is achieved, we must be reminded that a primary motivation for research on cancer patient CT decision-making is to 52 support patients in making well-informed and meaningful decisions about CTs and to inform the development of enhanced ethical procedures for CT recruitment. 53 Table 2.1 – Systematic Literature Reviews Reviewed Author, date, and country Focus of the review Search strategy Search terms Dates Primary inclusion/exclusion criteria # of articles reviewed Biedrzycki BA. 2010. US To describe what is known about the factors that influence cancer CT decision making Pubmed and the researcher’s review of the reference lists in identified articles “clinical trials”, “patient participation”, “decision making” 2004 - 2010 Inclusion criteria: • Written in English, • Study conducted in the US, • Included only adults with cancer, • Focused on cancer CT participation, • Reported original quantitative or qualitative research. Exclusion criteria: • Did not have an abstract, • Location outside US, • Not original research, • No cancer CT participation option, • Pediatric sample, • Sample did not target people with cancer 16 54 Author, date, and country Focus of the review Search strategy Search terms Dates Primary inclusion/exclusion criteria # of articles reviewed Cox K; McGarry J. 2003. UK To explore poor recruitment or non-participation in CT with specific reference to the field of cancer research CINAHL, Medline, Nursing Collection, British Nursing Index, Embase, Ingenta Journals, Swetsnet-navigator, manual journal searches, internet websites, search of reference lists from retrieved papers “cancer clinical trials”, “non-participation”, “participation”, “medical research” N/A Inclusion criteria: • Retrospective descriptive analysis of data from clinical trial non-participants; • Follow-up studies of non-participants and focus group studies with potential clinical trial participants • CTs outside cancer 33 Mills EJ; Seely D; Rachlis B; et al. 2006. Canada The attitudes and barriers of patients to participation in cancer CTs AMED, Campbell Collaboration, CINAHL, Cochrane library, Embase, ERIC, Medline, NHS EED, Clinicaltrials.gov website, UK national research register “cancer” or “oncol*”, “clinical trials”, “barriers”, “participat*”, and “enrol*” Varied according to database. Lower search limit: 1966; Upper search limit: 2005 Inclusion criteria: • Original research study; • Contained content addressing patient-identified barriers to participation in cancer CTs; • Paper included a semi-structured interview, focus group study, or survey of patients with cancer. Exclusion criteria: • Only addressed participant descriptions and patient demographics, • Vaccine trials 33 Schmotzer, GL 2012 USA Investigate barriers and facilitators that PubMed, CINAHL, PsycINFO MESH terms: “clinical trials”, “minorities”, 1995-2008 Inclusion criteria: • Descriptive analysis of data from clinical trial participants and 22 55 Author, date, and country Focus of the review Search strategy Search terms Dates Primary inclusion/exclusion criteria # of articles reviewed provide explanations for the low participation of women and minorities in clinical trials “minority groups”, “participation”, “recruitment”, “research subjects”, “neoplasm” nonparticipants obtained by survey or questionnaire, • Focus group studies with potential clinical trial participants and health care providers, • Prospective data obtained from health care providers and articles that analyzed the patient recruitment patterns of several studies. Exclusion criteria: • Not in English, • Did not address research participation or recruitment, • Did not represent an underrepresented group (defined as women, ethnic minorities, rural/older, low SES) Townsley CA; Selby R; Siu LL. 2005. Canada Barriers to recruitment of older cancer patients to CTs Medline, Embase, Cochrane Central Register of Controlled Trials “clinical trials”, “patient selection”, “cancer”, “neoplasm”, “tumor”, “aged”, “elderly”, “participate”, “recruit”, “enroll” 1994-2004 Inclusion criteria: • Primary research articles specifically addressing barriers to recruitment of older patients onto cancer CTs, • Described strategies to overcome recruitment barriers in an older population 9 56 Table 2.2 - Quantitative Studies Reviewed Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions Avis et al. (2006) USA 183 women breast cancer patients Survey Telephone interview with investigator-developed and standardized (Davis et al. 1993) survey items Sociodemographic: age, level of education, race/ethnicity, employment status, income, health insurance, and insurance coverage of the CT Medical: time since diagnosis (self-report), stage of breast cancer, type of surgery, type of treatment, and health (self-report) Knowledge of CTs: purpose of CTs, random assignment, eligibility, participant rights, and informed consent Attitudes towards CTs: beliefs about the benefits and drawbacks of CTs Personal: potential adverse effects of treatment, attitude towards random assignment, trust in medical research, amount of time and travel required, and others’ recommendations Personal factors • Perception of personal benefits • Perception of drawbacks Social factors • Physician recommendation System factors • Time commitment • Travel required • Trial type/phase • Insurance coverage • Cancer center Advani et al. (2003) USA 72 African-American and 146 white cancer patients (various diagnoses) Survey Standardized 31-item telephone survey (Lannin et al. 1998; Roberson 1994; Ware and Sherbourne 1992) Knowledge of CTs: awareness and knowledge about CTs, previously asked to participate in a CT; attitudes about CTs; benefits of CTs, Religious and/or spiritual beliefs: fatalism about cancer Satisfaction with their oncologist and clinic: doctor explains things, trust in doctor Financial and/or transportation issues Willingness to participate Demographics: age, level of education, race, household income, participation rates and reasons for refusal Factors affecting CT participation: advice from friends and family, advice from physicians, chance of side effects, chance CTs benefits themselves or others, , lack of treatment options, religious beliefs, transportation and cost Functional status: limitations in daily activity, disease stage Personal factors • Race • Income • Religious/spiritual beliefs (fatalism) • Age • Concern about experimental nature of the trial • Perception of personal benefit Social factors • Physician’s advice • Altruism • Friends or relatives who had participated in a CT System factors • Clinic site • Distance to the clinic and associated travel costs Agrawal et al. (2006) 163 patients who had Survey Structured in-person interviews using investigator- Alternative treatment options Personal factors • Desire to do something to fight cancer 57 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions USA consented to participate in a phase I cancer study developed 61-item survey Pressure to participate Prognosis Understanding of protocol Motivations for participation Risk/benefit preferences for cancer treatment Information-gathering behavior Sociodemographic characteristics Performance status using the Eastern Cooperative Oncology Group (ECOG) scale • Education Social factors • family pressure Significant factors • education and family pressure (more education related to less family pressure to participate) Baggstrom et al. (2011) USA 183 patients with non-small cell lung cancer who were eligible for CT based on histology and stage of disease Retrospective chart review Investigator reviewed factors associated with declining trials Demographics: age, gender, ethnicity, insurance status Medical: histology, stage, performance status, comorbid conditions, eligibility for trial following individual case review Descriptive factors Personal factors • Patient refusal (no other reason given) • Lack of compliance with therapy System factors • Ineligible for trial after individual review conducted • Lack of transportation • Distance from cancer centre • Insurance issues Significant factors • No significant difference in characteristics between those who participated and those who declined, including gender, age, ethnicity, histology, stage or insurance status Biedrzycki (2011) USA 197 patients with advanced gastrointestinal cancer Survey and retrospective review Descriptive, cross-sectional research design with one investigator-developed instrument, eight instruments from published research, and medical record review Sociodemographics: age, education level, race, sex Disease context: cancer stage, symptom burden Personal factors: hope, quality of life, preference for research decision control, understanding risks, information Social factors: trust in healthcare system, trust in healthcare professional Personal factors • Age (older patients more likely to accept CT) • Type of cancer (pancreatic patients more likely to enroll in CT compared to colorectal cancer patients) In logistic regression, all13 factors of the Research Decision Making Model predicted enrollment when included together Personal factors • cancer stage, • age, 58 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions • symptom burden, • education level, • race, • gender, • hope, • quality of life, • perceived risks and benefits, • preference for research decision making control, Social factors: • trust in healthcare system, • trust in healthcare professional, System factors • adequacy of research information Catania et al. (2008) Italy 102 patients affected with advanced breast or lung cancer Survey Investigator-developed 17-item questionnaire Demographics: sex, pathology, region of origin, level of education, type of work, living situation, family experience with cancer Accept/refuse CT Knowledge of CTs: new possibility of cure for themselves and for other patients in the future Motivations for participating in CTs: trust in doctors/institution, hope, perception of lack of alternatives Fears/prejudices: negative attitudes towards doctors and/or research, fear around receiving experiemental drug System factors • Distance from clinic (if asked if distance from clinic would influence their decision to participate) Catt et al. (2011) UK 40 patients who were offered a phase 1 trial Survey Descriptive (accepters vs. decliners), one study-specific questionnaire and two standardized questionnaires Reasons for participating in CTs Personal factors: optimism, general psychological well-being Primary reason for participating (given by at least 8% of respondents) Personal factors • Personal benefit • Best treatment option available • Hope • Thought that trial was only option available Social factors • Wanting to help with research System factors • More intensive follow-up Davison et al. (2007) Canada 122 prostate cancer patients who had never been asked to Survey Cross-sectional; investigator-developed 30-item questionnaire Demographics: age, education, marital status, employment, ethnicity, household income Medical: treatment status, time since diagnosis, PSA level, Results of Factor Analysis Awareness subscale Social factors 59 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions participate in a CT Gleason score, clinical stage Awareness factors: the recruitment process, informed consent, advice from physicians, and altruistic beliefs Acceptability factors: Impact of participation on current treatment preference, potential study side, and randomization Access factors: transportation, time, related costs, age, and influence of family and friends • Recruitment • Physician advice • Altruism Acceptability subscale Personal factors • Assessment of personal risks and benefits Access subscale Personal factors • Age Social factors • Family and friends influence System factors • Transportation • Costs • Time Personal factors associated with subscale responses • Stage of cancer • Treatment decision • Education Ellis et al. (2001) Australia 545 women attending a breast clinic for screening mammography or diagnostic assessment plus women with newly diagnosed breast cancer Survey Investigator-developed cross-sectional survey, which included the 14-item Hospital Anxiety and Depression Scale (HADS) (Zigmond and Snaith 1983) Demographics: age, marital status, education, occupation, ethnicity, and medical/allied health training Anxiety and depression Information preferences and decision-making involvement Knowledge about CTs Attitudes towards CTs: positive and negative aspects of CT participation, practical issues/inconvenience associated with CTs, attitudes towards altruism Willingness to participate in CTs Additional reasons Personal factors • Age • Decision making style • Time since diagnosis • Perception of positive aspects of participating • Perception of negative aspects of participating Social factors • Altruism System factors • Protocol burden:(time, inconvenience) Go et al. (2006) USA 1012 women and men newly diagnosed cancer patients at community-based cancer Retrospective review of records Retrospective chart review Demographics: gender, age, race, insurance status Protocol enrolment decisions Reasons for not enrolling a patient into a CT: protocol limitation, physician triage, patient decision Personal factors • Gender (more women than men were found to have CTs available to them; more women than men were enrolled in CTs) • Age (elderly patients were less likely to 60 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions centres who had not yet received radiotherapy or systemic treatment be enrolled) Social factors N/A System factors • CT availability CT eligibility Gross et al. (2004) USA 737 older women with breast cancer who had enrolled in a CT, and 7384 who had not enrolled in a CT (“community patients”) Case-control Case-control study Sociodemographics: patient age, race/ethnicity, zip code, date of trial entry, and method of payment Personal factors • Residing in high-poverty zip code • Residing in area with high population density • Age • Race • Medicaid • Proportion of county managed care penetration • Proportion of unemployment Social factors • SES System factors • Proximity to research centre • whether county had a teaching hospital Jenkins and Fallowfield (2000) UK 204 newly diagnosed and relapsed cancer patients Survey Investigator-developed cross-sectional design, postal questionnaire Demographics: age, gender Medical: cancer site, previous CT experience, previous experience of chemotherapy Anxiety Receive information sheet Accept/refuse CT; trial category Reasons for accepting/declining to enter a CT Personal factors • Fear of randomization • Perception that trial is best treatment available • Benefits of trial outweight side effects Social factors • Altruism (for both doctor and other patients) • Trust in the doctor System factors • Type of treatment offered (radiation/chemotherapy vs. hormone treatments) • More likely to participate in trials with active treatment or placebo than no treatment group Jones et al. (2006) Canada 100 ambulatory cancer patients Mixed-methods Cross-sectional design, mixed-methods, semi-structured interview survey Demographics: age, gender, education level Knowledge about CTs History of CT participation Personal factors • Convenience (other life obligations and extra visits) Social factors • Altruism • Express gratitude for good care and 61 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions Reasons for or against participating Willingness to participate in future CTs treatment • Confidence in the study (including confidence in the information and confidence in the physician) System factors • Access to better care and personal benefits (i.e. new drugs) Kemeny et al. (2003) USA 77 young and 77 old women with breast cancer and 44 physicians Case-control and survey Case-control design, investigator-developed questionnaire Sociodemographics: age, ethnicity, education, marital status, employment status, education level, household composition, occupation Medical: disease stage, functional limitations, comorbid conditions and depression Planned accrual/actual accrual Reasons for or against participating: treatment side-effects, outcome and cost, research-specific issues, the consent form, their doctor’s and family’s wishes, altruism Physicians’ reasons why their patients had not been offered participation: treatment toxicity, comorbid conditions, one treatment arm considered inadequate, lack of patient support, lack of available transportation, and lack of patient’s understanding of study requirements Physicians’ reasons why their patients had refused participation: treatment side-effects, outcome and cost, research-specific issues, the consent form, their doctor’s and family’s wishes, altruism Personal factors • Age (older patients are less likely to be offered a trial by physician) • Comorbidities (patients with greater comorbidities less likely to be offered a trial by physician) • Stage of cancer • Improve health Kim et al. (2008) Korea 524 cancer patients who initiated their first cycle of chemotherapy Survey Prospective study design, investigator-developed questionnaire Sociodemographics: age, sex, educational degree, marital status, economic status, distance from clinic, possession of private cancer insurance, religion Medical: diagnosis, disease status Patients’ awareness of cancer CTs and sources of information for cancer CTs Perceptions of patients and physicians associated with the benefits of CTs compared with conventional therapy Patient willingness to participate in CTs Reasons for participating or not in CTs Awareness of CTs Personal factors • Age • Education • Socioeconomic status • Health insurance Personal factors • Gender • Awareness of CTs • Benefits of CTs Klabunde et al. 573 cancer Retrospectiv Standardized log sheet Demographic and clinical characteristics of patients eligible Personal factors 62 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions (1999) USA patients successfully enrolled in non-bone marrow transplant cancer treatment or treatment-related protocols e review of records for CT enrollment: facility type, sex, race, insurance coverage, disease status, cancer site, cancer stage, trial phase • Stage of cancer • Type of cancer • Type of health insurance • Time since diagnosis System factors • Facility type (academic medical centre) Lara et al. (2001) USA 276 cancer patients Survey Cross-section design, investigator-developed questionnaire Patient characteristics: sex, age, race, insurance status, primary referral source CT eligibility: physician triage, protocol availability, patient eligibility Patient decision to accept/refuse CT Reasons for patient non participation: desire for other treatment, distance from clinic, no reason given, insurance denial, fear of randomization, other Personal factors • Type of health insurance LaVallie et al. (2008) USA 112 older American Indian and Alaska Native adults Survey Cross-sectional design, 37-item survey (Buchwald et al. 2006; Noe et al. 2005) including a hypothetical vignette Demographics: age, sex, education, current residence, marital status, previous participation in research, telephone in home Willingness to participate in CTs: institutional sponsorship, community involvement, human subjects’ issues, and convenience Factors that decreased or increased participation: convenience, community involvement, side-effects, friend or family member had this type of cancer, helping the community, informed consent, randomization, chance for cure, family support Personal factors • Threat that confidentiality would be breached • Personal experience with cancer Social factors • Physician with experience treating American Indians/Alaska Natives • American Indian/Alaska Native researcher • Family support System factors • Distance to study site Li et al (2010) China 578 cancer patient-family member pairs Survey Two study-specific questionnaires, one for patients and one for family members Demographics: sex, age, marital status, income, level of education, occupation, religiousness Clinical information: time since initial diagnosis, type of primary cancer, disease stage, therapies accepted, Eastern Cooperative Oncology Group performance status Experience with cancer clinical trials: Attitudes towards clinical trial enrollment: Personal factors • Level of experience with CT in early stage patients (those who participated in a CT and their relatives, those with prior therapy and their relatives, and those who had never heard of CTs and their relatives were more willing to participate) • Level of experience with CT in advanced stage patients (patients with prior therapy and their relatives, and patients who had never heard of CTs) • Education in advanced patients who failed to respond to therapy 63 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions • Level of experience with CT in advanced stage patients who failed to respond to therapy (those who had never heard of CTs were more willing to participate) Ling et al. (2000) UK 1206 cancer patients referred for consideration of entry into a palliative care CT Survey Prospective survey of standardized trial referral forms Demographics: patient name, age, hospital number, sex, diagnosis and the trial for which the patient was being considered Factors influencing decision regarding whether an eligible patient was appropriate or well enough to enter a CT: patients’ general performance status, patient understanding of requirements, patients’ likely prognosis and their ability to complete all study procedures Reasons stated for eligible patients declining a CT Personal factors • Patient prefers to wait • Patient unwell • Lack of interest • Objection to medication • Unwilling to complete forms • Unable to give informed consent • Too many pills • Anxiety • Previous experience Social factors • Family objection • Physician objection System factors • Distance • Research nurse unavailable • Eligibility criteria Llewellyn-Thomas et al. (1991) Canada 60 colorectal cancer patients with or without metastatic disease Probability trade-off; survey Questionnaire followed by one open-ended question. Decision Making Preference Questionnaire adapted from the literature. Two tasks developed from earlier work and adapted to the current study. Demographic data: age, sex Diagnostic and treatment characteristics: cancer diagnosis, current treatment Preferences for participating in decision making: physician dominant, patient participate Attitudes about CT entry: attitudes towards randomization, preferences regarding the risks and benefits associated with trial treatment Personal factors • Desire to avoid toxicity • Preference to prolong survival Social factors • Preference for physician to make treatment decisions • Advice of physician Mancini et al. (2007) France 267 non-metastatic breast cancer patients invited to participate in a RCT and 1888 patients not invited to participate in a RCT Survey Longitudinal design, mailed investigator-developed questionnaires (2, 6, and 12 months after being included in the study); standardized scales measuring depressive symptoms (e.g., CES-D, STAI-8, CASC) Sociodemographic data: gender, age, marital status, number of children, education, occupation, and perceived economic problems Characteristics upon inclusion of the patient in the study: centre consulted, chemotherapy start date, consultation duration Depressive symptoms Anxiety Social support Personal factors • Lack of fear about randomization • Previous trial experience • Preference to make own decisions about participation Social factors • Altruism (to other patients and physicians) • Inability to refuse System factors N/A 64 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions Satisfaction with care Preferences about decision making: shared or personal Willingness to participate in a hypothetical CT Reasons for agreement/refusal to be enrolled in CT: randomization acceptable or not, inability to refuse, and altruistic motivations, satisfaction with doctors’ communication Melisko et al. (2005) USA 297 patients with newly diagnosed or recurrent breast cancer and 127 physicians from private and academic centres from specialties of oncology Survey Longitudinal design, investigator-developed 28-item questionnaire (patients); 31-item questionnaire (physicians) Patient characteristics: year of diagnosis, age, race, educational level, stage of disease, marital status, and family history of breast cancer Patient survey factors: access to CT information, safety issues, logistics, confidentiality, attitude toward randomization, and feelings about investigation of complementary and alternative medicine Physician demographics: age, practice type, practice setting Physician survey factors: impact of CT on clinical practice, costs to the patient, and protocol merit Personal factors • Time for extra tests and visits • Concern about loss of control of treatments if in trial and interfere with choosing own treatments (earlier stage) Social factors • Prefer to wait for physician to introduce about trial (ethnicity) System factors • Transportation, child care, and income loss (greater for younger women) Physicians Personal factors • CTs stressful for patients • Control arm has inadequate therapy • Concern about side effects • CTs interfere with usual care • Lack of information about CTs to provide patients • CTs delay treatment System factors • Enrolling patients in CTs requires extra staff time and adds cost to practice • Strict eligibility criteria • Clinical trial process slow to answer questions as new treatments emerge Meropol et al. (2007) USA 137 medical oncologists and 170 cancer patients Survey Cross-sectional design, mailed questionnaire; investigator-developed questionnaire informed by the C-SHIP framework (Miller et al. 1999; Miller et al. 1998) Sociodemographic information: gender, age, ethnicity, and racial group, education level, marital status, employment status, and health insurance status Medical practice characteristics: professional setting, board certifications, number of individual patients seen in the past 12 months, number of new patients entered into practice in the Physicians Personal factors • Majority believe that patients would benefit from CTs, but community-based oncologists less likely to agree • Fear of randomization • Lack of patient knowledge about CTs 65 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions past 12 months, number of treatment CTs currently active at a site, number of patients enrolled in treatment CTs in the past 12 months Physician interest in participating in treatment CTs Physician barriers to CTs: staffing, strict eligibility requirements, financial constraints, time constraints Medical history: cancer diagnosis, age at diagnosis, and disease status Patient awareness of CTs interest and willingness to participate in CTs Cognitive-affective barriers to CT participation: uncomfortable with randomization, lack of trust and fear of being a guinea pig, fear of receiving a placebo, fear of side-effects Practical barriers to CT participation: logistical barriers (i.e. transportation), access Patients Personal factors • Concern over side effects from experimental treatment • Fear of randomization • Fear of side effects • Race (knowledge of CTs) • Education (knowledge and interest in CTs) Social factors N/A System factors • Logistical barriers (i.e. transportation to centre) Nurgat et al. (2005) UK 38 patients with advanced or metastatic cancer Questionnaire Cross-sectional design, questionnaire adapted from Daugherty et al. 1995 (includes both open-ended and close-ended questions) Demographics: gender, age, marital status, education, employment Medical history: tumour type, previous chemotherapy, previous CT participation Patient health status Patient expectations about benefits of CT participation Quality of life vs long term survival Sources of information: when sought more information about illness/treatment options, who patients contacted for more information, satisfaction with amount of information, understanding of prognosis Patient comprehension: given informed consent, having understood all or most of the trial information given to them, explanation of trial as research by research team, time to think things over, explanation of reasonable alternatives Influences on patient decision making for participation in Personal factors • Perception of personal benefit • Obtain more information • Surviving as long as possible • Obtaining more information about disease • Lack of alternatives (phase I vs. phase III studies and patients who previously had chemotherapy) Social factors • Benefit others • Trust in doctors • Trust in nurses • Altruism System factors • More monitoring • More follow-up and better standard of care • Access to better treatment 66 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions CTs: personal choice, with whom did you discuss, how difficult was the decision, would you say you made your own decision Reasons for participating in a CT: hope of remission, help me/help others, improve health, reduce tumour, trust in doctor, trust in nurses, being treated by the latest treatment available, better standard of care and closer follow-up, being likely to obtain more information about my condition, and closer monitoring of patients in trials Roberson (1994) USA 8 African Americans, 10 Hispanics, and 10 Native American cancer survivors and asymptomatic individuals Mixed-methods Cross-sectional design, investigator-developed telephone surveys (10 items, 7 open-ended) Patient knowledge and participation in CTs Benefits of participation: help find a cure, help others, assist with medical coverage, and educate families Why patients do not participate: fear, lack of information, mistrust of being treated like guinea pigs, ethnic background, do not like to get involved Role of culture in CT participation: mistrust of white people, religion, language, beliefs, moral values, and folk medicine Suggestions about recruitment and retention in CTs Personal factors • Lack of information • Fear of being a guinea pig • Mistrust • Ethnic background Social factors • Help others • Find a cure • Educate families System factors • Assist with medical coverage Sabesan et al. (2011) Australia 178 medical oncology clinic patients Survey Used validated questionnaires from Ellis et al. (2001), for questions about rural factors developed and pilot tested questions, descriptive analysis Demographics: age, gender, location (determined by postal code), educational level Medical history: prior cancers, history of cancer therapy, stage of cancer, previous experience or knowledge of cancer clinical trials Patient knowledge of clinical trials: Patients’ willingness to participate in a randomized clinical trial: Patient’s reasons for participating or not participating in clinical trials: Rural factors: inconvenience of travel, cost of travel, loss of work hour, need for family or friends to accompany, need for organizing carers for dependent children and adults, maximum acceptable number of doctor’s visits,, maximum acceptable number of blood tests Reasons for accepting trial Personal factors • Personal benefit from trial (66.% of participants) • Better chance at cure (64.0%) • Any treatment may help me (50.0%) • No other options (36.5%) • Best treatment (34.3%) Social factors • Further medical research (70.8%) • Altruism (67.4%) • Family would benefit if they get cancer in the future (62.9%) System factors • Receive more information (34.3%) • Access to unavailable treatments (43.8%) Reasons for declining trial Personal factors • Treatment may be too severe for me (39.3%) • Treatment may be worse on a clinical 67 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions trial (20.2%) • Interferes with daily activities (16.9%) • Clinical trial feels like a gamble (16.3%) • Fear of being a guinea pig (15.2%) • Extra inconvenience (13.5%) • Less say in what happens to me (12.9%) Social factors • Trial has negative effects on family (24.2%) • Doctor would be more interested in trial than in me (12.9%) System factors • Doctor may not know as much about trial treatment (18.0%) Differences between urban and rural patients System factors • Inconvenience of travel (41.9% for rural and remote vs. 9.7% for urban) • Cost of travel (41.1% for rural and remote vs. 23.5% for urban) Sateren et al. (2002) USA 24,332 patients accrued to National Cancer Institute (NCI)-sponsored cancer treatment trials Case-control Combined census data and NCI Surveillance, Epidemiology, and End Results incidence data Sociodemographics: accrual data: age, sex, race/ethnicity, insurance status, state market penetration by health maintenance organizations, place of residence, income, presence of oncology specialists, presence of hospitals with an approved multi-disciplinary cancer program County data: income, education, poverty, and unemployment Personal factors • Race Social Factors • Socioeconomic status System factors • Type of medical insurance • State of residence • Number of oncologists in state Solomon et al. (2003) Australia 100 patients with colorectal cancer admitted for surgery at a tertiary center; 43 colorectal surgeons and 103 medical oncologists Survey Structured face-to-face interviews with patients using questionnaire developed from previous research and pilot work; mailed questionnaires (to surgeons and medical oncologists) Patient preferences and willingness to enter a hypothetical CT: willing/unwilling to trade longevity and willing/unwilling to gamble mortality Patient willingness to enter into hypothetical CT Reasons for patient refusal to enter into hypothetical CT: dislike the concept of CTs, dislike idea of randomization, concern with the different size of the operations, concerned with risk of cancer recurrence, risk of complications of one therapy, concerned that quality of life will be too different with one therapy, others will not benefit from the results of the research, treatment given does not offer the best chance of recovery Personal factors • Type of participant (oncologists more likely to enter a CT than patients and surgeons) • Type of surgery (patients) • Quality of life (patients) • Fear of randomization (patients) • Risk of cancer reoccurrence (patients) Truong et al. (2011) USA 207 cancer patients and 49 parents of Survey Questions modeled after surveys by Daugherty et al. (1995) and the Advisory Demographics: ethnicity, education, marital status, first language Personal factors • Poor prognosis (negatively associated with altruism) 68 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions pediatric cancer patients Committee on Human Radiation Experiments (1996) Medical history: disease status at time of study entry, estimated probability of 5-year survival Outcome: altruism • Relapsed/progressive disease (negatively associated with altruism) • Gender (females more likely to be altruistic) • In univariate models, altruism as the motivating reason was associated with phase (phase III), gender (female) age (older), type of respondent (parent), prognosis (poor), stage of disease (relapsed/progressive) • In multivariate models only poor prognosis associated with altruism as primary motivation for trial acceptance Social factors • Altruism (help other cancer patients or to advance science) System factors • Phase of trial (phase III more likely to be altruistic compared to phase I) in both univariate and multivariate models Wallington et al. (2012) USA 944 patients from safety-net clinics serving largely immigrant Latinos from Central, South, and North America Survey Cross-sectional survey with quota sampling by age and gender Information-seeking behavior: from newspapers, magazines, television, radio, the internet, health professionals Personal cancer risk: Cancer worry: Information self-efficacy: Demographic s: age, gender, marital status, country of origin, highest grade completed, number of years in the US, language acculturation Knowledge about CTs Personal factors • Using internet for health information • Education • Marital status • Trust in health information services • Perceived likelihood of developing cancer Intention to join a trial Personal factors • Trust in health information • Worry about developing cancer • Information self-efficacy • Knowledge of what a CT is Social factors • Using cancer information line Wang et al. (2011) Taiwan 184 cancer patients Survey Cross-sectional survey developed by the investigators, based on pilot interviews, investigators’ experience, and published research Demographics: age, gender, education, occupation, religious beliefs, income, marital status Willingness to participatee: Motivational factors: medical issues,, subjective viewpoint, family and friends Barrier factors: Personal factors • Medical issues: especially chance of cure • Subjective viewpoint: contributing to medical research • Concerns about adverse effects from clinical trial drugs (barrier) • Patients’ past experience with clinical trials Social factors 69 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions Needs: Expectations: • Family and friends’ previous experience • Discussions with physicians System factors • Information from nurses Weckstein et al. (2011) USA 120 cancer patients who were offered a trial but declined to participate Survey Cross-sectional survey developed by investigators based on experience Barriers to participating: Personal factors • Adverse effects • Fear of randomization • Overwhelmed • Time Social factors • Physician recommended not participating System factors • Cost/no insurance Wright et al. (2004) Canada 189 cancer patients, their physicians (n=28), and clinical research associates (n=12) Survey Single-institution observational cohort study design; questionnaires with items based on factors from the medical literature and augmented by focus group response Patient sociodemographics: age, gender, living with spouse, level of education, travel distance to cancer centre, household annual income Patient involvement and levels of decision making: to what extent patients are involved in the decision, when the decision is made, factors that influence their entry decision (e.g. perceived personal benefit, when they believed the CRA helped with their decision, helping future patients, supportive of physicians or CRAs, more time spent discussing consent issues) Physician, patient, and clinical research associate factors influencing recruitment to CTs: general, trial-specific, encounter-specific Patient, physician, and CRA attitudes Personal factors • Personal benefit • Decision easy/hard to make • Expect to do well with conventional therapy • Study asking important question (physician) • Belief in no serious side effects (clinical research associate) Social factors • Clinical research associate/physician helped with decision • Time spent discussing consent • Doctor/nurse recommended entering study System factors N/A Factors MDs Identified Affecting Cancer Patient Accrual • !Kornblith et al. (2002) USA 156 physicians involved in the treatment of patients with breast cancer Survey Cross-sectional design, investigator-developed survey Physician characteristics: age, gender, ethnicity, medical specialty, the setting of their medical practice, and the proportion of their case load that involved patients age > 65 years Physician Survey of Barriers to Accrual: protocol requirements, treatment specific issues, social support, logistic issues, physician attitudes, medical and cognitive characteristics of older patients Personal factors • Comorbid!conditions!• Toxicities!• Patients!not!understanding!trial!System factors • Transportation!needs! Siminoff et al. (2000) USA 147 physicians (107 surgeons and 40 medical Retrospective review of records; case Chart reviews and structured in-person interviews with Chart review data: medical, disease, and demographic characteristics of patients technically eligible to participate in at least one of the phase III breast cancer treatment trials open Personal factors-surgeons • Race!• Practice!type! 70 Author, date, and country Study sample Strategy Data collection method Factors explored Factors influencing CT decisions oncologists) discussed 245 newly diagnosed breast cancer patient cases (patients eligible to participate in a phase III breast cancer treatment trial) study physicians for recruitment Physician case-based and attitudinal information about adjuvant therapy and phase III CTs Physician elicited decision criteria: general history of the patient’s breast cancer illness, examination of the treatment options the physician considered (including knowledge of whether the patient was eligible for a CT and if one was offered), whether there was a discussion of more than one treatment option with the patient, a visual analog scale to rate the strength of the physicians’ treatment recommendations, review of the physician’s rationale for his/her treatment recommendations, items considered (treatment benefits and side effects), decision-making style, final treatment choice, his/her impression of the patient’s reaction to the various treatment recommendations, the physician’s impression of the character and/or emotional status of the patient and ability to understand the information being conveyed, and the physician’s previous treatment relationship with the patient Physicians’ attitudes, knowledge, and general practices concerning CTs: perceived barriers to patient participation and their own participation Physician sociodemographics and professional characteristics: age, sex, ethnicity, hospital and cooperative trial group affiliations, training, and practice settings • %!of!practice!breast!cancer!patients!• Cooperative!group’s!support!• Patient!age!Personal factors-oncologists • Practice!type!• Number!of!hospital!affiliations!• Number!of!breast!cancer!patients!per!year!• Patient’s!family!history!of!breast!cancer!• Better!prognosis!• Finding!explaining!medical!uncertainty!as!part!of!trials!difficult!Social factors • Physician!involvement!in!decision!making!• Patient!delay!in!seeking!treatment!(surgeons)!• Patient!preference!for!tamoxifen!(surgeons)!• Patient!needed!more!time!to!make!decision!about!adjuvant!therapy!(oncologist)!• Patient!more!involved!in!decision!making!(oncologist)!• Discussed!chemotherapy!only!with!surgeon!(oncologists)!• Involvement!of!patient!and!clinician!in!decision!making!(oncologist)!• Referral!patterns!(surgeons)!• Level!of!patient!interest!(both)!• Comfortable!talking!to!patient!(both)!System factors • Patient’s!type!of!insurance!(oncologists)!• Eligibility!criteria!(both)!• Too!much!paperwork!associated!with!enrolling!patient!on!trial!(oncologists)!• Extra!work!and!expense!enrolling!patients!(oncologists)!! 71 Table 2.3 – Qualitative Studies Reviewed Author, date, and country Study sample Methodology and data collection method Purpose Cox and Avis (1996) UK 7 patients (5 men, 2 women) who agreed to Phase I and II trial participation Qualitative, longitudinal, semi-structured interviews To explore the psychosocial aspects of participation in early anti-cancer drug trials from the perspective of the patient Coyne et al. (2004) USA 17 patients (2 male, 15 female), variety of cancer diagnoses Qualitative, in-depth interviews To identify the attitudes and perceptions of rural cancer patients regarding CT participation from the patients’ perspective Ellis and Butow (1998) Australia 21 mothers or grandmothers of children attending a local primary school and 20 breast cancer patients Qualitative, focus groups To explore the knowledge of, and attitudes towards, randomized clinical trials among women in the community and breast cancer patients Eng et al. (2005) Canada 11 men with prostate cancer Qualitative, semi-structured interviews To examine the difference and similarities between the reasons for accepting and declining participation in a two-arm active treatment RCT comparing external beam radiation therapy versus cryotherapy Grunfeld et al. (2002) Canada 29 clinical research associates or data manager participants Qualitative, focus groups To explore the views of clinical research associates on barriers and facilitators to accrual Huizinga et al. (1998) Netherlands 14 cancer patients (12 women, mostly advanced breast cancer, 2 men) Qualitative, semi-structured interviews To gain insight into the decision-making process that patients go through when asked to participate in a cancer CT Kohara and Inoue (2010) Japan 25 cancer patients (mostly colon, lung, breast) Qualitative (grounded theory), semi-structured interviews and unstructured observations To reveal the decision-making process in patients considering participation in Phase I cancer CTs Kvale et al. (2010) USA 4 older white cancer patients (3 men, 1 woman), variety of cancer diagnoses and stages Qualitative (hermeneutic phenomenology) open-ended interviews To explore the experiences of older cancer patients with Phase I CTs. Madsen et al. (2007) Denmark 24 female patients with premenopausal breast cancer and 5 patients with advanced ovarian cancer Qualitative, in-depth interviews To explore a broader description and understanding of the meanings assigned to patients’ lived experiences during their treatment courses within or outside a trial setting Mills et al. (2011) UK 93 men with localized prostate cancer Qualitative analysis of audiotaped recruitment appointments to an RCT To explore how patients’ treatment preferences were expressed and justified during recruitment to a randomized controlled trial (RCT) and how they influenced participation and treatment decisions. 72 Author, date, and country Study sample Methodology and data collection method Purpose Schaefer et al. (2001) USA 26 women considering participation in a breast cancer prevention CT Qualitative (grounded theory), conversational-style interview To describe the process of decision making by women considering participation in a breast cancer prevention trial Schutta et al. (2000) USA 8 patients (3 men, 5 women) with various cancer diagnoses who were currently participating in a Phase I CT Qualitative, focus groups To explore factors that influence an individual’s decision to participate in Phase I cancer CTs Shah et al. (2012) USA 46 patients with localized prostate cancer Mixed methods, open-ended questions and a questionnaire To understand patient’s willingness to participate in an RCT comparing intensity-modulated radiotherapy to proton beam therapy. Shannon-Dorcy & Drevdahl (2011) USA 25 patients enrolled in early-stage 2 hematopoietic cell transplant research studies and 20 family caregivers Qualitative, semi-structured interviews. To examine how patients and their family caregivers decide to participate in hematopoietic cell transplants at a US cancer referral centre Stevens and Ahmedzai (2004) UK 22 newly diagnosed breast cancer patients who had declined entry into a Phase III CT Qualitative, longitudinal, in-depth interviews To explore the reasons why breast cancer patients decline entry into RCTs of adjuvant cancer therapy 73 Chapter Three: Methodology Given the limited understanding of the underlying decision-making processes and the experience of making clinical trial (CT) decisions from the patient’s perspective (see Chapter Two), exploration of how personal, social and structural factors interact and influence patients’ decision making about CTs is required. This involves uncovering the socio-political context of CT decisions, including inherent power differentials and inequities that may unduly influence patients’ decision-making processes. The two empirical studies included within this dissertation are meant to address this gap. The first study, the results of which are presented in Chapter Four, describes CT personnel’s perspectives of the influences on cancer patients’ CT decision-making process and ability to exercise relational autonomy, and best practices to support relational autonomy. The second study, the results of which are presented in Chapter Five, develops a theory of how cancer patients make CTs decisions, which is enhanced by their support persons’ perspectives. The purpose of this chapter is to describe the methodology and methods used in both empirical chapters. It is divided into the following sections: study design, theoretical framework and use of the literature, study population, recruitment strategies, sampling strategies, interviews, data analysis, validity and reliability and ethical considerations. Study Design Two qualitative approaches were used in this dissertation: interpretive description and grounded theory. Qualitative methodologies are appropriate when uncovering subjective experiences as they allow participants to define what is of interest, rather than starting with predetermined concepts or hypotheses, as is typical of quantitative methods (Creswell 1994). However, the use of a priori theory (such as relational autononomy in this study) in qualititve 74 research has been encouraged by some methodologists because it provides an initial framework that informs the research question and study design (Morse 2003). In addition to qualitative methodology, a relational autonomy and critical feminist perspective was applied to the analysis. Feminist theory is largely concerned with marginalized persons, particularly the domination and subordination of women in society. A critical feminist perspective encompasses not only a focus on gender but also other sources of social and cultural inequity that encompass both men and women. Moreover, critical feminist theory emphasizes the emancipation of those who have been oppressed and the transformation of oppressive social structures to promote social justice (Kushner & Morrow 2003). It was, therefore, well suited to the objectives of this research study and the ultimate aim of providing a foundation for the development of strategies to enhance patients’ relational autonomy within the CT decision-making process. Interpretive Description The qualitative methodology guiding the exploration of CT personnel’s perspectives, which is presented in Chapter Four, was interpretive description (Thorne 2008; Thorne, Kirkham & O’Flynn-Magee 2004). This methodology provided opportunities to explore CT personnel’s beliefs about the personal, social and structural influences on cancer patients’ CT decision-making process and how patients exercise relational autonomy. It also guided the categorization of themes related to best practices used by CT personnel to support cancer patients in exercising their relational autonomy within this process. Interpretive description is a form of inquiry that reveals knowledge relevant to nursing practice. It recognizes the contextual and shared nature of realities, fitting well with relational autonomy theory, where persons are understood as inherently social and 75 interdependent beings (Thorne 2008). This methodology allows for the identification of themes and patterns from the data in relation to clinical phenomena (e.g., CT decision-making) and for critical interpretation and explanation of the data, as required for a critical feminist analysis of the factors influencing cancer patients’ relational autonomy. Grounded Theory The qualitative methodology used to construct the findings presented in Chapter Five of this dissertation was grounded theory (Strauss & Corbin 1998). Grounded theory was the most applicable methodology because the goal for this chapter was to develop a theory of how patients with cancer make decisions about CT participation. Grounded theory draws on the tenets of symbolic interactionalism and is an appropriate methodology for exploring decision making and relational autonomy because of its emphasis on change, complex social processes and the active role of individuals and their social networks in shaping the world in which they live (Charmaz 2006). Rather than “discovering” grounded theory, this chapter is more closely aligned with an interpretive approach to theory development that recognizes the social construction of reality “through our past and present involvements and interactions with people, perspectives, and research practices” (Charmaz 2006, p. 10). This perspective acknowledges how research participants’ constructions of reality and grounded theorists’ meanings and views influence the interpretation and building of theory. Grounded theory provides a set of guidelines (described below, in sampling strategies and data analysis sections) to direct data collection and analysis, generating an overall theory that is “grounded” in the data. This methodology also examines processes; thus, it is well suited to generating an overall theory of patients’ CT decision-making process. Previous health research with cancer patients in the context of complementary and alternative 76 treatment decision making has also demonstrated the suitability and flexibility of grounded theory in exploring other forms of treatment decision making (Balneaves, Truant, Kelly, Verhoef, & Davison 2007; Balneaves, Bell, Truant, Verhoef, & Davison 2010). Theoretical Framework and Use of the Literature Glaser and Strauss (1967), the founders of grounded theory, differ in their views on researchers’ legitimate use of the literature and background knowledge in theory development. Glaser does not accept a review of literature prior to developing a theory because it could influence the inductive process of theory emergence, thus increasing the risk of a theory evolving from previous knowledge rather than the data. In contrast, Strauss allows for self-experiences and literature to inform initial understandings of a theory and guide theoretical sensitivity and concepts (Heath & Cowley 2004). Similarly, interpretive descriptive techniques generally discourage the use of a priori frameworks, such as relational autonomy, and the extant literature because it can interfere with the identification of themes. However, some have argued that the use of current literature and/or theory to inform an initial analytic scaffold is allowed (Thorne, Kirkham & O’Flynn-Magee 2004; Morse 2003). Morse (2003) believes including conceptual frameworks such as relational autonomy theory in qualitative inquiry supports the development of new understandings as it provides a “scaffold” for building upon previous assumptions, knowledge and research. Without compromising the validity of a qualitative study, conceptual frameworks allow investigators to acknowledge the assumptions guiding the research, identify known concepts, and inductively fill in gaps in knowledge. My approach in both the findings chapters is more aligned with Strauss, although I 77 incorporated elements of Glaser’s approach during the analysis phases as will be described below. In alignment with Strauss, and as discussed in Chapter One, the theoretical lens of relational autonomy was used to explore the ethical dimensions of CT decision making from the cancer patient’s perspective. Relational autonomy theory sensitized me to understanding power relationships and inequalities within the relevant cancer CT decision-making literature prior to embarking upon data collection for the interpretive descriptive and grounded theory studies. This allowed me to determine the scope of the problem at hand, identify known gaps in knowledge, target a study population and appropriately frame my research questions. I did, however, refrain from returning to the relational autonomy and cancer patient CT decision-making literature when identifying emerging themes and patterns that expanded upon personal, social and structural categories for the interpretive descriptive study. Similarly, when I was developing the grounded theory describing patients’ CT decision-making process and exercising of relational autonomy, I refrained from returning to the literature to further explore emergent concepts. I felt it was important to distance myself at that time so that the theory could emerge and concepts could be related based on the data in front of me. In this way, my methodology became more aligned with Glaser’s approach to grounded theory. In both studies, I positioned myself as open to new meanings – to be found in the data – relative to my initial research questions. However, I did ask additional questions of the data as a result of a relational autonomy (and critical feminist) lens. Cutcliffe argues that the combining of the two approaches (i.e., Glaser and Strauss) in this way can lead to a richer, more complete understanding of the phenomenon under study (Cutcliffe 2000). 78 Study Population The initial target populations for this dissertation were breast and prostate cancer patients. This allowed for an exploration of how relational autonomy was differentially enacted by men and women in the CT decision-making process. The primary inclusion criteria were women and men aged 19 years and older diagnosed with breast or prostate cancer and who had been approached to take part in a CT (Phase 0-IV) recruiting through an urban cancer centre in British Columbia. The study location was selected because of its affiliation with the university and also because it is a regional hub for cancer services with multiple trials ongoing. The selection of this centre allowed for a heterogeneous sample since patients travelled from across the province to access services and trials. The CTs from which patients were recruited were decided on through consultation with the Medical Director of CTs and a Senior Scientist at the cancer centre. Patients were excluded from research participation if they were deemed by CT personnel to be cognitively impaired, too ill to participate or did not speak English. The decision to broaden the sample to include support persons and CT personnel reflects the guiding theoretical framework of relational autonomy, which recognizes the importance of relationships for patients’ self-determination and exercising of capacities for autonomy. Eligible support persons were any individuals over the age of 19 years who had been identified by patients as being involved or consulted in their CT decision-making process. Support persons could include family members, friends, fellow cancer survivors and primary care providers. Eligible CT personnel were any staff or physicians at the cancer centre who were involved in CT recruitment or management. For the purpose of this dissertation, 40 patients (see Table 5.1. in Chapter Five) and 79 11 support persons were recruited. Twelve CT personnel were also invited to participate in order to provide insight into the relational nature of the decision-making process and the best practices that support relational autonomy. Based on previous qualitative research (Balneaves et al. 2007; Balneaves et al. 2010; Öhlén, Balneaves, Bottorff, & Brazier 2006), it was anticipated that a combined sample of this size would provide the opportunity to explore a sufficient range of decision-making experiences to uncover common patterns and themes and achieve data saturation in which new data do not reveal new codes or concepts or add to the description of existing concepts. Recruitment Strategies Several recruitment strategies were used to enrol participants into both studies. Nominal financial compensation ($25/interview) was offered to recruited patients and support persons respectively to acknowledge their participation in this study and to cover any related costs (e.g. parking fees). CT Personnel Recruitment Recruitment began with CT personnel since they were the most accessible population. The Medical Director of CTs and a Senior Scientist at the cancer centre compiled a list of CT personnel and researchers involved in cancer CTs. I approached these CT personnel and researchers via an emailed letter of explanation, consent form and invitation to participate in the study (see Appendix IV and V). Cancer Patient Recruitment Two graduate research assistants associated with this study (GRAs) were assigned to breast and prostate clinics at the cancer centre that were actively recruiting patients for CTs. Oncologists asked patients if they would be interested in hearing about the study after their 80 clinical consultation. If patients agreed, a GRA visited the clinic room after the oncologist had left to briefly explain the study and eligibility criteria. The GRA then provided those women and men who were eligible with a study package that included a letter of explanation and consent form, directions to the research office where the interview would take place and a demographic form (see Appendix II, III, IV, V and VI). The GRAs also sought permission to follow up with interested individuals within one to two weeks to provide them time to consider study participation, answer any additional questions and schedule an interview with interested patients. All interested and eligible individuals who contacted the research team (myself or the GRAs) by phone or email were provided with an overview of the study purpose and answers to any remaining questions. An interview was arranged if consent was provided. The GRAs were provided direction with regards to the types of patients to be recruited based on purposeful and theoretical sampling needs (see Chapter Three, Sampling Strategies). Posters about the study were also positioned in high-traffic areas in the cancer centre, such as public elevators. It was thought that this method of recruitment would be particularly useful for those individuals who had declined or withdrawn from a CT, since these patients might not regularly visit the cancer clinics. The posters, however, were not very successful at eliciting interested patients. Upon reflection, this might have been because the posters were overlooked or because patients who had declined or withdrawn from CTs were not interested in participating in any form of research. Support Person Recruitment Patients participating in the study were asked to identify a support person with whom they had discussed their CT decisions. A letter of explanation and consent form (see 81 Appendix IV and V) were forwarded by email to eligible support persons through the patients and those interested were asked to contact the research team for further information. Sampling Strategies Sampling and data analysis for both the interpretive descriptive and grounded theory study occurred in two iterative, closely linked phases. For example, in the initial phase of the grounded theory study, purposive sampling was undertaken to select a variety of participants with regards to their CT decision (i.e. acceptors, decliners and withdrawers) as well as type of cancer (breast and prostate). One bladder cancer patient was unintentionally included in the study due to insufficient screening. Her primary disease site was revealed during the interview and the decision was made to incorporate her perspective into the data analysis because her insights elaborated upon other participants’ (breast and prostate patients’) accounts. Purposive sampling allows the researcher to gather relevant data from diverse groups that will promote a more representative perspective on the phenomenon of interest (Strauss 1998). Attempts were made to recruit patients who represented a variety of backgrounds (with respect to age, education and ethnicity) as these factors may potentially impact relational autonomy and CT decision-making processes (see Chapter Two). A range of CT personnel were also selected based on profession (e.g., oncologist, CT nurse) and gender. Although my sampling strategy was purposive and focused on a variety of characteristics, I was unable to recruit a balanced number of acceptors, decliners and withdrawers into the study. Initially, the number of acceptors far outweighed the number of decliners and withdrawers that were entered into the study. The Medical Director of CTs and a Senior Scientist at the centre were consulted about this difficulty and they assisted in the 82 identification of additional decliners and withdrawers who could be approached for the study. With their help, I was able to recruit more decliners but the number of withdrawers remained low. Both breast and prostate cancer patients, however, were recruited sucessfully into the study. In the tradition of grounded theory, sampling continues until theoretical saturation is achieved, no new data emerges that is relevant to the identified categories, category development is dense and the relationships among categories are well explicated and validated (Strauss 1998). While theoretical saturation was attempted and achieved in many instances, there were still some concepts that remained not fully saturated due to the limits of the sample and time constraints within the context of a Ph.D. dissertation (particularly, in relation to concepts associated with decliners and withdrawers from CTs). Interviews All participants were provided a verbal explanation of the study and then given time to reread the consent form and ask any questions prior to providing consent. After signing the consent form, patients were asked to complete a demographic questionnaire to collect baseline data. This included demographic and disease information (e.g., age, marital status, education and cancer type) (Table 5.1). I conducted in-depth, open-ended interviews (see Appendix VII for interview guide) with each participant (patients, support persons, CT personnel) in space provided by my doctoral supervisor’s (Dr. L. Balneaves) research program. Open-ended interviews suited the qualitative approach because they invite unanticipated stories and statements to emerge that reflect participants’ experiences and beliefs (Charmaz 2006; Thorne 2008). Given the paucity of research that has explored the relationship between cancer 83 patients and support persons in making decisions about CTs, individual instead of dyadic or greater (e.g., triad) interviews were used to collect data about the interpersonal and social processes through which CT decisions are made. It was felt that individual interviews with patients and support persons would allow these participants to freely and confidentially discuss their experiences of CT decision making without creating tension within their personal relationships. However, recognizing that some patients may feel more comfortable participating with their spouse, dyadic interviews were conducted with two patients who indicated a preference to be interviewed with their spouse or support person. These interviews captured both patients’ and support persons’ perspectives on relational autonomy within the context of cancer CTs. Previous research supports this interview approach as being feasible and acceptable to cancer patients and their support persons (Balneaves et al. 2007). Interviews with CT personnel were conducted separately from patients and support persons because their questions had a unique focus and there was a shorter time allotted for the interview. Interviews were first conducted with CT personnel. These included a series of open-ended questions that explored how CT personnel understand the concept of relational autonomy within the context of the CT decision-making and the recruitment process, their knowledge and experience about why patients do or do not participate in CTs, and information surrounding best practices to support patients’ relational autonomy in CT decision-making (see Appendix VIII). For example, CT personnel were asked “Why do you think patients accept/decline/withdraw from clinical trials?” followed by exploratory questions related to the institutional context of patient decision making and how patients can best be supported in their CT decision-making process. 84 Interviews with patients began with a primary question (“How did you arrive at your decision to accept/decline/withdraw participation in the CT?”) followed by more exploratory questions, including what was important to them in their decision-making process, who was involved and influential in their decision making, and what challenges and facilitators did they experience in making the decision. Patients were also asked how they could have been better supported in making CT decisions (see Appendix VIII). Interviews with support persons included a series of open-ended questions that explored how they were involved in patients’ decision-making processes, their perceptions of the challenges and facilitators to CT decision making, and their recommendations regarding improvements to the CT recruitment process (see Appendix VIII for interview guide). These questions were important to understand how patients enacted their relational autonomy while situated within social relationship, and how these relationships influenced patients’ CT decision-making process. As data analysis progressed, questions became more specific to address beginning theoretical conceptualizations and validate suggested relationships and processes. For example, the practice of pre-screening was initially discussed in early CT personnel interviews but it was not yet clear what the practice specifically involved. Therefore, in additional interviews, I gently probed CT personnel further about pre-screening (e.g., “Could you describe what is involved in pre-screening patients for trial eligibility?”). Additional explications served to validate the relationship between CT personnel’s personal assumptions, structural influences and pre-screening practices that had begun to emerge from the data. These findings were then also probed in patients’ and support persons’ interviews to gain their perspectives on how they experienced pre-screening. 85 All interviews were tape-recorded and transcribed verbatim in order to facilitate close attention and understanding of participants’ responses. Interviews with patients and support persons lasted approximately 60 minutes, while CT personnel interviews lasted approximately 30 minutes. Data Analysis Transcription and analysis of the data began immediately following each participant interview and was facilitated by transcription services (Transcript Divas) and qualitative data management software program (NVivo 9). To ensure the accuracy of the transcription, the GRAs and I listened carefully to the digitally recorded interviews at places in the transcript where uncertainty was noted by the transcriptionist in what was said. We verified words and clarified meaning, as required. Data analysis was conducted in consultation with my doctoral supervisor. We had regular meetings to discuss transcript analyses and the emerging categories and themes, as well as sampling needs arising from the analysis, revisions to the interview guides, and directions for future data analysis. Interpretive descriptive study – data analysis. Multiple approaches to data analysis are allowed according to interpretive descriptive methodology. Interviews with CT personnel were subject to thematic analysis to uncover patterns of insight about the role of CT personnel in cancer patients’ decision-making processes and to inform understanding of the socio-political context of CT decisions and patients’ relational autonomy. This method of analysis is appropriate for interpretive description methodology that is characterized by a pragmatic desire to generate knowledge about common patterns and themes within human experience (Thorne 2008; Thorne 2000). To guide analysis, I developed an analytic template for thematically organizing and 86 analyzing data. Analytic templates are comprised of a list of codes representing themes and relationships identified in the data. According to King (2004), codes can be produced a priori based on theory or emerge from the data but all may be modified, deleted or expanded upon as the researcher interprets the data. Prior to developing the template, I read through all the transcripts to get an overall feel for the data. Next, to develop the analytic template, I began the analysis by reading three transcripts line-by-line and highlighting passages that reflected important ideas, themes and examples discussed by the participants. My doctoral supervisor also read the same three transcripts and identified themes. We separately sketched an initial coding framework and then together we asked ourselves what codes and themes could be condensed in order to arrive at a complete analytic template. This process often required us returning to the data to re-contextualize themes in order to understand the relationship between and among codes. To refine the analytic template, we applied relational autonomy theory to organize major themes according to the categories of personal, social and structural influences. These categories organized CT personnel’s responses to each of our interview questions. The relationships between codes were further conceptualized within the template by posing questions to the data such as “Does this theme fit under this category?” and by paying attention to the context in which the statement was made in order to determine where it best fit within the template. The analytic template was then applied to the rest of the transcripts. Two GRAs and I concurrently coded the remaining CT personnel transcripts identifying sections of text that corresponded to the template. If an issue emerged from the transcripts as important but the template did not yet cover it, a new code would be added after discussion and agreement was 87 reached between the GRAs and myself. Initial linkages within and between the data were examined and re-examined for new linkages and organizations. Questions posed to the data to determine the relationships between themes included: “What does this example mean?” and “How is this theme different or related to the other themes?” After reading and separately coding each transcript twice, the template was considered ‘final’ when all the transcripts had been coded and the template was regarded as sufficiently clear and comprehensive. Additional linkages were explored between and among the identified themes and rich descriptions and interpretations of CT personnel’s perspectives were formed (Thorne, Kirkham, O’Flynn-Magee 2004). Grounded theory study – data analysis. Transcripts of cancer patients’ and support persons’ interviews were analyzed using the constant comparative method that is characteristic of grounded theory methodology. The constant comparative method is able to capture social processes and relationships within the data by comparing all “pieces” of data with other pieces until a substantive theory is developed (Strauss & Corbin 1998). In this approach, data collection and data analysis are inextricably linked, with the interview data directing the coding and vice versa. Rather than following a linear approach, an analytical framework is developed in which several coding processes occur simultaneously to categorize the data, link the categories, and develop the core category, which is representative of the central phenomenon under study. The ultimate aim is to develop a theory that is faithful to and illuminates the phenomenon of interest (Strauss & Corbin 1998). I began the analysis by reading the transcripts line-by-line and highlighting passages that reflected important ideas, themes and examples discussed by the participants. This process was supplemented by other analytical strategies, such as asking questions of the data 88 (e.g., who? what? where? when?), considering opposite or extreme positions to bring out significant properties, and making systematic comparisons between incidents in the data to increase sensitivity to nuances (Strauss & Corbin 1998). Key factors and processes that appeared to influence cancer patients’ CT decision-making processes from CT personnel interviews were also identified and incorporated into the analysis to identify new concepts or relationships. For example, it became apparent early on that the patient-oncologist relationship was important to patients’ CT decision-making processes. Asking further questions of patients’ data, such as “Why was the oncologist-patient relationship important?”, “How was this relationship important to cancer patients?”, “What factors contributed to the importance, or not, of this relationship?”, revealed differences between accounts and illuminated key processes that facilitated this relationship, such as establishing trust and credibility. I continually questioned the data and strove to develop fitting categories. The aim of this initial stage of analysis, referred to as open coding, is to develop as many codes as possible to ensure “full theoretical coverage” of all important concepts and relationships. As analysis progressed, the codes were condensed into categories that were used to code the interviews in NVivo. This software program supports qualitative data analysis by facilitating the review and coding of the data, the storage of relevant attributes of participants (e.g., diagnosis) and the revision, refinement, and addition of codes (Morrison & Moir 1998). NVivo was also used to search the data for important codes, text, attributes and combinations of such to support the questioning of the data and further theoretical refinement. For example, the data was searched in NVivo using the keyword “trust” and narrowed to breast and prostate cancer in order to explore differences in the experience of men and women in 89 building trust in the oncologist. The second phase of the analysis, axial coding, was used to create, test, revise, and modify the associations identified within the data. Axial coding requires the identification of core categories early in the process of analysis that relates to a paradigm model, or a set of pre-defined subcategories that is used to categorize emerging concepts (Strauss & Corbin 1998). Asking questions about what conditions gave rise to each category allowed for organizing the categories of data in relation to each other and identifying properties within the larger framework. For example, questions were posed about the common actions in the codes “wanting access to a specific drug” and “considering health benefits.” This allowed the main category of “Barriers/Facilitators of CT Participation and Withdrawal” to emerge and explain what patients consider when making a decision about whether or not to participate in a CT. Finally, support person, patient, and CT personnel interviews were linked and important themes identified and contrasted within and across patient-support person dyads and CT personnel perspectives. This more fully elucidated the socio-political context of patients’ decision-making experiences and provided further insight into their relational autonomy within CT decision-making processes. For example, since CT personnel had identified the influence of protocol design (e.g., randomization) and the need to begin the trial within a specified amount of time as a barrier to patients’ CT decision-making process, questions were posed to the patients’ data about these structural influences to explore this aspect in greater detail. Linking interviews allowed for a more comprehensive picture of what factors could be contributing to patients feeling overwhelmed in the CT decision-making process and potentially undermining their relational autonomy. 90 The final phase involved selective coding. This required the identification of an overarching core category that subsumes all categories and explains the variation present in the data (Strauss & Corbin 1990). Techniques that I used to support the identification of a core category included writing a storyline, using diagramming, critiquing the theory for internal consistency and logic, reviewing the data to complete any poorly developed categories, and validating the theory through discussion with my supervisory committee (Strauss & Corbin 1998). Quite early on I identified the phrase “No man is an island” from a participant interview as an important concept that may be the central phenomenon of patients’ CT decision-making process and relational autonomy. I elaborated on this concept to include “wo/man” so that it was not gender-specific and sought out instances within the emerging theory that might contradict or further support my initial concept identification. On the whole, I was pleasantly surprised to find that the concept was able to explain a variety of personal, social, and structural influences and sub-processes within the model and also fit well with the guiding theoretical framework. It was sufficiently broad to capture a variety of influences on patients’ CT decision-making process yet also retained a ‘self-referent’ (i.e. “wo/man”) for a sense of patients’ personal enactment of relational autonomy. With the assistance of diagramming, I posed questions to the emerging theory such as “Does this process sufficiently capture the experience of CT decliners, acceptors, and withdrawers?” I continued to move back and forth between data collection and analysis until the categories were fully integrated into a substantive decision-making theory. Throughout both interpretive description and grounded theory data analysis, I paid particular attention to the impact of gender and power on patients’ decision-making processes and relational autonomy. To assist this, a critical feminist lens was employed (see Chapter 91 One) and questions drawn from both critical and feminist theories were posed to the data (Kushner & Morrow 2003). For example, normative questions that challenge social conventions and traditional ethno-cultural, class, and gender roles were posed to the data, including questions about the relationship of participants to others and to their community and larger society. This allowed for the exploration of any undue influence that may exist within CT recruitment, and more specifically, within patients’ CT decision-making process. In addition, a gender analysis framework was applied to examine specifically how gender may influence cancer patients’ decisions to take part in CTs and how relational autonomy is enacted within this process. Questions posed to the data included: • How do socialized gender roles differentially influence men and women's CT decision-making processes?!• How is relational autonomy differentially enacted by men and women in the CT decision-making process?!• How do men and women describe power hierarchies within healthcare and larger society and the potential impact on CT decision-making processes?!• How do class, age, and culture influence the effects of gender on the CT decision-making process?!• How do social meanings and expectations related to class and culture influence patients’ abilities to exercise their relational autonomy within the CT decision-making process? ! Validity and Reliability The manner in which the validity of qualitative research is evaluated differs substantially from the standards applied to quantitative research. In quantitative approaches, 92 validity refers to the degree a test or instrument measures what it is intended to measure. However, the truth-value of qualitative research resides in the discovery of the lived phenomenon rather than in the verification of a priori conceptions of those experiences (Leininger 1985). The validity and reliability of qualitative approaches can be framed in relation to the procedures of the research process through which the theory is generated, elaborated, and tested (Creswell 1994). The reliability of my research procedures in accessing and accurately representing my study participants’ lived experiences may be demonstrated by my use of theoretical sampling, the systematic approach to data analysis facilitated by the use of a software program, my commitment to filling in poorly developed categories until theoretical saturation is reached, and the range of expertise and theoretical sensitivity of my doctoral committee to support the adequacy of the research process. All procedures used and analytical decisions reached during the research, including the origins and development of major concepts and the identified associations among theoretical constructs, were also made explicit in this dissertation to allow others to understand and critique the analytical logic utilized. In grounded theory studies, a theory is deemed “valid” if it: (a) “fits” the situation or is faithful to the complexities of the everyday realities that are under investigation and (b) “works” when applied to real world settings (Glaser & Strauss 1967). To ensure the theory “fits” with the data, I employed processes of triangulation and peer examination. Triangulation refers to using two or more approaches to answer the research question, thus allowing for a comprehensive picture of the studied experience (Knafl & Breitmayer 1989). For this dissertation I triangulated my sources of data collection from different participant groups (CT personnel, patients and their support persons) to obtain convergence around the 93 information gathered. I also triangulated different research methods. I used grounded theory methods, enhanced by interview data from the interpretive descriptive study with CT personnel, to develop the theory “No Wo/Man is an Island”. I also used a critical feminist lens to elucidate the relevant power relationships and any undue influences or gender differences in the findings. Finally, triangulation of investigators occurred within the data collection and constant comparative analysis process. I relied upon two GRAs and my doctoral supervisor to validate the coding process and resultant template. This allowed for discrepancies in coding to become apparent in order to further refine and reach consensus around the interpretation of the data. Another way to assess whether the study “fits” is to conduct member checks with study participants. Member checking involves testing the theory with informants so they may comments on data, analytic categories, interpretation and conclusions (Lincoln & Guba 1985). Initially, I had thought to validate and extend the study findings using focus group interviews with cancer patients who had participated in interviews and were found to be excellent informants. Preliminary findings would have been presented, along with key aspects of the evolving theory, and participants would have been invited to respond. However, I was unable to conduct focus groups for this dissertation because I had relocated to a different city, which made returning to British Columbia impractical. However, I do not believe this necessarily detracted from the “validity” of the study. Some authors point to limitations of focus groups as a method for ensuring the “fit” of a grounded theory (Charmaz 2006; Sandelowski 1993). For example, Sandelowski argues that participants’ views and beliefs may change over time, thus making their perspectives inherently unreliable (Sandelowski 1993). Also, since the result of grounded theory is one main theory or core 94 category that describes the various “stories” throughout participant experiences, focus group participants may not recognize their own perspectives within an abstraction of combined perspectives. Therefore, instead of member checking using focus groups, I followed Charmaz’s suggestion that validation may instead occur in relation to developing codes and concepts (Charmaz 2006). For example, emerging relationships between codes and categories were tested in interviews with participants during concurrent data collection and analysis, thereby incorporating member checking into the research process. I relied on peer review to substantiate the theory and determine whether it “worked.” The findings were shared with my doctoral committee at various stages of analysis. This allowed me to discuss not only the emerging concepts and themes but also the research process and whether my findings were coherent. My doctoral supervisor continually pushed me to consider outlying or negative cases, so that concepts could become enriched and densely described. Any outlying cases that did not fit with the theory were carefully examined to ensure the associated variability was incorporated into the theory. This was important for the representativeness and dense description of concepts in order to offer deep insights about patients’ CT decision-making process (Strauss & Corbin 1990). The degree to which the theory is perceived by others to provide a useful guide to action (e.g., with respect to decision support interventions) was also assessed by peer examination. For example, the Medical Director of CTs and a Senior Scientist who helped facilitate recruitment at the cancer centre reviewed the diagram outlining the “No Wo/Man is an Island” theory. These professionals interact with cancer patients and families on a daily basis and are well positioned to determine how well the theory could apply to practice. Both professionals provided positive feedback on the model and indicated that the process of 95 decision making matched their reality and confirmed the theory could be useful from their perspective. Ethical Considerations Given the focus on cancer patients’ relational autonomy within the context of CT recruitment as a research objective of this dissertation, it might also be interesting to reflect on the ways the dissertation research process may itself have aided participants’ autonomy. There are two examples that I will expand upon here. First, purposive recruitment of marginalized perspectives, such as CT decliners and withdrawers and patients from lower socioeconomic groups, was potentially supportive of their autonomy because it sought to include them in the knowledge generation process. It demonstrated the value of their perspectives and offered these patients the opportunity to engage in meaningful discussion about their experiences. Moreover, the interview process allowed these patients to exercise their autonomous capacities of self-reflection and self-understanding in order to provide responses to the interview questions. A second way the conduct of this study supported patients’ relational autonomy was through the inclusion of support persons. This study affirmed these relationships as important for patients’ autonomy as is recognized by relational autonomy theory, and that family members, spouses etc. could be important influences on their decision-making process. I also abided by other, more traditional ethical recommendations provided by the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (CIHR 2010). For example, approval for the study was obtained from the requisite Research Ethics Board. Interview participants were provided information about the study, assured their participation was purely voluntary, and that their privacy and confidentiality would be respected. Privacy 96 and confidentiality was particularly important given that patients and support persons may be interacting with CT personnel who were also being interviewed. This was particularly important for CT personnel given the involvement of the cancer centre’s Medical Director of CTs in this study to facilitate recruitment. Confidentiality was assured by assigning all participants an identification number. A master sheet matching names to ID numbers was kept separate from the data in a locked filing cabinet at my supervisor’s research program office. All data was kept confidential and viewed only by the GRAs, my doctoral research supervisor, and myself. All data files associated with the study were password-protected. Explicit care was taken to mask potentially identifying information in this dissertation and in the use of participant quotes. For example, the specific geographical location of patient participants was not named. Subject matter of the interview may have been emotionally distressing for some patient and support person participants. As such, information about supportive care services at the cancer centre was provided during the informed consent process. Care was also taken to pause during the interview if the participant appeared distressed. For example, if the participant became teary I then offered to stop the interview completely or, if the participant desired to continue, I gained reassurance that he/she felt emotionally ready to continue. Patients, support persons, and CT personnel who participated in the study were invited to leave their contact details on the consent form to receive a copy of a lay summary of the study findings. Disseminating the knowledge obtained from this study so that it may be useful and advance the field is a key ethical concern of mine. Reports of this research project will be prepared for dissemination at conferences and health forums focusing on research ethics and 97 cancer care. Publications will also be prepared for peer-reviewed scientific journals to disseminate the findings to an international audience as well as a variety of health professions. An executive summary of the findings has already been shared with the NCIC CTG at their annual spring meeting in 2012 (attended by researchers, CT personnel, and pharmaceutical industry representatives from across Canada). I, along with Dr. Balneaves and Dr. Richardson from the current study’s CIHR grant, were successful in obtaining a CIHR Meetings, Planning and Dissemination grant to conduct an interactive workshop with key CT stakeholders (i.e., patient advocates, researchers, administrators, and pharmaceutical industry representatives) and research ethicists at the NCIC CTG annual spring meeting (co-sponsored by the NCIC CTG Quality of Life Committee). The primary goal of the workshop was to provide efficient knowledge transfer of study findings and to discuss future research projects that may further test the developed theory in other clinical sites and cancer populations across Canada. In addition, the workshop provided a venue to begin brainstorming potential decision support strategies that will aid patients in exercising their relational autonomy within current CT procedures in cancer care. A summary of the workshop outcomes is in development and will be distributed throughout the NCIC CTG network and the larger international cancer CT community (e.g., National Surgical Adjuvant Breast & Bowel Project). 98 Chapter Four: Cancer Patient Accrual to Clinical Trials: A Qualitative Analysis of Clinical Trial Personnel Perspectives Low accrual to cancer clinical trials (CTs) continues to challenge the translation of innovative research into clinical practice to improve cancer patient outcomes. Much quantitative literature has described barriers to cancer patient accrual to CTs. However, CT personnel perspectives, including oncologists and CT nurses, have informed little of this literature. When a trial opens at a centre, treating oncologists who are also part of the research team are often responsible for identifying eligible patients and making initial contact with them about the trial. If patients are interested in the trial, CT nurses will often then conduct the initial consent process. However, both CT nurses and investigators/oncologists may review the trial protocol with patients, discuss medical information with patients and discuss the intervention/drug and potential side effects. They will also review the time commitment and other logistics required for the trial. Both CT nurses and investigators/oncologists are also available in person or via telephone to answer any additional questions and concerns patients might have while they are making a CT decision or once they are enrolled in the trial. CT personnel are uniquely situated to possess knowledge of the larger CT and cancer care system (i.e., structural context) while also having insight into cancer patients’ personal and social situations that may impact CT decisions. Moreover, CT personnel often have close relationships with patients and are highly influential in their CT decisions (Avis et al. 2006; Joffe et al. 2001). Therefore, understanding CT personnel’s beliefs about what influences patients’ CT decisions is important since their beliefs may impact how they discuss and share trial information with patients. Finally, an exploration of CT personnel’s experiences allowed 99 for more patients’ perspectives to be represented in the overall research, since CT personnel were involved in recruiting patients with multiple tumour types, and from different socioeconomic status and ethnicities, to CTs. The purpose of this interpretive descriptive study was to gain comprehensive understanding of the personal, social, and structural influences on cancer patients’ CT decision-making. Interviews with CT personnel included open-ended questions that explored the challenges and facilitators experienced by cancer patients in making decisions about CTs and how health professionals could better support cancer patients and their support persons in the CT decision-making process. Twelve cancer CT personnel were interviewed for this study, including six nurses and six oncologists involved in multiple aspects of cancer CTs (e.g. patient recruitment, enrolment, and management of CTs). Half of the participants were female (n=6) and the rest were male (n=6). CT personnel were not tumour site specific; they were involved in trials related to breast, prostate, gynaecological, gastrointestinal, head, and neck cancers. Findings This interpretive descriptive study used thematic analysis guided by a relational autonomy lens to structure the CT personnel findings according to the personal, social, and structural influences on why cancer patients accepted, declined, and withdrew from CTs. Quotes from the interviews with CT personnel are used to illustrate major findings where appropriate. Personal Influences on Patients’ Decision-making CT personnel believed cancer patients’ personal beliefs about CT research played a major role in patients’ decision making about CTs. As one participant shared, “the vast 100 majority of patients have internally-driven beliefs and values that would contribute to the decision to go into a clinical trial.” Personal beliefs included beliefs about the potential benefits of the intervention being tested in the trial or general beliefs about what it means to participate in CTs. In addition, both CT nurses and oncologists thought patients’ hope, fear, anxiety, and distress informed their personal beliefs and decision making related to cancer CTs. The following sections elaborate on these themes in greater detail. Patients’ beliefs that they will receive better care. CT personnel thought the primary reason why cancer patients accepted CT participation was because of their belief that trials will benefit them directly by offering them a new curative treatment option, a treatment that could prevent the cancer from recurring, or a treatment that could offer them fewer side effects or better quality of life. CT personnel also noted that some patients might have a desire to try something different, which is often connected to their belief that “newer is better” and that CTs are at the frontier of medicine and offer the best and most advanced treatments. As shared by a medical oncologist: So I think the number one reason is because they want to have something that will benefit them. Obviously they want to try something different or they want to do something different, they want to do something new. So many trials are geared towards patients that don't really have other options. Or the current options, we know what they can do and so some people want to try to do something better than or try to get the best care as they can. So they want to try to benefit themselves. CT personnel believed that patients’ perceptions of personal benefit superseded altruism as a motivating factor for CT participation, although benefiting future populations was also perceived to be an important consideration. As a medical oncologist stated, “I will admit there’s probably the rare patient who is probably truly altruistic but ultimately most patients say, ‘Well, what does this mean for me?’” Later in the interview, the same participant 101 remarked “So, in the end, it always comes down to, as much as we say, ‘You’re helping the population’, people do it for themselves.” Patients’ beliefs about being a “guinea pig.” Patients’ concerns about the quality of care and the unknown risks posed by a CT and being “experimented” upon, with negative consequences, were raised by CT personnel. CT personnel believed some patients associated CTs with being a “guinea pig”, which was perceived to negatively influence their perspective of CTs and the CT decision-making process. One oncologist shared how he thought patients’ concerns about being the subject of an experiment might stem from misunderstandings about the nature of CTs and research in general: “I think it could be just a lack of understanding of what a clinical trial is and they hear the word “research” and they feel like they’re a guinea pig.” Other CT personnel perceived some patients to be worried about not receiving the best care if they are in a trial. This belief may be tied to the age of the patient, as one CT nurse recounted the views of her grandmother towards CTs: I think of someone like my grandmother, for example, who’s just, ‘Well, I don’t want to be someone’s guinea pig.’ She looks at it that way, that she’s not going to get good treatment if she’s on a study but she will if she – just as a regular patient. The same nurse described how it is important for CT personnel to communicate information to patients so that they understand they will still receive good care while on a trial and can provide informed consent: I think some [patients] think that it’s an experimental treatment, or they don’t understand that they’re still going to get a proven treatment on it. And a lot of people think, ‘Oh well, research is either you get the placebo or the real drug’ and they’re like ‘Well I don’t’ – They don’t understand or maybe they haven’t been given enough information that here’s your options, here’s the treatment you will get, or on a trial you’ll get this treatment plus this and this and this. So maybe they’re not shown the benefits of it. 102 Historical considerations were also perceived to play a role in patients’ understanding of and their beliefs about CTs. One nurse imagined that patients who had immigrated to Canada might be particularly concerned about being used for experimentation if their home country had a history of mismanaging research: And it could be too, you know, I’m just supposing too, if they’ve come from another country where perhaps research or experimentation on humans, it wasn’t terribly well managed at some point, they may have these particular components. Hope for a benefit/fear of risks. Emotions were identified by CT personnel as having a powerful impact on patients’ decisions to participate in CTs. CT personnel believed that many cancer patients persistently held hope that a trial would benefit them despite being informed that the treatment may not be individually effective. As one CT nurse remarked: There’s a hope that somehow this may be a wonder drug that may do it for them and so they’re willing to put up with all the other things, you know, even though it’s pretty clear – it depends on the drug. In contrast, anxiety or fear of cancer growth or recurrence was also perceived to motivate patients to participate in trials. As one radiation oncologist explained: They [patients] have a hope to get the best treatment, right? Or a hope to get the least side effects or a hope to be cured, you know? Or a fear or the flip side to all of those, the fear of not being cured, or a fear of not getting the best treatment. Hope for better outcomes or fear of illness progression was perceived to exist on the same continuum and to differ according to disease stage. CT personnel shared their belief that heightened emotions can sometimes interfere with patients’ understanding of the purpose of CTs and their CT decision making. This was believed to be particularly the case for patients with cancer recurrence or those who had advanced disease, where hope was seen to gradually transform into desperation or fear. This was perceived to contrast starkly with other 103 cancer patients, who felt hopeful about their prognosis but willing to consider alternatives to usual care. One CT nurse described the quandary of advanced cancer patients who had already exhausted several lines of chemotherapy and were now faced with limited to no standard treatment options. With nothing else to try, these patients were seen as being motivated to consider CTs because they are “at their wits end” and were facing huge risks to their mortality. As another CT nurse remarked, “patients are quite keen to continue because they don’t want to let go, they want to continue on treatment.” Other CT personnel believed some patients were more motivated by fear of recurrence because their initial diagnosis and treatment was so traumatic: They’ve been frightened to death because of this diagnosis of cancer. Often they’ve had their surgery and really at surgery, you’re cured. And anything else beyond that is just to ensure that if there’s anything small that’s out there that we can’t see, we can hopefully get rid of, so that’s what radiation and chemotherapy is for. So those people have been petrified and they want to do everything that they possibly can to make sure that they never have to face it again. Anxiety and lack of understanding. CT personnel reported that cancer patients were often concerned about potential toxicity or side effects from the drugs being tested in CTs and some decided not to participate in a trial after learning about the risks. As shared by a medical oncologist: One of the unknowns is the toxicity. It has not been well worked out many times and so they're concerned about the additional toxicity or other toxicity, side effects that might happen from a clinical trial drug. CT personnel could appreciate patients’ fears, particularly in relation to Phase I trials, in which there is limited knowledge about the intervention at this stage of testing. However, CT personnel expressed concern when cancer patients’ anxiety was seen to interfere with their understanding of CTs. Without guidance from a health professional or adequate 104 information and emotional support, one oncologist believed the decision-making process could be totally overwhelming for some patients, thus demonstrating the importance of social relationships for supporting cancer patients’ capacities for autonomy: Yeah, so some people can’t get their heads around flipping a coin and being randomized or the inhumanity of a coin or a computer deciding their fate and treatment, rather than the doctor who’s in front of them or at least somebody human. And they just can’t get their head around the business of randomization at all. Do what’s best for me, you know, I can’t decide, there’s too much pressure, there’s too much stress, there’s too much at risk to go that way. CT personnel believed that patients who were deciding between multiple treatment options, without a clear sense of which to take, also experienced anxiety and perceived the presentation of a CT as distressing. Newly diagnosed cancer patients were observed to be particularly vulnerable since they were also dealing with the trauma of receiving a cancer diagnosis while being inundated with treatment information. As one medical oncologist stated: A lot’s coming at them. Sometimes it's their first treatment for cancer and so a lot is coming at them. So I think the understanding of the whole picture, the bigger picture or the little picture, the details, is difficult. So that's what makes it most difficult. Inability to cope with logistics. The majority of CT personnel referred to the significant time commitment required of patients to participate in CTs. They believed this commitment discouraged many patients with early-stage cancer, with most viewing CT participation as a “hassle”, especially when there was no sense of urgency to enter the trial. As one oncologist remarked, “they look at all the things that you have to do, all the extra tests and they just don’t want to do it.” Cancer patients were perceived to withdraw from trials because the logistics associated with their participation became too burdensome. As one CT nurse commented, patients realize after they have agreed to participate “that [it’s] just too much of an 105 inconvenience for them.” As another CT nurse explained, “they just don’t like the amount of travelling, visiting and everything in order to participate on the trial. So that would be the main one.” CT personnel appreciated the relational aspects of trial participation, with patients needing to coordinate with family to help them attend CT appointments. CT nurses pointed to the difference between what patients expect when they sign up and the reality of participating in the trial, especially when travel to the study site and coordination with their social network was required: So especially patients here, especially oncology patients, they’re just not up for maybe that extra ECG they have to do, or extra scan that we may require for this particular trial. It is time consuming, so they may not be – they may have been eager to join up, but when the reality of – especially if they don’t live in the city and they’ve got to come in and maybe they’re staying at someone’s house and the parking and they’re – you know, a lot of them too don’t like coming into [the big city], it’s busy and scary for some people to come into a city. These findings call into question the effectiveness of informed consent procedures meant to inform patients about what is needed to partake in studies; however, the last quote also hints at the social and structural context of trial participation (e.g., staying as someone’s house, parking, etc.). Broadening the informed consent perspective using a relational autonomy lens appreciates the social and structural context of trial participation on patients’ CT decisions and their informed consent. Finally, other CT personnel believed patients accepted trial participation but then later withdrew because their condition worsened and they found it too difficult to cope with the demands of the trial because of fatigue or other cancer-related symptoms. As one CT nurse shared: 106 Sometimes if things go wrong, like the disease starts progressing and they’re having to travel from [distant city] all the way to here for scans and blood tests and doctor’s visit, it tires them out a lot. So I’ve had a couple of patients withdraw because of that. Social Influences on Patients’ Decision-making In addition to personal factors, CT personnel also identified social factors that they perceived to influence cancer patients’ CT decision-making. Altruism, trial presentation by the physician, and family opinions were the key social influences identified as impacting patients’ decisions to participate in CTs. Altruism. Although CT personnel perceived the majority of cancer patients to be motivated to participate in CTs because of their own self-interests, some individuals were perceived to be influenced by the desire to “give back” and help further science to benefit future cancer populations. Concern for society and important relationships, especially concern for the future well-being of their family, was perceived by one radiation oncologist specializing in prostate cancer to be an influential factor in some cancer patients’ decisions to participate in CTs: Patients want to contribute, especially in my practice, they’re often older men who’ve got young kids and they genuinely want to do what they can so that if the younger generation gets prostate cancer, they can do better than they’re doing. A few CT personnel shared experiences of consulting advanced cancer patients who were excited about participating in a CT in order to help science find a cure for the disease. These patients were perceived to make a quick decision without much thought as to how participating in the trial would impact them personally or socially. As part of the informed consent process, CT personnel described how they sometimes discouraged cancer patients from considering what the CT can do for others. Instead CT personnel re-oriented patients to reflect on how participating in the CT could impact their own situation. CT personnel thus 107 appeared to be privileging patients’ individual autonomy over their relational autonomy in these cases. However, CT personnel may actually be acting pragmatically and out of concern for the possible consequences of patients not considering the individual impact of CT participation, which could be their withdrawal from the trial. One CT nurse explained: A lot of people say yes to clinical trials because they feel like they are doing a service to humankind. I often say to people, look, you cannot view this as you are doing a service to someone. You really have to think about what this commitment’s going to be for you and it’s nice that you want to help others, but you have to be able to do all that’s required and if you can’t, because it’s getting in your way, or you’re too tired to make all these appointments then you’ve got to keep that in mind. You’ve got to think about that. Presentation of trials. CT personnel described the presentation of the trial by the physician, or what many CT personnel referred to as the trial “sales pitch”, as influencing cancer patients’ decisions about CT participation. When asked to describe the pitch, many CT personnel related the process to the selling of a car. However, CT personnel emphasized that unlike selling cars to make personal profit, the oncologist was perceived by his or her colleagues to be presenting the trial with the patient’s best interest in mind: Like I say, the oncologist may think that in some cases, it just gives the patient more options for future treatment and sometimes – and how it’s presented, I think may… First, do they think it’s in their best interest? Well, obviously they wouldn’t present it if it wasn’t. Sometimes, however, CT personnel believed physicians’ personal convictions about a trial could become apparent in the trial discussion and inadvertently sway patients’ attitudes and decision making. If physicians, for example, were questioning the science of the trial or whether the trial was a good option for the patient, their unease may become apparent in the consult. CT personnel thus demonstrated an awareness of the relational nature of CT recruitment and the power held by physicians to influence patients’ CT decisions. As one radiation oncologist shared: 108 I mean, you might be doing a study that involves pushing the radiation program harder to treat prostate cancer but as a clinician, you’re also rather worried about the toxicity. So when you present this study to the patient, your fears and anxieties about the study are evident and if that comes over, then the patient is more likely to decline the study. While many CT personnel believed oncologists try not to pressure or influence patients’ decision making, some CT personnel recognized that subtle communication factors such as “the way the trial is presented to the patient, the body language and the terminology used” might encourage patients to privilege one option over others. Many CT personnel suggested that physicians may not recognize their own communication nuances and how this might sway patients. As a radiation oncologist explained, patients may not be aware of the ways in which physicians’ subtle conveyance of their personal beliefs can influence their decisions: It’s [communication] really subtle and they [doctors] may not even recognize it’s going on themselves. In fact, I’m sure most of them don’t. It really isn’t often somebody comes and says, well, this doctor said this to me. Another radiation oncologist provided a telling example of subtle communication, demonstrating how the way things are phrased could influence patient decision making about CTs: For example, we’ve got a great trial here of an extra high dose boost of brachytherapy [internal radiotherapy] for prostate cancer, and all of us who are doing the trial in our hearts believe that it will likely give better cancer cure rates, at a cost of more toxicity. The trial basically comes down to the trade off, if we’re right between curing a few more people at a cost of more toxicity. So there you see, you could present that trial as, ‘We’ve got this great new treatment which we think, but we have to prove it with a study, is going to cure more people.’ Sounds great! Or you could present, ‘You’ve got this trial which is looking at this new treatment but we’re pretty sure it’s going to have more side effects, and most men with prostate cancer don’t die of prostate cancer, they die of old age, and so if you participate in this study, you may just get all the side effects.’ You could say something like that, so I think those are the subtle ways in which it’s presented. 109 A few CT personnel did share the concern that if a physician was an investigator on a study, he or she might experience a conflict of interest. This conflict was associated not only with the pressure to accrue patients to the trial, but also with the scientific rewards of achieving the targeted sample size and recognition from funders and peers. As one radiation oncologist explained: Investigators have a lot invested sometimes in trials. Sometimes it is about the per case money to keep their clinical trials’ program going. I’m not pointing a finger at Dr. X, okay. That’s not what I mean to do here but all of us, right and also another investment will be…I mean, there is some recognition if you put patients on trial. Flip side is, you know, people rain on your parade if you don’t put patients on trial. So there’s pressure there. Trust and preservation of the physician-relationship. CT personnel perceived patients’ trust in their doctor enhanced their receptivity to subtle clues or encouraged patients to follow their physicians’ recommendations related to CTs. One radiation oncologist elaborated on how patients’ trust in physicians was implicated in their decision-making: I think it’s the trust in their doctor […] and if it’s a randomized trial between two different treatments, I think it may come down to how the standard treatment is presented and how the experimental treatment is presented to the patient in terms of potential benefits and side effects […] I think patients appreciate the honesty that comes across, that builds the trust and then they’re more likely to do what you recommend. CT personnel also reflected on the possibility that some patients were accustomed to deferring to physicians in the treatment setting and might interpret the invitation to participate in a CT as itself a recommendation, even if the oncologist has taken precautions to present the trial impartially. This may reflect the power dynamic between physicians and patients, where patients accepted the advice of their physicians because they were perceived to be the expert or authority. As one oncologist shared: 110 They [patients] see [it as] an authoritarian thing and they don’t think it and then they’ll find it easy to go on, cause they say if the doctor mentions it, it must be good. So it becomes almost an authoritarian thing or they’ll just say yes to it. Other CT personnel discussed examples of patients who appeared to make decisions out of a desire to please their physician or to preserve their relationship with their physician or the cancer agency. One oncologist commented on how patients may make CT decisions so as not to compromise care they receive from their physician: Some patients actually stay on trial because they’re worried we’ll get upset at them and then not give them any more treatment and I really try to emphasize we are not like that at all, which is – it’s kind of sad that they think of us that way. And I think it’s more of a fear. I don’t think they see anything horrible of physicians or the clinical trial’s process, I think it’s just a fear that, if they say no to something, then I’ll be upset, not give them chemo any more. I’ve actually had a few patients tell me that. I’ll say, ‘Where did you get that idea from?’ And they’re, like, ‘I don’t know.’ Because I never said that. And, yeah, patients just have that fear sometimes. This misunderstanding by patients about what was communicated by physicians was believed to arise from the fear or distress associated with the cancer diagnosis rather than the hierarchical patient-oncologist relationship. As one oncologist explained, it matters less how well physicians present all the options because patients will hear what they want to hear in their time of need: I think the patients themselves are going through so much with the diagnosis so you mention one thing and all they hear is ‘Wah, wah, wah, wah, trial, wah, wah, wah. I think it’s a good idea.’ And so then they end up hearing, ‘My doctor thinks this is a good idea if I do this.’ Right? And they don’t hear everything else. Socioeconomic factors. The complexity of CTs and the consent process was perceived to pose a barrier to informed decision making for certain patients. CT personnel perceived patients’ educational background or “cultural differences” affected their ability to understand CT information. Additionally, language barriers and literacy were thought to limit 111 patients’ understanding of CT information. As shared by a medical oncologist and another CT nurse: It's very medical and scientific and a lot of people just don't understand. And in my patient population, which may be different from others, a lot of these guys are older, 'cause it's prostate cancer, and sometimes when I'm talking to them, I don't know if they really understand what the implications are and so on. Maybe they don't understand the details. I think they understand the big picture but I think sometimes it's difficult about the details. Illiterate […] you do get people that are illiterate that will make a snap decision and think that they don’t need to read or they won’t ask for someone else to read it to them because maybe they’re embarrassed. Other CT personnel believed “general social issues” made it difficult for patients to participate in CTs. For example, lack of income confounded with geographical distance were perceived to be significant barriers to patient participation in CTs, since patients may not have the money to pay for travelling to and from the trial site. As one participant stated, “high risk population, like low income, that would be an issue because even transportation to them is a huge factor for compliance.” These beliefs about socioeconomic factors led to CT personnel forming assumptions about patients’ abilities to participate in CTs. For example, one radiation oncologist spoke about how geography was linked to socioeconomic resources and how his assumptions about a patient’s occupation and financial resources influenced his decision about whether or not to present a trial to this patient: You can tell a lot about whether it’s going to be a fit also on incorporating where they live and what they do. I mean, obviously a logger who lives in [city up north], you can just guess he’s not going to have extended health benefits to bring him down so I look at things like that…So definitely geography, that’s a huge piece. Gatekeeping by physicians. The previous example illustrates the role of CT personnel in determining who is offered a trial. This process, called gatekeeping, refers to the process whereby CT personnel determine which patients they will approach about a trial. CT 112 personnel remarked on the nature of gatekeeping by physicians and how not offering a trial to a particular patient immediately forecloses the accrual of that patient into a trial. From the interviews, it became apparent that gatekeeping involved pre-screening patients for trial eligibility and was a strategy some physicians used prior to initiating a trial discussion in order to save time and resources. Thus, CT personnel recognized the structural component to patients’ relational autonomy, where oncologists decide who receives a CT offer, and that this decision is influenced by limited CT and healthcare resources. As one radiation oncologist shared: So even if I’m not thinking about the patient in front of me, I’m thinking, okay, is it a worthwhile investment of the extra time to offer this trial to this patient and in those sorts of cases, it isn’t, right. So I can move on with my day much more quickly if I pre-filter. CT personnel talked about two kinds of screening practices based on trial protocol eligibility criteria: patient pathology and social factors. CT personnel perceived the assessment of pathology as relatively straightforward; however, assessing patients’ social situation was viewed as more challenging. As one oncologist described: So we can break it down into the patient characteristics and the pathology. So one is, do they actually meet the eligibility? So do they have the right pathology -The eligibility criteria? And that’s the easy part because that’s just a simple checklist. The next is probably the harder intangible stuff. Is this a person who really understands what a clinical trial is? Is this a person who truly is suitable for someone who – and it depends on each trial. Some trials have much more rigorous tests. Some are quite easy and simplistic in terms of the logistics. So not every trial will suit every patient. Cancer patients’ mental competency was seen by CT personnel as having an important role to play in the suitability of patients for CTs and in physicians’ determinations of whether or not to offer a trial to a particular patient. As one CT nurse remarked: Yeah, they don’t have any major mental health issues that would be troublesome while in a clinical trial or questioned, because there’s always that part of the 113 eligibility that they have to have, like, certain mental competence to be eligible for the trial. Another oncologist commented on how, in his mind, the criterion of mental competency for CT participation outright excludes certain patient populations with psychiatric disorders or addictions: Definitely the ones who, you know, drug addiction or the downtown east side population12 are not patients who I put on trials. And patients with depression or psychiatric disorders I tend not to, either, (a) because of the medications and (b) they’re just – you don’t know whether they really truly understand the trial and it can get very complicated. CT personnel recognized that the sort of person who is a good fit for a trial is an “exceptional person”, someone who is competent and able to think clearly about the trial. These characteristics mimic the ideal autonomous person, or someone with all the resources of the upper echelon of society. However, acknowledging the relational aspects of autonomy, CT personnel also believed it was desirable to have patients who have family nearby to provide social support during the patient’s CT participation. As one oncologist elaborated: Living with someone who supports – so, ideally, a married couple is always a good thing because you know a spouse is normally there for support. And that’s not to say single people shouldn’t go on trials. Well supported single people – and even if they don’t have families but have a very supportive friend network, you know, these are people who normally come with a friend. They might be taking notes. Where I tend to be a bit more worried is sort of the person who lives on their own, doesn’t seem to have a lot of friends, may have a lot of pets. Those are the ones you tend to worry a bit more. Family member influence. Relational influences were most clearly seen when CT personnel spoke about the influence of family and friends on patients’ CT decisions. In 12 The downtown east side (DTES) is considered one of the poorest areas in Canada. It has a very high incidence of drug use, poverty, crime, violence and sex trade (Brethour 2009). 114 addition to physicians, family members or close friends were also perceived to influence patients’ decisions to participate in CTs. As one participant remarked, “some people go into studies to please the doctor or to please their spouse or…that framework and a sense of doing what is right for others.” Sometimes the influence was seen as being a quiet nudge in the direction of a trial or a suggestion to look into a trial that the family member heard about in the media or online. Other times a friend or a neighbour of the family might have had a good experience with being involved in CTs. Advice from family or trusted others was seen as opening patients to the possibility of participating in a trial even before the oncologist presents CTs as an option. One CT nurse reflected upon the role of trusting relationships with family members and how this impacts patients’ CT decision making: They’ve had family members say, ‘hey, when you go, ask, because this is good stuff. I’m reading all this good stuff about this particular trial. They’re doing this trial, ask if you can be on this trial.’ So they would already be coming in then with an open mind and wanting to be a participant if they can because they’ve already been convinced by someone they trust in the family to already ask about it. On the flipside, CT personnel thought that if a family member has had a negative experience with trial participation, cancer treatment or cancer care, this could influence patients’ decision making by limiting patients’ willingness to hear more about CTs. As one CT nurse explained: Some people who are convinced that they don’t want to do trials, you can almost see it when you’re talking to them. They have already made up their mind. They’re closed to that option so they’re not really, they’re open to hearing about anything that might be there in the first place and those people might have had other experiences happen in their lives or with other family members being on stuff… If that family member had a grandpa that died horribly after being radiated or had chemo and had horrible side effects or whatever, in their head, chemo is horrible. At times, the support or encouragement related to participating in CTs provided by a family member or a significant other was seen to move well beyond a simple suggestion. 115 Instead, some CT personnel spoke about patients being “pressured” by trusted others to enter into a trial. This was perceived to be problematic because, as one radiation oncologist shared, “you’ll see people do it because their family really wants them to, so it isn’t even for their own reasons. It’ll be because other people are telling them they should.” CT personnel particularly noted that spouses of prostate cancer patients tended to be intimately involved in patients’ decision making about CTs and sometimes were perceived to outright influence patients’ decisions. As an oncologist described: I think certainly for the men in prostate cancer, their wives and family and caregivers, the primary caregiver, has a huge impact on that decision. More often than not—this is my bias—the wives tend to push the husbands into doing clinical trials and kind of actually help make that decision for them. More often than not, they actually do push I think for the patient to go on a study. When asked why family members were motivated to encourage or push patients to participate in trials, CT personnel believed it is because families value their relationship with their loved one and were looking out for the patient’s best interests. Other times, family members were perceived to be influenced by their own personal fears associated with losing a loved one to cancer. One radiation oncologist mused how some loved ones may also be overcompensating for the guilt they feel for not living close to the patient: And a lot of it can be tied into their own belief system and also their own fear that they’re going to lose their relative, tied up in the fact that they don’t live there and they feel guilty and they wish they could be there. There’s all kinds of personal issues for them and they really have nothing to do, if anything, with what’s good for the patient or him or herself. Some CT personnel believed patients might make CT decisions that place the needs of the family before their own because they wanted to avoid further upsetting their loved ones. As one oncologist shared: I think sometimes some patients are worried about being a burden to their family. The other aspect of it, some patients are afraid to decline a trial because then they feel like 116 they’re not doing everything and their family is going to get upset with them as well. So it’s not so much that we’re [CT personnel] coercive but that the family is coercive. When asked specifically what being a ‘burden’ to one’s family meant, another oncologist perceived it differently and described burden as being patients’ worry about imposing upon their family to attend CTs and related activities: Some patients say, ‘Well, I don’t have a ride to come in. I feel like I’m burdening my family with all these extra tests and coming in. I’m going to experience more side effects.’ I think sometimes some patients are worried about being a burden to their family. Other cancer patients were perceived to decline CT participation because of the time commitment and how it might interfere with their daily lives or family responsibilities. As one CT nurse mentioned, patients can be very protective of their family time: And then the other force is of course the family. Like, you know if they have little kids. Those doors will be first and you would be silly, you will not have a patient if you tell them that you can’t accommodate for their immediate family issues. And if you cannot accommodate, then you can’t do it. In contrast, CT personnel explained that some patients considered participating in a trial because of the responsibility they felt towards caring for their family or young children. An oncologist recounted an experience with a young woman with advanced cancer who wanted to go on a trial in order to have a future with her children: She came over because her oncologist had given her two cycles of chemo, she’s a young woman, she’s got three young kids and he said he had nothing else to offer her and ... But she could come over and talk to me about experimental therapy. So she came over in the hope that there would be other treatment options for her and she said – I’ve got three young kids, I’m a mother, I want to do everything I can with this. Structural Influences on Patients’ Decision-making In addition to personal and social reasons, CT personnel also identified structural factors that they perceived to influence cancer patients’ CT decision making. For example, 117 CT personnel believed many cancer patients participated in trials because they correctly perceived them as a gateway to better treatment and specialized care. Access to drugs. CT personnel described situations where patients accepted CT participation because it was the only way to gain access to new treatments or promising drugs that were not yet approved for standard care or available as part of the cancer agency’s formulary. CT personnel also described patients as entering a CT in order to receive a drug earlier than they would have otherwise been eligible for given their disease stage. These patients saw CT participation as being a “loop hole” in the system because it allowed them to bypass the wait until they were eligible to receive the drug as part of their standard care. As one CT nurse shared: Say there’s a new drug...I’m not sure if you’re familiar with the Phase I, II and III part of it. So after Phase I, sometimes we do the phase two studies; so those drugs, patients can get access to. So it can be a new targeted therapy drug that, technically, won’t be covered if someone, let’s say, is only at stage I cancer and they’re getting their first line of treatment. This drug can only...say, this new study drug or another drug is only...they can only access it on the third line of treatment. They can actually access it earlier in the study. CT personnel also believed hospital or provincial-level policy making about oncology drug coverage impacts patients’ CT decision making. Patients were perceived to consider the cost-savings of accessing a promising drug through a CT instead of paying for it out of pocket. As one CT nurse explains: Well, like, if they’re in a clinical trial, they don’t pay for any drug but, if they were off the clinical trial and if it was a drug that was on the market, sometimes they’re not through the systemic therapy program. The clinicians that are involved with the systemic therapy program, they decide every year which drugs are going to be covered by MSP and which are not, right? So there are some drugs out there that people go elsewhere and purchase and have in I.V. infusions off site but they’re paying for those drugs themselves. So, if they’re in a clinical trial, they don’t pay for any drugs, they’re all paid for. They’re free. 118 Access to care benefits. In addition to drug access, some cancer patients were perceived to want to participate in CTs because they believed the extra clinic visits, blood work and tests would lead to better care. One CT nurse shared her belief that patients appreciated the more intensive follow-up if they participated in CTs: They feel confident that someone is looking at their results of their scans and their blood work and somebody’s on it, somebody is following it closer than if they weren’t on a clinical trial. Many CT personnel also believed patients would receive access to “preferential treatment” within a trial. As one participant stated, “I think there’s a correct perception that if you’re in a trial, your care is better.” When asked where patients might get this perception, a radiation oncologist explained it might come from the consent form: It’s often implied actually within consent forms that the follow-up care will be more in a randomized trial than it would be for routine care. I’m not quite sure that I’m happy about that as a statement but it’s certainly at least implied in all the consent forms I’ve seen. Patients who participated in CTs were perceived to receive additional nursing support, including a singular trial nurse that organized their appointments and addressed their questions and concerns. This additional support was seen to improve patients’ continuity of care and helped them navigate an otherwise increasingly fragmented cancer care system. Sometimes this added benefit was not fully appreciated until patients came off a trial or withdraw from a trial, as one CT nurse explained: But I think that generally when patients are not on clinical trials, they do realize the care, the attentiveness that they had for one person organizing their appointments and one person to contact when they had problems is kind of value versus them calling different numbers for different problems. In general, CT personnel believed CTs addressed a structural gap in care and were better able to support vulnerable patients than standard care. One oncologist reflected on how 119 the current cancer care system does not provide the care and attention all patients required in the standard treatment setting. This makes the option of a CT and all the associated care benefits considerably more appealing to most patients: And I think there’s perhaps a little bit more of that now than there used to be because the cancer care system has been hacked to bits and you know, people who aren’t on clinical trials don’t tend to get the TLC perhaps that they used to. Complexity of trials. Some CT personnel identified trial complexity as a barrier to centres being able to host CTs because there were insufficient resources to support and coordinate complex CT procedures. As one CT nurse explained: But do they want to do it or is it a really cumbersome trial, or is it – or we can’t make it work in our building, i.e. that we can’t, if they need three CT scans or something a week, well we can’t do that here because there’s such a wait list for CT scans for example. As a consequence, patients located in these centres were denied the opportunity to participate in such CTs. Instead, complex CTs were typically offered in urban hospitals or cancer centres that have more resources for their execution. However, CT personnel believed the predominance of trials within urban cancer centres made CTs less accessible to rural patients. Patients who live far away are required to pay more out of pocket to attend a trial because compensation for travelling to the study site is not supported by CT funds, especially if it is a collaborative oncology trial as opposed to a pharmaceutical drug trial. As one CT nurse elaborated: There’s not a lot of pockets of money for people to come to [city] to participate in a clinical trial. We do occasionally pay travel expenses but it’s not total reimbursement, it’s usually a partial pay. And that would only be trials that are drug company trials, run by pharmaceutical companies. The Cooperative Group trials through, like, National Cancer Institute of Canada, there’s usually no payment other than we can occasionally pay for parking. 120 Complex versus simple trial protocol design was also seen to influence patients’ ability to understand what CTs involve since the presence of randomization was perceived to make a trial more complex and difficult for patients to understand. Complex protocols also influenced physician’s gatekeeping decisions about which patients should be offered the option of CTs based on their assumptions about patients’ competency and ability to grasp randomization. One oncologist believed studies will only increase in complexity as cancer care itself is getting more complicated; he perceived this clinical reality as a structural barrier to patients’ CT decision making: I think the nature of some of the trials we’re doing are just so complex now. Even in our site, we’ve got a couple of these British trials where they have multiple randomization and sub randomizations. I mean, one randomization that people get, what…what do you mean a randomization? But then when you get into multiple randomizations and they can choose arms and all this kind of stuff. That’s confusing, that’s a hindrance. And then if the modalities that are being studied are complicated, people just…I mean, cancer care is getting more complicated. And it’s not just the treatment they’re studying but also the extra tests required. There’s a lot more translational sub-studies and stuff, and people go, oh, my God…so it’s overwhelming. That’s a hindrance. So they’re already loaded with multi-modality treatment programs and multiple appointments and stuff, and the fact that the system around them has become much less well resourced. Consent forms were also seen as particularly problematic. Despite the research ethics requirement that forms should be written in lay language and typically at a grade six reading level, CT personnel perceived them to be too detailed, which made concepts such as randomization difficult to understand. As one CT nurse stated, even basic information about CTs is difficult for patients to assimilate: There’s stuff in the consent that the patient may not be able to put into terms of a real life context in terms of the amount of time that’s going to be required for their participation and what the visits might look like, what the tests for the workup are while they’re on study and making sense of a long list of side effects. 121 Oncologists believed too much information in consent forms was actually a hindrance to patients’ autonomy and caused some patients to feel “paralyz[ed]”, where they felt unable to make a good decision. As explained by one oncologist: You know anything happening to a person, you get into what’s going to be done with their tissue or whatever and then you get into where the documents go and who has access. It [the consent form] just goes on and on and on and somehow that doesn’t, I think, work. Somehow I think that makes people less autonomous in terms of decisions because I think you’re over informing them and initially at least, confusing them. Difficulties with understanding information provided in the consent form was perceived to be more apparent among non-English speaking patients who often relied on family members to interpret complex trial information. This underscores the need for CT specific translation services or education for non-English speaking populations: I cannot judge what the interpreter has interpreted to that person and you know usually if there’s a family member, then you want the family there also because it’s just better to have more, but yeah. And that’s becoming a bigger issue with more and more of the studies and the more complex, the harder it is on the interpreter and the interpreting of the information. Finally, the timing of when to provide CT information was also seen to be important for patients’ understanding. For patients who were newly diagnosed, CT personnel perceived these individuals as not being ready to learn about CTs. Instead, as one radiation oncologist described, these patients are focused on their prognosis and standard treatment options: So to then give them additional information about a study, which in the big scheme of things is much less important than things like, am I going to live or die, what treatment am I going to have, what are the side effects, which are really the key things the patient needs to know, so it’s probably not the ideal time to talk about a trial. When CT personnel did discuss trial information with patients, one of their main concerns was about having sufficient infrastructure to support their discussion. Sufficient time needed to be built into clinic hours to provide information and decisional support to 122 patients, especially when patients were very anxious and needed a lot of support. As shared by one CT nurse: A lot of people are very anxious when this is going on, especially if it’s – they feel that they’ve got disease, they want it treated and that and so you can talk to them and kind of explain something to them and you can’t explain, even though they read it, you can’t explain every detail of it in 15 minutes, you know when time is a huge factor. Even though CT personnel tried to give patients enough time and space to digest the information and explore both experimental and standard options, CT personnel sometimes experienced pressure to obtain informed consent quickly due to pre-defined windows of eligibility to enroll patients into trials. As one CT nurse stated; “there’s a bit of a time crunch for getting patients on a clinical trial.” If too many weeks or months pass, patients may not still be eligible for the CT. Other times, trials may be dependent upon, for example, receiving tests within a specified window. As one CT nurse commented, this can create a “high pressure” situation to obtain consent within a matter of days: For the ones [CTs] with the shorter turnaround, then I need to know much sooner than that. So usually they’ve been seen on a Wednesday or a Friday, Tuesday, Wednesday or a Friday. If they’re seen on a Tuesday, I’ll call them on a Friday. If they’re seen on a Friday, I’ll be calling them on Monday, kind of thing, to know, okay, do you want to do this? All right. Now you need to come in. This is what you need to do. Because often when they’re quick turnaround, we need them to sign consent before we can send tissue, it’s, like, nothing starts until you sign that consent. So you need to come in, you need to think about it, you need to work on that. So there’s high pressure if there’s fast turnaround. Recommendations Interviews with CT personnel explored best practices for supporting patients’ relational autonomy in their decision making about CT participation. The recommendations discussed below include suggestions from CT personnel about ways patients can be personally, socially and structurally supported in their CT decision making. 123 Provision of Clinical Trial Decision-making Support CT personnel spoke about the need to provide “individualized” support to patients when they are faced with making decisions about CT participation. This could include tailored support for synthesizing information, psychological support for decision making and promoting patients’ sense of control and autonomy over the decision. CT personnel believed support from the healthcare team was especially important when families were not willing or available to take on a supportive role. One medical oncologist described his strategy for supporting patients: And I think I spend a lot of time talking to my patients about the study and what it means and the benefits and the risks. And I think they appreciate that […] I actually sit down and I go through the whole consent and study and the pros and cons as well of going on a study versus not going on a study. And I make myself available for any call that they want. Other CT personnel explained the necessity of alleviating the fear or burden of making a wrong decision through the provision of information. This was an important part of supporting patients’ relational autonomy and CT decision making. As one medical oncologist explained: I find if you give them more information and relieve them of the burden of making a wrong decision, that any decision they make is the right decision for them. More often than not, that actually helps for them to decide to do a trial in fact, once they have that belief taken away from them. CT personnel also described how they might offer reassurance to patients by reminding them that they can always withdraw from the CT and that “they will still get good care even if they’re not in a clinical trial.” This gives patients a feeling of control, and relieves their anxiety, as described by one CT nurse: So those people who can’t make up their minds, to help with their autonomous decision is reassurance and often, the only way that I reassure people is by saying, and the doctor says this too, at any point in time you can get off the trial. So here’s 124 what we think, here’s what we believe, anytime, you don’t like the drugs, even if the drugs are not affecting you. You just don’t like being here, whatever, you don’t like all the appointments, the way they’re set up, you can leave. Any time at all you can leave and that helps people a lot to feel like that they are making a decision that they have a bit of control. Mobilization of Social Resources CT personnel recognized that it is often necessary to provide culturally sensitive decision-making support for patients. One radiation oncologist emphasized the importance of appreciating the patients’ perspective and then mobilizing resources from the patients’ cultural community in order to facilitate their decision making: So for some patients who need the interaction with their culture, with cultural caregivers or community in order to derive a decision, I mean my job is to facilitate that. I mean I’ve got to help people come to a decision using whatever resources they have, not whatever resources I have, whatever resources they have. One of the key recommendations from CT personnel was to recognize how important family and community are for supporting cancer patients’ relational autonomy and decision making about CTs. One CT nurse remarked “they’re [patients] more able to be autonomous if they have good family bonds and good social backup, good connections out there.” She explained the role of social connections in the following way: I think they help them to function. They might be supportive in other ways, you know, they might provide some kind of a service to them so I think, if patients are socially connected, then they’ll have a GP. They’ll also have possibly some home care coming in or some kind of assistance. And, if they have family, then I think the family members can help to support them emotionally but also financially sometimes. Some of these patients are actually living with family members so finances are a big problem when you’re ill, especially with cancer because, you know, people sometimes have to go on disability. Sometimes they have to retire early or sometimes they’re even let go or their business falls apart if they’re self-employed, they have their own company. So family can help out in financial dire straits, too. 125 Attention to Structural Barriers CT personnel considered overall system issues and how CT infrastructure could be re-developed to support patients’ assimilation and appreciation of CT information. One radiation oncologist suggested a triage system whereby patients who were eligible for a CT were followed up separately in a clinic that allowed for more time to discuss the trial: I think it’s the whole sort of lack of infrastructure and timing, time issue. If we had, if we knew up front who was definitely not eligible for a study and who might be eligible for a study, then you could conceive of a system whereby, instead of having one clinic, say as a follow-up clinic, and patients scheduled every fifteen minutes, or every ten minutes or whatever it is, you could have like a clinic where people are scheduled every twenty minutes, and there’d be other clinics where there’s no question of a trial every five minutes you know, so that you could create more time. Reflecting on structural influences, and how to better support patients’ relational autonomy in trial decisions, particularly CTs that require a quick turnaround, one CT nurse expressed some exasperation about how to allow more time for patients to digest CT information. This nurse decided that it might come down to re-designing trials to make them more patient-friendly: I think that if we can give them as much time as possible, to not pressure them. But I don’t know how to fix that, because with some trials you can and when they have the time, they are given the time, and with others you can’t because that is just the way the trial is built. So if you want to help with that, then you have to go back even further and say, don’t write trials that are like this. Many CT personnel believed communication resources to support the provision of trial information in a patient’s own language was also important for patients’ relational autonomy and decision making about CTs. This suggestion could apply to all trial offers, not just cancer CTs. For example, when patients are not English speaking, having a doctor or a family member “who speaks their language, can read the English and then talk to them about 126 it” was seen as helpful. This may be particularly important for consent forms and other trial material, as suggested by a CT nurse: We could try to provide them with consents in their own language. Questions in their own language. We try that now. We do. We ask, because lots of clinical trials have lots of quality of life questionnaires and they are supposed to be only done by the patient, not even a family member. And so it’s hard to do if you do not speak the language or read the language. And it’s hard to understand. And now most of our population, we have a huge population that doesn’t speak the language. And so we should really, to make it easier for them, we should provide them with information that they can communicate in. Summary of Findings The present discussion has identified several overarching themes within the CT personnel findings that involve personal, social, and structural influences on cancer patients’ relational autonomy and CT decision making. In this summary, these themes are explored in greater detail in the context of the findings. In Chapter Six, these themes will be more broadly situated in relation to the current literature and theoretical lens of relational autonomy. In examining the personal, social, and structural factors in CT decision-making, three major themes are apparent from interviews with CT personnel. First, there is an apparent power differential between patients and oncologists that may place undue pressure on patients’ trial decision-making. Patients are dependent on oncologists for their continued care, which creates a strong incentive for them to make trial decisions that do not jeopardize their relationship. Second, findings demonstrate the pervasiveness of the therapeutic misconception in trial discussions, both for patients and CT personnel. Oncologists may present and discuss CTs in a way that makes it difficult for patients to distinguish a CT from usual/standard care. This may be intentional or inadvertent, as oncologists may also be confused about the distinction. Linked to this is the concern that patients do not understand 127 the nuances of what is presented to them, often because of their anxiety or fear, and lack the ability to provide truly informed consent to trial participation. Third, justice issues arise in relation to selective gatekeeping by oncologists and equitable access to CTs across different socioeconomic, geographic and demographic groups. Such practices and barriers have implications for appropriate randomization and fair access to CTs especially for populations marginalized by factors such as geography, poverty, immigration status and language. The following sections elaborate on these themes in greater detail and touch on the ways in which personal, social, and structural influences are all intricately involved and overlapping in these issues. Power Differences Between Patients and Physicians Findings indicate that patients with cancer are vulnerable to many strong emotions throughout their cancer trajectory. CT personnel identified hope, fear, anxiety and distress as especially prevalent emotions that a majority of patients experience in response to their cancer diagnosis. These emotions make it difficult for patients to appreciate CT information, and many cancer patients are thus dependent upon CT personnel for support to understand and undergo care. For example, CT personnel interviews revealed that cancer patients often experience psychological distress and anxiety as a result of receiving their cancer diagnosis, which places them at a disadvantage when it comes to feeling capable of receiving information and empowered to make decisions. In addition, patients often struggle to comprehend their diagnosis, prognosis and treatment options because of their unfamiliarity with the cancer care system. This creates unease and uncertainty for patients when they approach consultations with their oncology healthcare provider. Finally, CT personnel have 128 recommended that patients need substantial social (i.e., provider as well as family) support to assist them with the logistics of CT participation. For all these reasons, cancer patients are heavily reliant on CT personnel, especially their oncologists. These physicians hold medical knowledge that patients require to make treatment or trial choices but that they themselves lack. Patients’ attempts to seek medical information such as their risk of cancer recurrence/growth if they participated (or not) in a trial from CT nurses were unsuccessful because of their restricted scop of practice. Patients were instead referred back to their oncologist to have more nuanced medical discussions. This reinforced a power imbalance where patients are dependent upon physicians. It would come as no surprise then that patients would make trial decisions to maintain good relationships with their oncologists so that they may continue to receive care and assistance. Furthermore, if patients are under emotional distress they may not want to engender conflicted relationships that may add to their distress. This is played out in the findings with CT personnel where it was discussed that some patients accept trial participation out of a desire to please their oncologist or out of a duty to give thanks for the care they have received. Other patients are reported to make CT decisions out of the fear that they will suffer consequences to their care if they do not participate. These findings attest to the significant influence oncologists have on patients’ relational autonomy and CT decision making, and raise concern that patients may feel pressured to make trial decisions to preserve relationships with their oncologists. Therapeutic Misconception Another important issue identified from interviews with CT personnel is the perceived dual role of the treating physician who is also the physician-researcher. This has 129 the consequence of encouraging patients’ therapeutic misconception of trials, which is defined as the mistaken belief that the purpose of CTs is to provide individualized care rather than further scientific understanding (Appelbaum, Roth, & Lidz 1982). When oncologists introduce the concept of a trial in the clinic, the patient is already conditioned to engage in a discussion about treatment rather than CTs. There is also the perception that oncologists may downplay the experimental nature of trials and the associated risks and sometimes use the terms “trial” and “treatment” interchangeably in accrual discussions. Many CT personnel, however, perceived oncologists to have the best of intentions in blurring the line between treatments and CTs. They may perceive a trial as the only way to access a treatment option for their patients and may downplay the experimental nature of CTs if they are also aware of patients’ hesitancy to participate because of their fears and beliefs about trial participation. However, presenting trials in this manner may mislead patients into believing that the trial intervention is known to be efficacious when, in fact, there is still medical uncertainty about its effectiveness and risk profile. Even if recruiting physicians genuinely believe the trial offers the most effective therapy for their patient, if other members of the expert community disagree they are obligated to disclose this to their patients. This moral requirement is rooted in clinical equipoise, which Canadian research ethics guidelines stipulate must be present in order for trials to be carried out ethically (CIHR 2010). Clinical equipoise is obtained when there is genuine uncertainty in the medical community over which arm of a trial is superior. Once it is established with certainty that one arm is more beneficial, all patients are entitled to receive the superior treatment (Freedman 1987). Structural issues exacerbate the therapeutic misconception and impact patients’ understanding necessary for CT decision making and informed consent. For example, the 130 cancer care system and CT appointments are not structured in such a way that oncologists, who are often involved in recruiting to their own studies, are able to discuss trials separate from treatment. Therefore, oncologists have no choice but to discuss trials alongside treatment and within a limited time slot that makes having a more thorough discussion with patients difficult. Language or literacy barriers may place further challenges on the trial discussion and on patients’ ability to understand the information imparted to them. In terms of providing unbiased information for patients’ informed consent, findings suggest there are professional and institutional pressures for oncologists to recruit for or champion trials that are running at the agency or in their department. This may influence the trial discussion and encourage patients’ therapeutic misconception even if the physician presenting the trial is not directly involved or is careful to remain neutral in his or her discussion with patients. As the findings indicate, subtle clues and biases are difficult for providers to identify and control. Oncologists’ body language, amount of time they allot to discussing trials, and how they phrase the CT discussion may reveal information to patients about how they feel about a trial, suggesting the need for professionals to ensure that these pressures do not unduly influence the trial discussion and encourage the therapeutic misconception in patients’ CT decision making. Equitable Access to Clinical Trials Finally, findings indicate that not all cancer patients are equally able to access CTs. Strict trial eligibility requirements, pre-screening or gatekeeping behavior by oncologists, and lack of trial resources are social and structural influences that are perceived to affect cancer patients’ access to trials. Because of insufficient resources directed towards CT accrual activities, oncologists are forced to pre-screen patients for recruitment into trials. Persons 131 with mental health issues and addictions are often selectively denied access to CTs because they are perceived to not have the up-front capacity to understand the trial and therefore require more resources to support their understanding of and participation in CTs. Lack of family or other social support and poverty can also negatively impact patients’ ability to access CTs due to the assumptions oncologists develop about these patients’ ability to participate in CTs. Also, interviews suggest rural cancer patients are approached less often to participate in CTs because of assumptions made about their financial means and the lack of funding to cover the cost of travel to the study site. However, the resultant pre-screening practices by oncologists may be based upon erroneous assumptions about the capacity of certain groups to take part.13 The result is that already marginalized groups, those who have mental health issues or addictions and those who have lower socioeconomic status or who live in rural areas, will unfairly have decreased or no access to CTs. This impacts not only the ability of these persons to access potentially helpful interventions but it also compromises the scientific integrity of CTs since interventions would not be appropriately randomized across all patient groups. Thus, in an era of evidence-based medicine where the synthesis of the best available randomized controlled trials inform practice and policy decisions, the generalizability of study findings may be limited, which may result in unknown risks and benefits to these populations (Rogers 2004). 13 Pre-screening practices are also a result of systemic barriers to oncologists’ time to discuss trials with patients. 132 Chapter Five: "No Wo/Man is an Island"—Cancer Patients' Experience of Autonomy Related to Their Decision-Making Process about Clinical Trial Participation This chapter presents the results from the grounded theory study of cancer patients’ decision-making process and explains how patients made decisions about clinical trials (CTs). Support persons and CT personnel perspectives were used to provide further insight into patients’ decision-making process and relational autonomy. The overall aim of the study was to develop a theoretical model that illustrates patients’ decision-making process in order to provide a foundation for the development of decision support strategies that will enhance cancer patients’ relational autonomy and inform how they are approached, recruited, and enrolled in cancer CTs. A total of 24 breast, 15 prostate, and 1 bladder adult cancer patients who had accepted (n=32), declined (n=6), or withdrawn (n=2) from a CT participated in this study.14 Most of the CTs discussed by patients were phase II or III pharmaceutical trials (n=37). The majority of participants were between 50-70 years old (n=23), were Caucasian (n=35), and were married or common law (n=29). Almost all the participants (n=34) had a college/trade or university degree, including graduate degree. Patients were representative of a range of disease stages. Some had advanced disease (n=16), others had intermediate disease (n=10), and the rest had early stage of disease (n=13). Additional demographic characteristics of the patient participants can be found in Table 5.1. 14 The percentage of acceptors, decliners, and withdrawers from CTs reported here are reflective of patients’ current or most recent CT decisions. However, during the interviews, some patients reflected upon their previous CT experience, in which they made different decisions. Every effort was made in the reporting of patient quotes to cite the relevant CT decision (past or most recent decision). 133 Table 5.1 – Reported Patient Characteristics Characteristic Frequency (%) Characteristic Frequency (%) Age Self-Reported Ethnicity <40 years 1 (2.5%) Caucasian 35 (87.5%) 41-50 years 6 (15.0%) Asian 2 (5.0%) 51-60 years 11 (27.5%) Aboriginal 1 (2.5%) 61-70 years 12 (30.0%) African Canadian 1 (2.5%) >70 years 7 (17.5%) Unknown 1 (2.5%) Unknown 3 (7.5%) Type of Cancer Gender Bladder 1 (2.5%) Male 15 (37.5%) Breast 24 (60.0%) Female 25 (62.5%) Prostate 15 (37.5%) Marital Status Stage of Cancer Married/common-law 29 (72.5%) Early 13 (32.5%) Divorced/ widowed/ single 10 (25.0%) Intermediate 10 (25.0%) Unknown 1 (2.5%) Advanced 16 (40.0%) Education Unknown 1 (2.5%) High school diploma 5 (12.5%) Trial Decision College/trade school 7 (17.5%) Accepted 32 (80.0%) University 13 (32.5%) Declined 6 (15.0%) Graduate or professional degree 14 (35.0%) Withdrawn 2 (5.0%) Unknown 1 (2.5%) Introduction to the Theory The following paragraphs provide an overview of the theory describing cancer patients’ decision-making process related to CT participation. The theory identifies key phases, processes, and sub-processes of patients’ decision making. More fulsome explanations, along with supportive participant quotes, will be used in the body of this chapter to elaborate on the theory and central core construct “No Wo/Man is an Island”. No Wo/Man is an Island “No wo/man is an island” was the central core construct describing breast and prostate cancer patients’ CT decision-making process. This construct captured the relational complexity of cancer patients’ CT decision-making and made visible the various key influences on this process. In particular, significant relationships, including relationships with 134 partners, family and extended family or friends, as well as the broader social and structural environment influenced patients’ CT decision making. A patient explained: …no person can do everything by themselves - you know, no man is an island. But anyway, I took advice from many people and they pretty well all had the same idea [to accept CT participation] (male, accepted). A CT nurse who participated in the study described how social support (e.g., supportive family, friends, and physicians) intertwined with structural support (e.g., accessible healthcare, economic resources) to create a relational system that facilitated cancer patients’ decision-making process about CTs: I think they [family bonds and social connections] help them [patients] to function. They might be supportive in other ways, you know, they might provide some kind of a service to them. So I think, if patients are socially connected, then they’ll have a GP [general practitioner]. They’ll also have possibly some home care coming in or some kind of assistance. And, if they have family, then I think the family members can help to support them emotionally but also financially sometimes. Three general phases15 made up the “No Wo/Man is an Island” theoretical model (see Figure 5.1). Consistent with the core concept, a relational component was imbedded throughout all phases, processes and sub-processes. This reflected not only how relationships influenced how patients moved through the phases in the model but also encompassed the larger socio-political influences (e.g., availability of resources, geographical location) embedded in the theoretical model. The first phase occurred prior to the trial offer and related to the influence of predisposing factors on patients’ level of receptivity to receiving CT information from their physicians. In this phase, social and cultural norms around cancer CTs, along with familial 15 Bolded or italicized text in this chapter highlights major themes and processes that were identified during data analysis. 135 experience were key influences on patients’ beliefs and attitudes. For example, personal and cultural understandings of research and medical technology, family and support networks’ opinions, previous CT / healthcare experiences, and social media were highly influential. These factors then interacted with one another and affected how “open” or “closed” patients were to learning about CTs.16 The second phase of the model was the trial offer phase, which referred to the physician’s initial presentation of the CT to the patient. Within this phase, physicians’ framing of the trial and their non-verbal cues proved to be important for patients’ initial assessment of the trial and was helpful for them in gleaning their physician’s opinion of the trial. Whether a trial was even offered as an option at this point was an important structural precondition for patients’ further engagement in the “No Wo/Man is an Island” model. This phase highlights the important role of the patient-oncologist relationship related to patients’ decision-making process and the socio-political context in which these relationships were imbedded. The third phase was composed of patients’ CT decision making within which there were three processes: seeking information, establishing credibility, and weighing pros and cons. The weighing pros and cons process further consisted of several sub-processes. These included doing what is right for me/considering altruism, weighing adverse effects/benefits, grappling with uncertainty, considering logistics and maintaining relationships. Depending upon whether the type of decision making was snap or reflective, patients had more or less 16 How “open” or “closed” patients were refers to how receptive patients, who received a trial offer, were to hearing about CT information. As such, these concepts/themes do not apply to patients who were not offered a trial due to gatekeeping behaviours by physicians. 136 engagement in these sub-processes. Snap decision-makers tended to seek minimal information and established the credibility only of their physician or care team before making a decision about whether or not to participate in a CT. These decision-makers bypassed or only minimally engaged in weighing pros and cons and related sub-processes. In contrast, reflective decision-makers engaged in all processes and sub-processes. Within this phase, larger social norms and expectations around cancer CTs, socio-political barriers and facilitators of patients’ CT decision-making, and the role of social networks were important influences in guiding patients through these processes. Finally, a broad range of influential factors informed the decision-making phase of the model. These factors included patients’ stage of disease, trial type (whether it was an exercise or a drug trial), previous and current healthcare experience, and gender. These factors influenced what kinds of and how much information was important to patients, to what extent trust and credibility was significant, and how patients weighed the pros and cons of CT participation. Additionally, stage of disease and trial type influenced patients’ weighing of pros and cons through the mediating concepts of risk tolerance and choice perception. Patients who reported more advanced stage of disease, for example, demonstrated greater tolerance for uncertainty associated with the trial (i.e., whether or not they would receive the active agent, whether or not the agent would be effective for them) and minimized the risk of adverse effects because they perceived no other treatment options available. At the end of the three, mostly sequential phases, patients ultimately reached a decision to either accept or decline trial participation. However, some participants who originally accepted returned to the decision-making phase, reconsidered their decision, and 137 withdrew from the trial. The overall theory highlights how CT decision making is not only personally but also socially and structurally located and demonstrates key phases and processes that patients go through in order to make a CT decision. In the following sections a more in-depth description is provided and supported by quotes from participants. 138 Figure 5.1 - "No Wo/Man is an Island" 139 Phase I: Prior to the Clinical Trial Offer - Predisposing Factors Prior to being offered a trial by their physician, predisposing factors such as patients’ pre-conceived beliefs and attitudes about CTs influenced whether patients were receptive or not to participating. These included patients’ beliefs about being a “guinea pig” or general orientation towards medical technology that was informed by their previous healthcare experiences, conversations with friends or family, cultural understandings of CTs, and social media. For example, a few patients expressed a fear of being a “guinea pig”, which they negatively equated with being experimented upon. This typically stemmed from their cultural understandings or personal or familial experiences of research experimentation as potentially exploitative. In contrast, other patients tended to view experimentation as advancing medical technology and as the answer to unlocking the mechanisms of disease and treatment. These patients perceived volunteering for a trial as heroic, which was supported by one patient’s spouse as something to “feel a little bit proud that maybe you can hopefully make a bit of a difference and help others after you.” Another patient with prostate cancer explained how he was thinking of CTs in relation to the social good: Because like I say, I don’t mind being a guinea pig. If I contribute, you know, it doesn’t matter whether it’s successful or it’s failing. But if you don’t offer yourself to be part of this experiment, how could they tell the end result? In other words, the volunteer is very important, they need the volunteer to gain the approval. I mean the result is definitely through the data of the patient, right? (male, accepted) Patients also described how their attitudes about trials were informed by their conversations with family or friends who had strong opinions about CTs. A close relative’s, friend’s or family member’s previous experience in a CT often informed whether a patient’s attitude towards CTs was positive or negative. Patients’ own previous CT experience or 140 beliefs about research also coloured their perspective going forward. One breast cancer patient discussed how her professional experience allowed her to gain an understanding of research and fuelled her initial positive appreciation of the study: I also worked as a nurse in the research department at the cardiac department at [name of hospital] and they [studies] were all pharmaceutical ones. I was not impressed with all of them. So I don’t see all research as being for the best for everybody, but in general, I will say that, yes, I am particularly for breast cancer and other women’s cancers that the research is what helps move things forward so there’s better care and less toxicity and less damage to the person. So I’m willing to participate (female, accepted). Another patient discussed her previous experience on a fitness trial for cancer patients and how she found it extremely positive, which informed her current decision about whether to go on a drug trial: And then I would say I have had the luxury thus far of participating in a clinical trial that really had no downsides to it… there was only an upside, in my opinion, to participating in a fitness trial. And you know, it brings along with it a support group and everything (female, accepted). The above quote illustrates how positive (or negative) previous experience can potentially mislead patients into thinking that all CTs are conducted in a similar manner. Therefore, it would be important for patients and oncologists to acknowledge previous experience but also focus on the details of the current CT being offered. Patients’ beliefs and attitudes, previous CT experience and conversations with others influenced their degree of receptivity or how “open” or “closed” they were to receiving a trial offer. Rather than discrete concepts, patients’ level of receptivity existed on a continuum whereby some patients were adamantly closed to learning about CTs whereas other patients were very open. Those patients who were very open to CTs typically had positive beliefs and attitudes or previous healthcare experiences that encouraged them to orient themselves in this way. Some believed oncology trials represent the cutting edge of cancer research and were 141 generally more optimistic about studies having positive results and helping others. These patients may have actively sought out CTs or approached their oncologist about them instead of waiting for a trial offer. One patient discussed his feelings of optimism about participating in CTs: I like to be on the leading edge. So far they haven’t found the magic bullet, but you do get this feeling of optimism when you know a lot of work and research has gone into developing these compounds, and you might be one of the first people to be taking it, so I generally feel sort of optimistic, knowing that when all else has failed, this might be the substance that will succeed in slowing down the growth of the particular cancer that I have. I go to them generally with a feeling of optimism for those reasons (male, accepted). For some patients, social media, such as general news broadcasts, radio or television shows and newspaper articles, had a positive or negative influence on personal beliefs and attitudes towards CTs. For example, a few patients developed initial concerns about a trial’s safety after learning about drug-related side effects in a newspaper article. Other patients heard more positive messages about CTs. The following quote illustrates the power the media had in fostering a prostate cancer patient’s optimistic attitudes towards CTs: Oh I think it’s [CTs] a marvellous thing because I mean every time you turned a radio on or turned the TV on, there’s always some new developments with cancer. You know, they’re getting a step up on this; they’re getting a step on that. Oh yeah, it’s very positive (male, accepted). Patients who were closed to CTs typically had beliefs and values or previous healthcare experiences that caused them to be more skeptical of research and cautious about participating. Relating back to the core construct, these patients’ beliefs and values were, in turn, often influenced by their social network. For example, a family member might have had a negative experience in a CT (e.g., experiencing numerous side effects, received placebo) and passed along his or her anxieties to the patient in conversation about trial participation. These personal and social factors made patients more hesitant about taking part in a trial: 142 And my experience has been - not in the trials, I only did do one trial, not with the [name of cancer agency], it was a pain trial—when they said 2% of patients have these side effects - I had them all! Same thing when I would go for surgeries. Most patients tolerate this well but 2%…so I was always on that cusp of the bell curve, way off the other end. So I got to dislike the term ‘most patients’ and I got very weary of it because I learned that what I had to look for is what are these 2% side effects. What are they. And if they didn't meet my criteria, just out immediately. Period. And that was the basis of most of my readings. It wasn't that the trial didn't look good, the results looked good, the results looked promising. But I wasn't gonna trade side effects for the cancer. That was not my motivation (male, declined). Patients’ initial receptivity was a powerful influence on how they then engaged with the trial offer and decision-making phases of the model, such as what and how much information was sought. For example, patients who were closed or skeptical of trials because of previous experience with debilitating side effects on a prior trial sought in-depth information about side effects for the current trial. In contrast, patients who were more open to trials had a broader interest in many aspects of the trial. Level of receptivity also provided an initial starting position for receiving and interpreting information. As one patient with breast cancer explained, her initial openness to a trial became her “default”, and all other information was considered in terms of whether it would give her a reason to reject her initial decision: I just went through the thought process, is there any reason why I would decide not to do this, if that makes sense? I left with the default switch on I’m probably going to do this, and all the other stuff was just around whether or not something would move me off that position (female, accepted). Phase II: Trial Offer This phase of the model was complex and inherently relational. Patients interpreted the way in which a trial was presented and the demeanor of the oncologist in making a judgment about the relevance and value of the trial for them. For example, a patient with 143 breast cancer reflected on her oncologist’s clear presentation of the trial but believed her overall positivity reinforced her initial openness to participating in the study: Well, I think she [the oncologist] was very positive about the studies. I think she was very fair in that it wasn’t presented in a manner that would have made me worry if I hadn’t gone on the study. Like, I don’t mean to make it sound like she over-sold it; she was very clear on the pros and cons, but she was very positive about it and I guess maybe that’s sort of the overarching element; she was quite positive that there’s this opportunity, you know… kind of discussion about who else and what else and where it was in its program and that it seemed to be a very well-tolerated drug and stuff. I think her positive approach and good initial descriptions kind of right away gave me a good feeling that, well, this is probably not a bad thing to do. I left with a positive feeling about it (female, accepted). Another patient recalled how the positive presentation of a trial supported his initial openness to participating in the study. Reflecting on his decision, he acknowledged that potential side effects were not presented to him in a clear and straightforward fashion. He believed that if they were, this would have caused him to “think twice” about the trial and reflect more on the consequences of CTs, thus calling into question how informed was his consent to participate: I was very eager to do these trials. And, in fact, with one of the trials I signed up before I read it. Probably not a good idea, but … Where do I sign? I think things were never presented to me in fear terms either. They were presented in positive terms. If somebody had approached me and said, ‘Here’s a great trial. You’re going to be sick as a dog’, I would have thought twice. But they were always presented in positive terms (male, accepted). Some patients interpreted the mere mention of a trial as a recommendation to participate. This may have been partly a result of structural influences that created confusion between treatment and research, such as the CT presentation by the oncologist occurring during a patient’s regular clinic appointment. In one or two cases, oncologists were reportedly very explicit in conveying participation as a “no-brainer”, that the drug would be effective and recommending the patient participate. This statement supports the findings from the study with CT personnel, where patients were influenced by oncologists’ subtle and 144 overt “messages” about CTs. Most patients believed the oncologist was looking out for their best interests and appeared to extend the principle of beneficence that informs treatment decision making to the context of CTs. In other words, patients assumed the trial would not have been presented to them unless their oncologist thought it was a good thing for them to do: In all of these doctors that I’d been seeing – I’ve got three; I’ve got my GP, the urologist, and the oncologist. In all three cases, I’ve come to embrace what they are saying as the truth, and I agree with it, and I don’t have any qualms about following their advice. I understand that they all have my best health in-, you know, it’s what they’re trying to do, is make me healthy (male, accepted). The above quote highlights the relevance of the core construct driving the theoretical model. It appears that patients, although encouraged to demonstrate personal autonomy by making their own decisions, are likely to embrace relational influences on their decision-making processes. This is particularly the case given the nature of the patient-physician relationship and the vulnerability of cancer patients. What is striking in the above quote, however, is the potential for misunderstanding CTs as individualized medicine. This raises the issue of whether this patient was able to provide truly informed consent to participate in the trial and/or whether the patient was exhibiting well-placed trust in his physicians. These issues will be explored further in the discussion (Chapter Six). Within the context of the trial offer, the majority of patients assessed their oncologist’s level of enthusiasm for the study in order to ascertain whether he or she thought it was a good thing for them to do. Was the oncologist excited about enrolling the patient in the trial or did he or she have concerns or reservations? Some patients outright asked whether the oncologist thought it was a good option for them, others were attuned to more subtle indicators that the oncologist seemed more or less keen, either through the physician’s body 145 language or non-verbal behaviour. One patient described how he “sensed” his oncologist needed someone in the trial, which was confirmed for him through a handshake after he agreed to participate: It took him about 20 minutes. He sat right beside me and he told me about this. And I knew he—I sensed he needed somebody badly. […] He sort of quietly talked to me about it. I put money as I told you earlier on him and I said okay, I'll do it. And he shook my hand warmly and said he was pretty happy about it. So he obviously needed another patient (male, accepted). Another patient was very attuned to her oncologist’s presentation of the trial and also recognized the need for participant accrual if trials are to move forward. This sensitivity to physicians’ attitudes and behaviours may reflect the perception found within the CT personnel interviews that cancer patients often seek ways to please their oncologist to demonstrate their appreciation for the care they have received. In this example, the patient detected a heightened level of enthusiasm from her oncologist’s explanation of the trial, which suggested to her that there was some bias towards her participating in the trial: The doctor wanted me in the trial very badly. They're trying to get their number of people that they want in the trial. And she had a lot of information, explained to me how they came about with their numbers and how many people were going to be on either side and how it was chosen and this and that. So she was quite encouraging in trying to have me take the trial (female, declined). This patient ultimately decided not to participate because her current treatment was working well and she did not want to take a new drug with uncertain risks and side effects. This decision, however, was very difficult and she required the assistance of another doctor, who encouraged her to trust her “gut instincts” and carry on with a treatment that had worked for her in the past. Although most of the patients I recruited into the study were offered a CT, during the interview some participants reflected upon the implications of not being offered the option of 146 a trial. Patients believed this would immediately foreclose the option of trial participation and any further decision making unless the patient was to actively seek out a trial on his/her own. One participant did have the experience of not being offered a CT initially. She believed her oncologist made assumptions about her ability or means to travel to the study site, which was four hours away by car. Because this patient was actively searching for CTs, had no other treatment options left, and had advanced disease, she confronted her oncologist and asked to be enrolled in the trial. This patient was fortunate in the sense that she was well educated and had an active social network that assisted in her identifying a CT for which she was eligible, thus allowing her to overcome the gatekeeping behaviour of her physician. From a relational autonomy perspective, other patients who are in less fortunate circumstances in relation to socio-economic and social support factors may have been less able to advocate for themselves and confront their physician’s assumptions. Finally, throughout the trial offer phase there was slippage between how oncologists and patients framed the trial. Sometimes the trial was referred to as “treatment” in the discussions, so it was unclear that the trial drug was actually experimental. As demonstrated by one oncologist’s general recollection of CT discussions: But if patients, with that particular trial, they say to me, ‘well which do you think is the better treatment?’ And I’ll say ‘I think this treatment’s going to have better cure rates, but it’s going to have more side effects, and so you as a patient need to decide if you’re willing to accept the risks of more side effects for a better long term cure rate’ (emphasis added). This lack of clarity encouraged some patients to consider the trial as potentially offering the best treatment option. This was particularly the case amongst patients who reported they had advanced disease and whose options were already limited. As shared by a prostate cancer patient: 147 I mean, that was a treatment option. I mean, surgery or radiation, those are all gone for me. So trials, that’s great. Go for it man, you know? What else am I gonna do? At my stage, it’s easy (male, accepted, emphasis added). These quotes illustrate the potential for the therapeutic misconception and the possibility that patients are not making informed decisions about trial participation. These issues will be explored in greater detail in the discussion (see Chapter Six). Phase III: Decision-making Once patients were informed of the option of trial participation they then moved into a series of processes that made up the larger decision-making phase. This included seeking information, establishing credibility, and weighing pros and cons. Seeking information and establishing credibility were complementary and informed each other. The majority of patients moved back and forth between these two phases. Many patients then progressed to weighing pros and cons but returned to seeking information and establishing credibility if new questions arose. Other patients did not continue into the weighing pros and cons process. As will be described below, these patients went directly from seeking information and establishing credibility to making a CT decision. Seeking Information Type of information sought. Patients sought a variety of information that they believed was important for their decision making. For the majority of patients, physical effect was foremost in their minds. Patients wanted to know what their oncologist thought about the trial and whether the trial intervention would be effective in helping them survive their cancer. The likelihood of experiencing serious or unpleasant adverse effects, as well as whether participating would require prolonging or temporary stopping their current treatment, which could pose a risk regarding the spread of disease, was also valuable 148 information for the majority of patients. Consistent with the relational autonomy framework, the oncologist was also perceived to be very important for supporting some patients’ interpretation of the CT information, as demonstrated by one patient with breast cancer: So reading all that stuff about the possible side effects, it is true and they have to tell you that, but it’s less of an influence than the conversation with your doctor. Because you need to be able to put some perspective on the likelihood and the severity of these [side effects] (female, accepted). Whether the trial design involved a placebo and the chances that they would be randomized to a non-experimental agent was also information that patients prioritized in the information-seeking process. Other health benefits they might receive from the trial, such as closer physiological monitoring (i.e., extra tests and blood work), additional nursing support or follow-up care were also important. Similarly, understanding the logistics of the trial was significant so that patients could assess the potential burden on themselves and their family members. Logistics included how long the trial would last and how far and how often the patient would need to travel to the study site. Sources of information. Patients sought information from their oncologists, CT nurses, GPs, fellow patients, and family. They also sought information from social media and, occasionally, scientific journal articles. After receiving trial information from their oncologists, patients typically went home and searched online for further information about the trial intervention. They used their social networks to direct them to relevant information, support them in distinguishing credible from non-credible information sources and help them assess the information they collected in order to situate it in the context of their own lives. This often entailed weighing risks and benefits in light of their personal and family goals and values. Thus, seeking information was a cyclical process where patients continually asked questions and sought further information to address gaps in understanding. One patient with 149 breast cancer shared how she consulted with her husband at multiple times to seek his opinion and help her interpret information and seek further understanding, thus underscoring the importance of close relationships in this phase of her decision-making process: I found news stories related to the study and then I found more scientific information about the study parameters, not that I fully understood everything or most, but it gave me a pretty good overview. And it was over a course of a week I would say I kept checking stuff on the internet and talking to my husband about it and so it was a back and forth (female, accepted). Another patient, who declined a previous chemotherapy trial but accepted another subsequent drug trial, described the advantages of involving others in the information-seeking process: Well, I process through talking. So just talking to people and then, you know, you’re kind of going down a path and somebody might point something out and you go, ‘Oh yeah, okay, I’ll think about that one’ (female, declined/accepted). Some patients shared that they preferred to consult with healthcare providers to gain assistance in interpreting information rather than seeking help from family members because providers were viewed as the experts. This was particularly the case with some men, in whom traditional gender roles and previous professional experience appeared to influence information seeking behaviour. One prostate cancer patient stated the following when asked about his involvement of family members in his CT decision: Oh, not very much. Other than the fact that I wanna be around 'em. I don't ask 'em for advice or anything. I'm kind of a patriarch of the system and I don't seek advice from anybody. I was a manager, six departments in a company, I had 200 people working for me. What the hell. You get used to making your own decisions without input (male, had both accepted and declined CTs). The kind of information patients sought was heavily influenced by the type of CT being offered. For example, with exercise trials, patients primarily considered logistics rather than adverse effects because these trials were perceived to be generally less risky and the objectives more easily understood. Drug trials, on the other hand, were perceived to involve a 150 “life or death” decision and patients prioritized information about the potential risks and physiological side effects. These trials also required deciphering complex information. As shared by one patient with breast cancer who made decisions about both an exercise and drug trial: I think it’s a different decision to take internally a drug and participate in that kind of trial, than it is to participate in some time of external study, like the fitness trial, or I’ve seen ones to do with relaxation and other things like that. To me, those are different decisions. I think people in general have a better understanding of the outside-your-body ones than the inside-your-body ones, in that the inside-your-body one is really a black box for the non-medical person, right? You’re simply taking this drug and you don’t specifically understand the mechanism. For me, it was quite a different decision to go on a drug trial and it does just have to do with the… kind of the unknown effects on your body, perhaps (female, accepted). Barriers and facilitators to information seeking. Many patients found the ease of the information-seeking process to be dependent on the “newness” of the experimental intervention and the subsequent availability of information. For example, patients were more limited in their search for information about experimental drugs in the first phases of testing because the human data was non-existent. On the other hand, if the trial intervention had been used previously to treat other cancer types or patient populations, or if it had already been subject to Phase I or II trials, then there was usually some information available to inform patients’ decisions: I started out with news links about the study, because news links, you know New York Times, various papers had talked about the study. And then of course they would reference other sites as well, for more information such and such, you know how you hop around doing searches. And then after that I just brought up Metformin in itself and just started reading about it generally, trying to get a sense of what it was about. It was fairly easy to find information about the study. I guess because with three countries participating, there is a lot out there about it (female, accepted). Experience of receiving information. The experience of receiving CT information varied across the cancer trajectory. Patients particularly struggled with CT decisions that 151 occurred around diagnosis. Fear and the emotional adjustment to the cancer diagnosis were identified as barriers to understanding information because it blocked patients’ ability to process information and appreciate how it related to their situations. In addition, there was often a sense of urgency surrounding the decision because patients did not want the cancer to advance. When emotions combined with the pressures of having to make a quick decision so as not to lose ground on fighting their cancer, patients expressed a sense of being completely overwhelmed and destabilized during the time when they were supposed to make a CT decision. One patient with breast cancer discussed her previous experience of declining a chemotherapy trial: It’s a bit overwhelming …first of all, dealing with diagnosis, which is pretty traumatic, and then all the information around course of treatment and taking all that in. And then to be given an option for a study. That’s just another whole kind of…it’s all happening at once. And you’ve got to make the decision fairly quickly because you want to start treatment very quickly and then in my case, this time around, I started as quickly as possible. So yeah, it’s a lot of information to take in on that first meeting with the oncologist and so you’re dealing with all these issues and then there’s always that study that’s presented. And so it’s just more information to have to deal with (female, declined). As a result, some patients appeared to limit their search for information because they were overwhelmed with information and just wanted to “get on with it.” As shared by a patient who received a drug trial offer shortly after being diagnosed with breast cancer: I had already been inundated with stuff from going through my surgery and everything and the biopsies…it was already so much stuff coming that by the time I got to this it was like, ‘Yeah okay, just sign me up’ (breast, accepted). Having had cancer before, being educated (some university or college level education), or having worked in a profession that required assimilation of complex information and navigation of risk probabilities may have helped some patients to manage their fears and consider CT information within their current context. Those whose cancer was 152 recurrent had some familiarity with being diagnosed and reported being able to simultaneously process information related to their diagnosis and the trial option. Patients who worked in healthcare, or had jobs that involved mathematics or science, also appeared to be at an advantage to understanding risk information, which was often expressed in statistics. For most patients, consistent with the “No Wo/Man is an Island” model, having support persons available to also ask questions and take notes during the trial offer and review online literature to filter useful information was a facilitator of patients’ information-seeking processes, especially for those patients who did not have previous cancer experience or professional expertise. As one husband who was a scientist stated, his role was to help his wife “figure out what she wanted and to make sure that she knew that whatever she chose was okay with me.” This man supported his wife’s relational autonomy by not telling her what she should do but rather by helping her seek appropriate information that was relevant to her goals of treatment: I think to give as much information as possible and easy sources of information for further understanding for those who what that would be helpful. I happen to be able to surf the web and have a scientific background so I could find the studies and we have people in the family [who could review the studies] (male spouse, supported female patient who accepted a CT offer). Lastly, some patients expressed feeling “sheltered” from certain information by health professionals or not receiving the information they felt was important to their decision making. This was perceived to be a result of oncologists withholding or not elaborating on information related to severe adverse effects, including mortality rates, because of the concern that such knowledge might cause patients distress. A few patients were sensitive to CT personnel’s struggle to provide the necessary information while minimizing their fears, which was supported by the perspective of a few prostate cancer patients who did not want to 153 know certain information about their life expectancy, for example. However, other patients expressed a desire to be provided this information because they felt it was important to their decision making. As discussed by one patient: Now I do believe at the time I got some information on the drug and its side effects and [the oncologist] certainly pointed out it was a Phase II drug experimental program. But he didn’t say how many patients failed and have since died or anything like that. That would have been the information I would have liked to have known but that may have scared me away, so… (male, accepted). The above demonstrates the interesting ethical dilemma of CT personnel not wanting to share information that will cause patients to experience additional anxiety or fear but still needing to support patients in making fully informed decisions. This issue will be discussed more fully in the final discussion chapter (see Chapter Six). Establishing Credibility Most patients engaged in a process of establishing credibility or trustworthiness as they sought information about a CT. This was largely a social process that required input from others. Patients attempted to determine the credibility of not only the oncologist presenting the trial, but also the institution hosting the trial and the drug or company sponsor. There was a circular relationship between the process of seeking information and establishing credibility. In order to establish the credibility of an information source, such as a health professional or a clinical institution, patients had to first seek information about their reputations. Was the oncologist a leading scientist in the treatment of their cancer type? Was the institution a world-renowned centre for cancer care? In this way, patients’ relational autonomy extended beyond the patient-physician relationship into the organizational and structural context. Once credibility was established, patients described feeling more confident in seeking further information about the CT from these sources. 154 Establishing the credibility of their oncologist, who patients believed also had a role in recommending trials, was especially important to patients since these physicians were their point of care for cancer treatment and patients relied upon them for psychosocial support. Patients established credibility of their oncologists in several ways. Some patients turned to a physician whom they already trusted (often their GP) to recommend a reputable oncologist. The GP would “bridge the credibility” by identifying the oncologist as a leading expert in the field of research and treatment of the patient’s type of cancer. Credibility provided patients with the foundation for the development of trust in their oncologists. Some oncologists, in turn, earned patients’ trust by investing time and energy into building a caring relationship with the patient and family, and going above and beyond in their care. One patient described how her oncologist, whom she believed to be credible and well-respected in her field, also treated her as an individual and made her feel cared for and well looked after, thus enabling a trusting relationship: She was very open, she was so warm and so confident and I liked the fact that she was straightforward, I could ask her anything, she called me late at night - it surprised the heck out of me! I’m in bed and I get this phone call…it was actually her calling from home, so stuff like that. She always asked about my kids and so it wasn’t just…I wasn’t just another number or just another person, and I know how busy she is, so I hate to take up any more of her time, but I appreciated that relationship with her. And knowing who she is, that to me meant a lot too. That I could trust her (female, accepted). Trust was important because it made patients feel safe and that their best interests were being looked after. Patients who trusted their physicians were more likely to seek their physician’s opinion and recommendation regarding CTs. Only a few patients were able to trust their oncologists without any history of relationship or recommendation from another physician; however, these patients often had a medical professional in the family and were 155 thus “more open to the concept of trust my doctors and off we go, kind of thing” (female, accepted). As another cancer patient with breast cancer stated: I trust my healthcare professionals. I’ve been lucky enough to have very good people treating me over the years and so if you start with that level of trust, then if it’s something they suggest, obviously that carries a lot of weight. And I followed pretty much all of the suggested treatments that I’ve been offered and that’s again because of that level of trust. These people are the experts and they know what’s best for me. I mean I don’t follow everything unquestionably but that certainly carries significant weight with me and so the fact that [name of oncologist] thought this was an important study – it was the same thing, it carried a lot of weight with me (female, accepted). For some patients, establishing trust in their oncologist was an “essential” component of their CT decision-making process as it allowed them to accept their oncologist’s opinion about the trial and bypass some of their own decision-making processes. These patients are called “snap” decision-makers in the “No Wo/man is an Island” model because they privileged the doctor’s knowledge and thus spent less time specifically on researching information and weighing pros and cons. Some snap decision makers appeared to blindly trust the oncologist to know what was best for them (often their trust was based primarily in oncologists’ social authority and professional status), while others had spent more time establishing the trustworthiness (and credibility) of their oncologist through their social networks and structural resources. For example, one male patient asked his cousin, who was a researcher at the cancer centre, to confirm the credibility of his oncologist. Once the patient had his cousin’s confirmation, he proceeded to trust his oncologist and made a snap decision to participate in a CT. In general, patients recognized their oncologists as specialists who had far greater knowledge, skill and expertise regarding cancer treatment decisions than they could ever have. They also assumed their oncologist would carefully consider their case and would not 156 recommend something that would harm them or not be worth their while. As expressed by a patient with prostate cancer: I had the information that I wanted and that information sort of manifests itself in a trust of the people that have been treating me at the cancer agency, so I already have a fair amount of information on those people and I trust them, so that work is done. I know that, as specialists, if they think it’s a worthwhile proposition to test a drug on me, I don’t for one minute think that they are doing that or make these decisions not without thinking about them fairly carefully and coming, first of all, to the decision that it’s the right thing to do. I am fairly sure that if they know of something out there that would not be effective in my case, then they wouldn’t offer me the treatment. I am fairly confident that the clinical people look at my case and look at me as an individual or my disease as an individual phenomenon and decide what is worthwhile trying and what is not (male, accepted). The process of establishing credibility of the institution hosting trials also involved seeking the opinions of others. Patients sought opinions and reassurance from family members, the media, or GPs that the cancer agency was a centre of excellence or a “world-renowned” research institution. One patient discussed how he consulted with his family, who had previous experiences of being treated for cancer at the centre: I actually spoke to my mother about it, and my brother. My brother actually went through the prostate clinic. And that’s how I ended up here, actually, is just through his coaching. And then I spoke to my mother about her experiences with breast cancer, her clinical trials. […] And my brother has got a fairly high regard for the clinic itself and sort of coached that if they’re conducting trials then it’s probably … it’s probably state of the art kind of treatment, I guess (male, accepted). With regards to establishing the credibility of specific trials, some patients wanted more information about the study, including whether the trial was multi-site, how many subjects had been enrolled in the study to date and whether the intervention had shown previous success in other disease sites or cancer populations. If a drug had already been approved for use in another country and was in a later phase of testing in Canada, for example, this satisfied patients that it was credible and relatively safe. It's been around for 40 years, so it's not something brand new that hasn't been 157 adequately tested in clinical trials for the other purposes for which it's used. So, presumably they've figured out all the negative problems that it can have otherwise it's just generally in your system. So it's not like you're taking something brand new (female, accepted). Weighing Pros and Cons A variety of sub-processes made up the complex weighing pros and cons process, depending on what concepts/themes patients prioritized as important considerations of trial participation. These sub-processes included doing what is right for me/considering altruism, weighing adverse effects and benefits, grappling with uncertainty, considering logistics and maintaining relationships. These sub-processes often overlapped, with patients engaging in many of them at any given point in time. Whether patients considered the concept as being a pro or con appeared to be induced by influencing factors, predisposing beliefs and/or social networks. Doing right for me/Considering altruism. Many cancer patients considered whether the trial would benefit them or benefit others - doing what is right for me or considering altruism. To what extent patients were internally (self) or externally (other) oriented depended on a variety of factors, such as their stage of disease or pre-disposing beliefs and attitudes about research. For example, patients with more advanced stage of disease in this study appeared to be more internally focused on their own health since they are facing their mortality. Doing what is right for me. The concept, doing what is right for me, involved assessing one’s needs and prioritizing the self. Prioritizing the self meant thinking of one’s own health status and emotional needs first, instead of others’ needs (e.g., social good), in order to fight the cancer. Typically, patients who prioritized themselves were aggressively seeking treatment for advanced or recurrent illness or they had just been diagnosed with 158 cancer and were focused on finding a curative treatment. Often, doing what is right for me was equated with a hope for a cure. One patient shared the following: I was motivated by the fact that I thought that this could really help me. That's what I was motivated by more than anything else. There really wasn't even a secondary thought in my mind. I thought this might be a hope for me (female, accepted). A CT nurse confirmed that this self-focused orientation was not necessarily “selfish” but was a result of wanting to extend life. It most frequently occurred when patients were scared, had severe disease and when they had limited remaining treatment options. These patients were perceived as prioritizing themselves as they needed to find a treatment that would work for them because they wanted to live; however, being selfish could also mean refusing a trial because the burden was too great. As explained by a CT nurse: Often they’re facing large risks. They have either huge cancer and that is scary and has potential for coming back very fast and so they are looking at aggressive treatments, or they metastasized and there’s nothing left. They’ve got nothing else to try, they tried everything and like then, they’re at their wits end, they have nothing else. So those are the things that motivate people to think of clinical trials. Many patients considered what was best for them in the context of their family. Patients wanted to be around for their loved ones. Some felt they were also needed to support raising children or to provide financial assistance. In turn, the patient’s family was invested in helping their loved ones find the best care available to extend their lives. Thus, doing what is right for me could be interpreted through a relational autonomy lens as doing what is right for the family. As one patient stated: “Well, it's because of my family that I chose to do this. It was to give me the hope of a little bit of life extension” (male, accepted). Some spouses of prostate cancer patients went so far as to make the decision for their husbands or strongly encouraged the patient to participate in the study so that he would receive the best medical care. Thus, choosing a trial to receive the best care for oneself was 159 sometimes the responsibility of a family member. The tendency to strongly influence CT decisions was perceived by a CT nurse to be gender specific: I think certainly for the men with prostate cancer, their wives and family and caregivers, the primary caregiver, has a huge impact on that decision. More often than not—this is my bias—the wives tend to push the husbands into doing clinical trials and kind of actually help make that decision for them. More often than not, they actually do push I think for the patient to go on study. I used to do breast cancer a long time ago and I found it was different for a woman. The woman made the decision and men were usually not there and usually did not make any comment. Considering altruism. The concept of considering altruism fits well with the “No Wo/Man is an Island” theoretical model, since it is not just about serving the interests of the individual patient but also the good of society that is important in the CT decision. These cancer patients were primarily motivated to consider trial participation because it might help future cancer patients and contribute to the scientific advancement of cancer treatment. They held the belief that they would not currently be benefiting from drug therapy if patients before them had not volunteered for research. A few expressed a sense of obligation about “doing their duty” (breast, accepted), perceiving trial participation to be a “good thing to do” and a way to give back to others and advance science. As one patient with prostate cancer who had participated in many oncology trials stated: The only reason I went on it [CT] was 1): I don't believe people don't go on it. It's kinda their obligation. If they have cancer and they're benefiting from experiments done before they got on a drug, then they should carry it forward and help someone down the line when their time comes. So that was the main thing I considered. Other than that, there's nothing else to say, you know? I just thought that it should be done. It's a good thing to do (prostate, accepted). Patients who reported their cancer as localized, stable, or in remission were more often motivated to consider trial participation for altruistic reasons. These patients were typically less in a state of crisis with regards to their disease stage and illness management. They also believed strongly in the value of research. Patients who reported they had 160 advanced cancer but were not aggressively seeking further treatment were also more likely to consider trial participation for altruistic reasons. These patients were generally older and less concerned about outcomes or achieving their goals because of their later stage in life. As one patient reflected: When you're 88½ years old, almost 89, you've had enough of this life, so it doesn't matter if it kills you. I was very successful in my life. I've had a good job, I got a house that's inflated by $100 a day in [name of town] and my kids will be well off, so what the hell (prostate, accepted). Education level also appeared to play a role in patients’ ability to understand the purpose of research and to identify with the larger goal of helping others benefit from advancements in science. Those patients with university/college-level education and higher, tended to have an easier time understanding the role of research in informing standard treatment and that ultimately someone needs to volunteer to participate in research in order for care to advance. The following quote highlights how an educated woman with breast cancer was able to reflect on the purpose of CTs: On the other hand, I figure that, with the treatments that I've had, I've benefited from a whole bunch of work that's gone on before. And I know that the only way that the doctors and scientists can properly assess a given treatment, is to do the randomized double blind medication, placebo or whatever they're testing to see if, in fact, it does make a difference. From a scientific standpoint I understand the concept and why they do it, the way they do it and why it takes time (breast, accepted). Personal benefit and altruism were not mutually exclusive considerations. The majority of patients who were motivated to help themselves were also concerned about helping others, and vice versa. Patients often talked about framing these as primary and secondary motivations: Well, very selfish initially. But then after you're chosen you think, ‘Oh yeah, and plus I'm doing something for the world of research’ (prostate, accepted). 161 A few patients stated that they experienced a moral struggle between doing what was best for them and what was best for society. This moral struggle was most pronounced in female cancer patients who expressed distress and guilt when they prioritized their needs and declined a CT: Oh gosh. I just remember having such a tough time…There’s a huge part of me that wants to do the right thing and contribute to science and all that stuff. So my natural tendency is to go for these kinds of things. [But] there was an aspect of it that was kind of throwing my lot into sort of that randomness of…so that was drawing me away from the study and that’s in the end what made me not participate was the randomness…the big possibility that I was going to be, you know, kind of at the toss of a coin as to whether I had chemo or not (breast, declined). However, it is unclear to what extent their distress was the result of an internal moral struggle versus a social imperative stemming from gendered norms and expectations. Social relationships had an impact on these female patients’ experiences since the way information was framed by CT personnel appeared to impact their levels of distress or how they perceived CT information. Weighing risks and benefits. For most patients, weighing the risks and benefits of a CT was extremely salient to their decision-making process. Of paramount concern were physiological risks. These included the risk of disease progression if patients decided to stop their current treatment in order to participate in a trial or potential adverse effects of the study drug, especially organ failure, blindness and chance of death. These effects were considered by both those who accepted and declined CTs: Well, the key thing is the side effects and potential side effects of the drug. They’re not particularly nice drugs so … the probability of the down sides, or side effects, I guess tried to weight that. That was probably the biggest thing (male, accepted). The most important thing was that there would be no side effects regardless of their statistical significance. As an engineer and so on, I know statistics very well and I know that they have their uses for selling and they have their misuses as well. I just have no interest in statistics that involve side effects. Serious ones. I'm not talking 162 about, you know, headache, diarrhea, that kind of thing. I'm talking about where you could have kidney problems... (male, declined). Patients considered the probability and magnitude of adverse effects and to what extent their quality of life would be affected. This was very much a personal and individual judgement but often had relational implications. For example, a few patients had planned to travel with their spouses after cancer treatment. They did not want to participate in a trial that could cause them to have adverse effects because it could interfere with their travel plans. Also, fatigue was a significant outcome for some patients because it meant not being able to engage in activities that were important to them. As explained by a prostate cancer patient who initially was closed to a CT and declined participation because of adverse effects but later retracted his decision after talking with his oncologist about fatigue: He [the oncologist] told me that I'd have no problems but fatigue. And I don't want fatigue and I have a hobby of collecting and I spend 10 to 12 hours a day on my computer, sortin’ those entries and I don't want any interference. So he guaranteed me that I would only have a little bit of fatigue (male, declined/accepted). Perceived benefits of CTs that patients described as being influential to their decisions included improved access to new and state-of-the-art therapy that could potentially be effective for treating their cancer. This was particularly beneficial because of the current restrictions to new drugs in the healthcare system that made CTs sometimes the only way to access a therapy. A positive by-product of trial participation that many patients also commented on was access to additional nursing support, follow-up care and tests to monitor their physical health. Many patients appreciated the idea of continuous health monitoring while on a trial and saw this as helping them cope with the fear or anxiety they had about cancer spread or recurrence, particularly once standard treatment ended. As mentioned by one breast cancer patient: 163 To me, I find it a reassuring thing to participate [in CTs] because it’s getting me probably more oversight, somewhat more oversight than I’d otherwise get in terms of my breasts to make sure – because obviously a concern I have is that it will recur. So if I’m getting looked at more frequently as a part of participating in the study, I see that as a good thing. And continuing to get more attention to my condition. I find that comforting (female, accepted). Follow-up care by nursing staff and monitoring was especially important for some patients who had previous experience with care that was fragmented. Returning to the “No Wo/man is an Island” metaphor, patients shared stories of being primarily an “island” or left to themselves in their experience of health care. Having to see several doctors and residents within one appointment, wait several hours for blood work to be completed and travel to different sites to receive test results often made them feel alone and unimportant. The lack of coordinated care experienced by some patients led them to view CTs as offering coordinated and comprehensive healthcare that they could not get outside of a trial. One patient explained this perceived benefit in the following manner: So navigating the system is really challenging. And being part of a trial, being looped in to a nurse who navigates that system for you is awesome. It’s really awesome... also knowing and understanding that being part of a trial, you have a stronger link with the system because of the study nurse and just being able to have access to queries and questions (female, accepted). The benefit of being closely monitored and receiving additional follow-up was particularly important to patients who lacked family support or did not have family living nearby. In other words, CTs provided the social support of being at the clinic and in frequent contact with nurses. As one breast cancer patient shared: It was that I was going to get probably the best treatment that I could get, meaning probably a few more tests that would probably take a little more time for me, but to me it was like there was nothing that was going to missed, it wasn’t like oh well maybe we should have done this, I knew I was going to be followed through the whole procedure step by step, and that was really important to me because I felt I don’t have any family here, so it was more important that I had all the treatment I could get and know that I was being looked after (female, accepted). 164 Perceptions about the efficiency of the healthcare system also played a role when patients, especially those newly diagnosed, considered the risks and benefits of CT participation. One patient explained how she thought participating in a CT would address her concerns of long waiting lists and delays in receiving treatment. When I found out I had cancer, of course that was the first thing, ‘Oh my God, am I going to be on this huge waiting list? How long is this going to take?’ and this thing is growing and the trial was going to start right away, so that did influence me. Because I was going to get my treatment right away, it was going to start right away. I don’t know if it would have been right away without the trial, but in my mind I thought the trial would be my best bet of getting the best medical care. (female, accepted) It is important to note that in describing her experience, this patient appeared to equate receiving a trial intervention with treatment for her cancer, thus also raising questions about her level of understanding the difference between standard care and research. In some instances, patients also considered the benefit of satisfying a sense of personal responsibility to be doing absolutely everything to treat their cancer. This would allow patients to avoid decisional regret. As one patient elaborated: Again I think a couple of things, one was that I didn’t want to have to look back on my decision and wish I had done this, because that would have been the easier way of doing it. I knew that the clinical trial may be a little more time consuming and I didn’t know another option that I…like I knew there was an option, yes, I could do it, or no, I couldn’t, but I couldn’t imagine not doing it because it was going to give me, again the best care…just getting all the tests and having everything sort of put through quickly… (female, accepted). In weighing risks and benefits, some patients whose current treatment was not working considered a CT because it would give them access to a novel and potentially life-saving medication, even if there was a chance they would experience adverse effects. These patients downplayed the potential side effects of participating in a CT by comparing them to the adverse effects associated with any chemotherapy treatment, thereby interpreting those 165 from the study as “not a big deal” or explaining how it is “part and parcel” of CT participation. The following patient was aware of the risk of placebo but focused more on the adverse effects of receiving the study drug: If I did do this study, I had a 66% chance of accessing this drug, ‘cause it was two-thirds, one-third for this study that I’m on now. One-third placebo, two-thirds the drug. So I was, yeah, I really wanted to have this medication, again, because it’s my nature to go for it and do everything that I can do to stop this cancer. And so it was my only way to have access so this medication was to participate in this study. So I didn’t see any downside, apart from that, you know, small chance of having a significant reaction to the drug, but that was just a chance you’ve got to take (breast, accepted). For some patients who were undecided about participating in a trial because of concerns about adverse effects, being informed by their oncologist or CT personnel that they could withdraw at any time was an important consideration. Other patients were aware that a trial had to be stopped if significant adverse effects were identified during the study, as discussed by one patient with breast cancer: I guess maybe the side effects, or like whether you’re taking something that you later wish you hadn’t. I feel quite confident that… that’s not a big concern for me, in that I feel very confident that if there were any negative side effects, that these would be identified early and we would be all asked to stop doing it [the study] (female, accepted). Clarifying the potential risks and benefits of participating in a CT was a process that for some patients occurred over an extended period of time. These patients returned to the information-seeking process, consulting with their oncologist or CT personnel about the potential risks and benefits. For example, the following patient identified delay in treatment as a concern for him but after talking with his oncologist, he was persuaded that the risk to his health was not significant: I basically weighed the pros and cons, the upside and downside. When I was talking to Dr. X and he was explaining what it was … I was a bit concerned because the operation was delayed for four months, that was potentially a downside. And so we 166 discussed that and he explained, basically, the theory and the mechanism, what was going to happen. So when I was first speaking with him…decided there wasn’t much of a downside” (male, accepted). Grappling with uncertainty. Unique to CT decision making (as opposed to treatment decision making) was patients’ heightened sense of uncertainty associated with trials and the influence this had on their decision-making process. There were many unknowns associated with these studies, which was precisely why trials were needed – to answer outstanding questions about an intervention’s safety and efficacy. Moreover, in order for trials to be conducted ethically in Canada, clinical equipoise must also be met (CIHR 2010). That is, the community of oncologists must be in honest disagreement about whether or not the trial intervention will be helpful, harmful, or no different than alternative treatment options already available. From the patients’ perspective, the fact that so much was unknown presented a challenge to their decision-making. Answers to questions that were important to patients were difficult to arrive at with certainty. For example, which adverse effects, if any, will the patient experience? What is the likelihood of benefit from the intervention? If it is a placebo controlled trial, will the participant receive the active arm or placebo? If the patient forestalls standard treatment to be in the trial and receives the placebo, what is the chance the cancer will progress or recur? Patients’ uncertainty about an intervention’s effectiveness and safety was more pronounced for earlier phase trials because so much more was unknown about the intervention and outcomes at this stage of testing. There was also little to no expected benefit to participants, including a risk of receiving no intervention. Some patients managed this uncertainty by seeking information and establishing credibility of the study drug. For example, if the trial had been conducted in other countries and had shown promise in earlier 167 phases or for other diseases, then this assuaged some patients’ anxieties about the unknown efficacy of the intervention or fear of adverse effects. One breast cancer patient described how her familiarity with a study drug for treating diseases other than cancer allayed some of her fears about unknown side effects: And with this Metformin study that I’m going on, for example, I was quite comforted to know that it is a well-established drug from other fields of application, so there’s quite a long history of its application. If this was a drug that only had two years of people taking it, then I would be a little bit more concerned about unknown consequences twenty years from now (female, accepted). Patients grappled with uncertainty and ultimately managed it in one of two ways: either deciding to stay with the tried and true treatment, thus declining to take the risk of the unknown associated with CTs, or rationalizing they had nothing to lose if they accepted the offer to participate in a CT, thus facing uncertainty head-on. Patients’ perceptions of illness severity and their previous or current healthcare experience (e.g., if patients were already taking a known effective treatment) influenced whether patients were willing to take the “gamble” and try something they really did not know much about. Patients who were motivated to stay with the tried and true treatment shared that they felt that standard care was the most effective course for them because they believed their cancer was, for the most part, under control. As a result, they were less willing to take a risk where they could experience unpleasant side effects in the face of uncertain benefit. As described by one patient: That was basically my thought process was going through what I could expect from the different choices. And again, deciding to go with the tried and true rather than take chances on something that might give me horrible side effects that I wasn't quite…my argument had always been it was a matter of whether I was going to go through with any treatment at all was the quality of life that the treatment would allow me. If I was going to spend six months flat on my back sick to have an extra six weeks of misery at the other end then I would have said, ‘Thanks, but no thanks’ (female, declined, emphasis added). Another patient described how her understanding of the severity of her illness influenced her 168 decision to continue with standard care: Because at that time, I did not have metastasis and so I would look at what was the best treatment options and what was the best success rate, and I would rather go with something that I knew than something that I didn’t know (female, declined). On the other hand, patients who reported they had late stage disease and fewer treatment options perceived they had nothing to lose by venturing into the unknown. In these cases, patients downplayed the risks of a CT and hoped the extra tests would at least lead to early detection of cancer progression that would help inform their oncologist’s treatment plan outside of the trial. These patients framed uncertainty as a potential benefit since there was also a chance the intervention could be helpful. As both a breast and prostate cancer patient explained: And so I would probably have wanted to be more ill to participate in that trial than me currently feeling perfectly healthy. I probably would have been less willing to go on one where I didn’t have as good a view of what the likelihood of an unknown effect on me was. So I imagine, in a way, the more ill you are, the more embracing you are of the unknown, because there’s also the unknown upside too (female, accepted). Of course, you know what we get into but then you have to pay a price. Paying a price is getting into the unknown, getting into uncertainty. And there’s no guarantee. But again, the final, I think the rationale is no venture, no gain. If you don’t get into it, then you don’t know what happens (male, accepted). Considering logistics. The majority of patients considered the logistics of trial participation as part of their decision-making process. This sub-process also relates back to the core construct in the sense that patients are whole persons. They have lives and responsibilities that they need to consider in addition to their disease and whether the trial is the best option for them. Patients reflected upon the amount of time that would be required for them to attend the trial, how many visits they would need to make to the cancer agency and how long the trial would last. Driving to attend clinic appointments was a major concern for many patients who were considering trial participation as well as whether they could 169 afford the cost of parking and fuel, which could be significant, especially for longer trials that required frequent visits to the agency. Even those patients taking advantage of volunteer driving services had to consider the associated challenges: Initially, it was quite a commitment because I had to come in every week and then… Yeah, I think it was every week or every two weeks initially. And so that was time consuming. So, it was a matter of arranging rides and volunteers to help me come in, because at that time I was still quite weak, I wasn’t coming in on my own. And so that was certainly a concern, but felt that once we go through that then just coming in once a month was manageable (female, accepted). Most patients relied on family or other support persons to drive them to the clinic. This was particularly the case for patients who lived outside the city centre. From the perspective of these patients, much coordination with their social network was required in order for them to participate in CTs. They needed to request certain appointment times to fit with their driving schedule. They also needed to deliberate about how best to manage their fatigue after a long day of travel and appointments. For some out of town patients, they had to rely on family members or friends to provide a place to stay over night, if required: Because we come in from [name of city], I had requested morning appointments for the fact that I come and stay at my sister’s place because from her place it’s a half hour drive. From [name of city], it’s a two-hour drive. It’s a lot more relaxing for me as well too because if we have to drive in for two hours, you get here, you’re tired. Then you see the doctor and you get your prescription, whatever, then you turn around and drive back home again. I mean that’s 200 kilometres in one day. So I mean that’s a tiring day. And so that’s why I made the arrangements with my sister to go and stay the night before (male, accepted). Patients were concerned about the time that they would need to dedicate in order to participate in a CT and whether they would be able to fit it into their daily lives. Many needed to negotiate flexible work hours and childcare in order to participate. Particular challenges existed for those younger patients who were still working and had children at home that required care. As explained by one patient with breast cancer: 170 It was not close to home for me, so travelling back and forth to downtown was a lot… it was just the time and making sure I had child care because a lot of it couldn’t just be a short appointment. Even though I would be here in the morning and my kid would go to school, but he would be home at noon, so I always had to make sure I had somebody covering that. So that to me, that was the only real issue, was the travelling back and forth (female, accepted). Those who worked for employers that allowed some flexibility regarding work hours were more able to commit to trial participation. Other patients were able to go on disability or were retired, so balancing work with CT participation was not an issue. As one patient explained: I don’t work anymore, so maybe you should make a note of that. It would be a lot harder if I still had kids at home, in school or I was still working full-time. I used to practice as a lawyer. So I don’t have as many demands on my time. I do some not-for-profit board work and that’s about it. And so, that’s probably another big factor in my participation, is the fact that I do have more free time than the average working person or working parent out there. So I thought I don’t really see why I shouldn’t participate (female, accepted). Although some patients might have been able to negotiate time off work, there still was the issue of lost wages if they needed to be absent for long periods: I’m very lucky, because I’m self-employed – I run two businesses – so those businesses were capable of running on their own if I needed to be away. I was away from work for quite a bit during the chemo, especially. The radiation not so much – it was very easy compared to chemo. But yeah, they could run quite well without me, so I wasn’t… that wasn’t a concern. And I didn’t have to worry about money. I imagine for some people that’s a huge, huge thing (female, accepted). In addition to the above logistical issues, some patients described feeling overwhelmed just managing their standard treatment regimes. Further coordination of a CT would have meant more visits to the clinic and even more coordination with their social network. Some patients did not have the economic and social means to coordinate visits to the cancer centre for trial participation, thus raising the issue of inequitable distribution of resources across patient groups. Even though the study intervention might be beneficial, one 171 breast cancer patient explained that she felt the chances she had of actually receiving the active arm was not worth the added inconvenience because coordinating her participation in the trial would be difficult: I had heard positive things about taking the bisphosphonates but I only had a 1 in 3 chance of actually landing that in the clinical trial and I thought for all the pain and effort… the inconvenience of going in for all these extra tests and things. I was having a hard enough time going in for the tests I needed to just follow up on the chemo. And then parking and, in some cases, having to get people to drive me. I had my plate full already with just getting to my chemo treatments. It was just too logistically demanding and difficult under the circumstances I was in at the time (female, declined). A few patients spoke about prioritizing other activities that they perceived as being more important to their quality of life than visiting the cancer agency for a trial: No, I was asked if I wanted to participate in the sports one, which I would have liked to have done, but you have to go I think two or three times a week in town and during your chemo for it and I thought since that got into the summer – actually it was longer than May, June, it was May, June, July, August. What am I thinking? It just got – life into 3-week chunks at that point. We have a place [outside of town]. Psychologically, I really wanted to be able to escape to there between my treatments. So if I had participated in the sports study, I wouldn’t have been able to have gotten out of town for the summer, which was really important to me (female, declined). Patients identified ways in which CT personnel addressed logistical challenges to their participation in a trial. If the trial tests and follow-up were incorporated into already scheduled clinic appointments for standard care, then this would lessen the additional logistical burden of CTs. Also, if CT personnel scheduled appointments for when it was most convenient for the participant to attend, thus working around the patient’s schedule, this also facilitated patients’ participation. As described by one patient: I asked about how it would impact my time and would I have to do extra things and would I have to be at the clinic more often…all the logistics of it. I didn’t have to do anything different. I had more blood tests done, I had more blood taken and that sort of things and that’s what she had told me initially, but she said that would happen on days that you have to be here anyway. So that part of it was very convenient for my life (female, accepted). 172 Logistics, however, were balanced against perceived need and limited treatment options, especially among those with more advanced disease. For example, patients who reported they had late stage cancer were adamant that just because they lived far away from a study site, CT personnel and oncologists should not assume that they would not want to participate. As one patient explained, advanced illness creates pressure to make logistics work - the cost and inconvenience of travel is minimal compared to the possibility of death from advanced illness: It is a serious business, terminal illness, and if I think that I could have advanced my position, or in fact retreated from terminal illness one inch, by making several trips across town and having several bone scans, I will do it. It is an easy decision to make (male, accepted). Another patient, who discovered her oncologist had not offered her a trial because of assumptions made about her ability to travel to attend appointments, was very concerned that she was not initially presented with the option. The following quote highlights the potential for oncologists’ assumptions to result in inequitable access to CTs and potentially damage the trust relationship with patients. Furthermore, as the following quote demonstrates, it is a mistake to assume all patients are only “islands” unto themselves, and not integrated with a socio-political system that can potentially facilitate their access to CTs: So I did all the background information myself and when I told her about it, she was like, ‘Well, it's in [major city].’ I'm like, ‘Yeah, I know. So? You should at least have offered it to me.’ You don't know what my financial situation is. You don't know if I have family there. You don't know anything. I do have a lot of support. I used to live in [major city], I lived here for six years so I have lots of friends. I am lucky that way. But she didn't know that when she didn
UBC Theses and Dissertations
Cancer patient decision-making and relational autonomy related to clinical trial participation Bell, Jennifer Ann Harrison 2014
Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.
- 24-ubc_2014_spring_bell_jennifer.pdf [ 39.64MB ]
- JSON: 24-1.0166923.json
- JSON-LD: 24-1.0166923-ld.json
- RDF/XML (Pretty): 24-1.0166923-rdf.xml
- RDF/JSON: 24-1.0166923-rdf.json
- Turtle: 24-1.0166923-turtle.txt
- N-Triples: 24-1.0166923-rdf-ntriples.txt
- Original Record: 24-1.0166923-source.json
- Full Text