Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The idealization of dying at home : the social context of home-based palliative caregiving Stajduhar, Kelli Isabel 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2001-71540X.pdf [ 20.17MB ]
JSON: 831-1.0090464.json
JSON-LD: 831-1.0090464-ld.json
RDF/XML (Pretty): 831-1.0090464-rdf.xml
RDF/JSON: 831-1.0090464-rdf.json
Turtle: 831-1.0090464-turtle.txt
N-Triples: 831-1.0090464-rdf-ntriples.txt
Original Record: 831-1.0090464-source.json
Full Text

Full Text

THE IDEALIZATION OF DYING AT HOME: THE SOCIAL CONTEXT OF HOME-BASED PALLIATIVE CAREGIVING by Kelli Isabel Stajduhar R.N., Misericordia General Hospital (Winnipeg), 1986 B.S.N., University of Victoria, 1990 M.S.N., University of British Columbia, 1995 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES School of Nursing We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA September 2001 © Kelli Isabel Stajduhar, 2001  In presenting this thesis i n p a r t i a l fulfilment of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It i s understood that copying or publication of this thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission.  The University of B r i t i s h Columbia Vancouver, Canada  II  ABSTRACT THE IDEALIZATION OF DYING AT HOME: THE SOCIAL CONTEXT OF HOME-BASED PALLIATIVE CAREGIVING The impetus to provide home-based palliative care is growing in an effort to curtail health care costs and to provide a more therapeutic milieu for dying and death to occur. Despite an expanding body of research focusing on the provision of palliative care at home, there has been little attention given to the historical, social, economic, political, and ideological structures and processes that play a role in shaping family caregivers' experiences with providing home-based palliative care. This ethnographic study approached this problem by examining the social context of home-based palliative caregiving, with the overall purpose of uncovering how larger structures and processes influence family caregiver experiences with providing home-based palliative care. The investigator conducted 130 hours of field work in the homes of 13 palliative caregivers (and 12 patients), and interviewed 13 active caregivers, 47 previous caregivers, 28 health care providers, nine health care administrators, and one publicly known lobbyist for the assisted suicide movement. Through this research, home-based palliative caregiving was revealed as more complicated than typically represented. Under ideal conditions, home-based palliative caregiving would result in life-enriching experiences, marked by negotiated decision making, careful consideration of caregivers' needs, respectful interactions with health care providers, and equitable access to health care. However, a number of factors stemming from idealized assumptions about dying at home and the contexts of health care provision resulted in some caregivers feeling "pressured" to provide home care, and consequently, left some of them feeling that their obligations  to care were exploited by the health care system. Recent shifts toward moving the provision of health care closer to home not only changed some of the caregivers in profound ways but also transformed the home setting into one that was infused with the influences of biomedicine. Long-standing philosophical divisions between mainstream medical practice and the ideals upheld by palliative care have inhibited the development of coordinated and consistent systems of palliative care that could ultimately benefit both patients at the end-of-life and their caregivers. The findings of this study demonstrate a need to re-think the ways in which home care for the dying is conceptualized, pointing to a need to foster critical consciousness in order to explore how taken-for-granted assumptions influence and sometimes dominate the provision of health care at home. Such awareness will ultimately pave the way toward a health care system and society that genuinely respects and honours the obligations, choices, and needs of family members who provide palliative care at home.  iv TABLE OF CONTENTS Page ABSTRACT  ii  TABLE OF CONTENTS  iv  LIST OF TABLES  viii  ACKNOWLEDGEMENTS  ix  DEDICATION CHAPTER ONE: Introduction  x  Situating the Study in a Historical Context The Emergence of the Modern Hospice and Palliative Care Movement Background to the Problem Purpose of the Study The Research Questions Definition of Terms Organization of the Thesis  1 2 4 7 8 9 10  CHAPTER TWO: Review of the Literature Introduction The Burden of Family Caregiving Physical, Psychosocial, and Financial Burdens Health Outcomes Ethical Implications of Home Care for the Dying The Home as the Ideal Care Setting for Dying Transformations of the Home Choice in Dying and in Caregiving Dying at Home Dying in Canada The Demography of Dying at Home Factors Influencing the Location of Dying Functional Status Symptom Management Access to Palliative Care Services The Nature of Support Family Caregiver Needs Preferences for Place of Dying and Death Home-Based Palliative Care: Mapping out the Macro Level Influences Summary of the State of Knowledge  11 12 13 18 22 22 24 25 27 28 31 34 36 37 37 41 43 46 53 58  CHAPTER THREE: Research Design and Implementation Introduction The Intersection of Ethnographic and Critical Perspectives Ethnographic Characteristics Critical Perspectives Instituting the Study Negotiating the Field Negotiating Access and Entry The Sample Data Collection Techniques Engaging in Participant Observation Writing Field Notes Conducting Interviews Collection of Documents "Making Meanings": Data Analysis Scientific Integrity Reflexivity Triangulation Face Validity Auditability Generalizability Ethical Considerations Summary  61 62 62 66 76 77 77 83 87 88 97 99 104 105 108 109 110 111 111 112 114 116  CHAPTER FOUR: The Individual Context Introduction Making the Decision Variations in Decision Making Making Promises Influencing Factors Maintaining Normalcy Institutionalized Health Care Practice Knowing What Lies Ahead Living With the Decision Constructing Caregiving as a Life-Enriching Experience Constructing Caregiving as a Life-Draining Experience Self Sacrifices Financial Sacrifices Lasting Memories Re-creating the Hospital at Home The Physical Space Home Invasions Reinventing Self Feeling Abandoned  117 118 118 123 126 127 129 137 142 143 147 147 150 154 155 156 159 165 169  vi Discourses on Euthanasia and Assisted Suicide Changing the Decision Summary  173 179 182  CHAPTER FIVE: The Community Context Introduction The Organization of Health Care Health Care Regionalization The Organization of Palliative Care Services Continuity of Care Organizational Approaches to Palliative Care Discharge Planning Visioning the Future: Palliative Care as a Core Service The Historical Context: Exploring Organizational Tensions Health Care Reform and the Rationalization of Health Care Services Pushing Out and Pushing In: The Scarcity of Institutional Practice The Rationing of Community Care: Home Support Reconstructing Home Support Reducing Access to Home Support Summary  183 184 184 190 191 194 197 199 204 212 213 219 220 225 230  CHAPTER SIX: The Sociopolitical Context Introduction The Power of Ideology The Illusion of Equal Access Palliative Care for Marginalized Populations Affordability The Ideology of Free Choice The Hidden Biomedical Imperative Paternalistic Practices The Health Care Provider Knows Best The Myth of Holism The Idealization of Dying at Home Summary  232 232 233 234 237 240 250 250 257 261 264 273  CHAPTER SEVEN: Discussion: Locating the Findings in a Broader Social Context Introduction Discourses About Choice Mediating Influences of Health Care Reform Shifting Health Care Closer to Home A Concern for Equality A Focus on Efficiency and Cost Containment  276 277 282 283 287 289  Moving Toward Health Care Regionalization Medicalization of Home-Based Palliative Care Summary  290 294 299  CHAPTER EIGHT: The Path Ahead: Conclusions and Recommendations Introduction Conclusions Recommendations Promoting Critical Consciousness Nursing Education and Nursing Practice Working with Decision and Policy Makers Expanding Public Perception Expanding Choices for Palliative Caregivers Fostering Informed Decision Making Expanding the Range of Options for Palliative Caregivers Towards the Development of a Comprehensive Continuum of Palliative Care Transcending Organizational Tensions Enhancing Continuity of Care Directions for Research Directions for Policy Development  301 301 303 304 307 314 317 319 320 323  REFERENCES  345  APPENDICES  401  Appendix Appendix Appendix Appendix  401 403 406  Appendix Appendix Appendix Appendix Appendix  A: Definition of Palliative Care B: Letter of Invitation to Participate: Health Care Providers C: Letter of Invitation to Participate: Family Caregivers D: Letter of Invitation to Participate: Health Care Administrators And Lobbyists for the Assisted Suicide Movement E: Fieldwork Calendar F: Demographic Data Collection Form G: Confidentiality Agreement: Transcriptionist and Research Assistants H: Interview Guide: Family Caregivers, Health Care Providers, Administrators I: Consent Forms  328 329 331 334 340  409 411 413 416 418 421  viii LIST OF TABLES Page TABLE 1:  Demographic Profile of Caregiver Participants  85  TABLE 2:  Demographic Profile of Health Care Providers, Administrators, and Assisted Suicide Lobbyists  87  ACKNOWLEDGEMENTS The process of completing this dissertation and my doctoral studies has been an incredible experience. Not only have I had the privilege of being mentored by some of the most outstanding academics, but also, I have had the opportunity to be supported by some extraordinary people. I am deeply indebted to the caregivers and health care personnel who participated in this study and gave so willingly of their time and energy to share their experiences. You have my heartfelt admiration and respect for your strength and selflessness. I would like to acknowledge the British Columbia Medical Services Foundation, the University of British Columbia School of Nursing, and the Xi Eta Chapter of Sigma Theta Tau for providing funding to support this research. I have also been fortunate to have remarkable support from my employer, the Capital Health Region, and from associated agencies like the Victoria Hospice Society and the Family Caregivers' Network Society. If I had to choose one word to encapsulate the experiences I have had with my Dissertation Supervisory Committee, it would be mentorship. I have had the distinct privilege to have two stellar research supervisors. I am deeply grateful to Dr. Sally Thome and Dr. Betty Davies for their unbelievable support and belief in me, for their persistent enthusiasm, and for their commitment to excellent nursing scholarship. I thank Dr. Joan Bottorff for her thoughtful comments and questions and for continually challenging me to think in new and different ways. Dr. Pat Coward has offered me unfailing support and guidance and has provided me the opportunity to begin to learn how to function in the world of health services research. And, I extend thanks to Dr. Elvi Whittaker, for her important scholarly contributions and for the intellectual challenge. I have been truly fortunate to have worked with such an outstanding team of people who were there for me at every important juncture. I am also grateful to the many friends who have supported me through this journey. Thank you to my good friend Liza McGuinness, for always believing in me. To my fellow colleagues in the doctoral program, thank you for your inspiration and guidance. I am also grateful for the ongoing encouragement from my colleagues, Darcee Bidgood, Lynn Cummings, and Terry Downing. And, I extend a heartfelt thanks to my friends Blair Paterson and Lauren Knoblauch for being committed to my project when it was most needed. I am blessed to have the most wonderful family. Thank you to Dad and Rosie, to Debbie, Larry, Jonathon, and Caitlin, and to Bonnie, for having faith in me, championing me, and keeping me grounded in what is important in life. And, to my husband Doug - You have been incredible and I would not, could not have done it without you. Your love, understanding, and support were of immeasurable sustenance to me.  DEDICATION This work is dedicated to My mother, Mary Milko (1931-1990) and my father, Clifford Milko You have: Inspired me with your strength and wisdom. Given me the confidence to succeed. Taught me about what is meaningful in life.  1 CHAPTER ONE: INTRODUCTION Situating the Study in a Historical Context Dying at home is not a new phenomenon. For centuries, many cultures have supported the idea that people should remain at home while they are dying. The home was where family, friends, and neighbours congregated to take part in an inevitable and integral part of life (Thorpe, 1993). The shift from home-based to institutional-based care for the dying can be traced to 475 A.D. when Fabiola, a Roman matron, opened a place of refuge for the sick and the dying (Campbell, 1986). These early hospices, run mainly by Christian religious orders, proliferated in the Middle Ages (Manning, 1984). In the mid 1500's, however, the advent of widespread poverty and disease transformed societal attitudes, and those who were dying were often beaten and enslaved because of their illness (Stoddard, 1978). At the same time, the support of Christian religious orders was weakened by the Protestant reformation and this resulted in secular rulers or physician groups taking over many of the services offered by hospices (Phipps, 1988). Consequently, many hospices were closed and the dying were left, for the most part, ignored and unattended (Campbell). In response to societal attitudes and the stigmatization of dying, the Sisters of Charity revived the European hospice tradition in the early 19 century, opening a th  30 bed in-patient hospice unit (Campbell, 1986). Located in London, St. Joseph's Hospice worked in tandem with a long-standing tradition of home nursing care. Throughout the early 19 century, family members became formally recognized as th  2 care providers and played a critical role in supporting those dying at home. Yet, despite changes in societal attitudes and a relative familiarity with dying and death, health care institutions gradually usurped the role of family members. These institutions became the socially accepted, and preferred, place to die (Aries, 1976; Thorpe, 1993). Dramatic changes in attitudes to death and dying again took place in the mid19 century. With the development of modern medicine, death was perceived as a th  failure by health care professionals and many began to view death as the enemy of life (Aries, 1976; Capra, 1982; Cowley, Young, & Raffin, 1992). Institutionalized death, which became the norm, was increasingly complex and highly dependent on medical technology in an effort to preserve life at all costs (Thorpe, 1993). Health care providers, in their devotion to curing, began to de-emphasize and dehumanize death (Campbell, 1986). The Emergence of the Modern Hospice and Palliative Care Movement  1  The emergence of the modern hospice movement has been linked to the inadequacy of Western medicine to face death and to care for the terminally ill (Dudgeon, 1992; James, 1992). In response, Dr. Cicely Saunders, the founder of the modern hospice movement, opened St. Christopher's Hospice in England in  T H E TERM PALLIATIVE CARE WAS ADOPTED IN CANADA IN THE EARLY 1 9 7 0 S TO DELINEATE A PROGRAM OF CARE FOR THE DYING. ALTHOUGH THE TERMS "HOSPICE" AND "PALLIATIVE CARE" ARE USED INTERCHANGEABLY THROUGHOUT THE WORLD, PALLIATIVE CARE IS MORE COMMONLY USED WITHIN THE CANADIAN CONTEXT. VARIOUS DEFINITIONS OF PALLIATIVE CARE EXIST AND ARE USED ACROSS CANADA (E.G., HEALTH CANADA, 1989A; WORLD HEALTH ORGANIZATION, 1 9 9 0 ) , ALTHOUGH THERE IS CURRENTLY NO NATIONAL CONSENSUS ON A DEFINITION. IN 1 9 9 3 , THE CANADIAN PALLIATIVE CARE ASSOCIATION ( C P C A ) BEGAN A PROCESS TO DEVELOP A NATIONAL DEFINITION FOR PALLIATIVE CARE ALONG WITH STANDARDS OF PRACTICE ( C P C A & FERRIS, 1 9 9 8 ) . FOR THE PURPOSE OF THIS THESIS, I HAVE ADOPTED THE C P C A WORKING DEFINITION ( 1 9 9 5 , APPENDIX A ) . THIS DEFINITION EXPLICITLY ACKNOWLEDGES BOTH THE PATIENT AND THE FAMILY AS THE UNIT OF CARE AND IS INCLUSIVE, COVERING ALL POSSIBLE SITUATIONS IN WHICH PALLIATIVE CARE MIGHT APPLY. 1  3 1967. Following this, programs of care for the dying proliferated in England and, almost without exception, hospice care was provided in free-standing facilities with home care services playing a secondary, but supportive, role (Seale, 1998; Torrens, 1985). Similar programs of care for the dying began to develop in Canada in the 1970s. In contrast to the United Kingdom's hospice model and home care models in the United States (U.S.), the foundation of the palliative care movement in Canada began within the acute care hospital setting (Ajemian, 1992). Few links to home care services were developed at that time, and death continued to be an institutionalized experience. Since the 1970s, there has been enormous growth in the number of hospital-based palliative care programs in Canada. The development of home care programs for the dying has gained momentum only in the past decade . This shift from institutional care to home care has been advocated by 2  policy makers and practitioners alike who maintain that people who are dying prefer to spend their final days at home, in familiar, comfortable surroundings, and prefer to be cared for by family members or those closest to them (Collett, 1999; Gilbar & Steiner, 1996; Gomas, 1993; Mann, Loesch, Shurpin, & Chalas, 1993; McWhinney & Stewart, 1994; National Forum on Health, 1997).  T H E DEVELOPMENT OF H O M E CARE PROGRAMS FOR THE DYING HAS BEEN LINKED TO AN UNDERLYING BELIEF THAT THE HOME SETTING PROMOTES NORMALCY; WHERE PEOPLE WITH LIFE-THREATENING ILLNESSES AND THEIR FAMILY MEMBERS CAN MAINTAIN FREEDOM AND CONTROL IN THEIR OWN ENVIRONMENT (DAVIES, REIMER, BROWN, & MARTENS, 1995; O'HENLEY, CURZIO, & HUNT, 1997). WIDESPREAD CRITICISMS OF HOSPITAL-BASED CARE FOR THE DYING (COWLEY, ET AL., 1992; DESBIENS, MUELLER-RIZNER, HAMEL, & CONNORS, 1 9 9 8 ) HAVE ALSO CONTRIBUTED TO THE DEVELOPMENT OF HOME CARE PROGRAMS. CANADIAN RESEARCHERS AND LEADERS IN PALLIATIVE CARE SUGGEST THAT THERE IS ALSO A PERCEPTION THAT H O M E CARE WILL RESULT IN CONSIDERABLE COST SAVINGS TO THE HEALTH CARE SYSTEM (CHOCHINOV & KRISTJANSON, 1998; R O E , 1992; SCOTT, 1992) ALTHOUGH THE EXISTING EVIDENCE IS INCONCLUSIVE IN THIS REGARD (EMANUEL, 1996; HUGHES, ET AL., 1997; SIU, 1997). 2  4 Background to the Problem  Family caregiving is a prominent issue in Canadian society (Aronson, 1998). Irrespective of whether a country has a universal, comprehensive health care system such as Canada's, estimates suggest that between 75% and 85% of all caregiving occurs in the home and is provided by unpaid family members (Kane, 1990). While Canadians have a long history of providing care and assistance to family members, the work of caregivers is not only unpaid, but also invisible (Canadian Coalition for Caregivers, 2000). Family caregiving , as an issue of 3  concern, has gained attention in the last decade as the Canadian health care 4  system has struggled to meet the demands of an aging population and an increase in the prevalence of cancer, Acquired Immunodeficiency Syndrome (AIDS), and other life-threatening conditions (Mount, Scott, & Cohen, 1993) . At the same time, 5  1 USE THE TERM "FAMILY" QUITE BROADLY, TAKING DIRECTION FROM CONTEMPORARY FAMILY THEORISTS WHO MAINTAIN THAT "IT IS QUITE POSSIBLE FOR PEOPLE TO HAVE A FAMILY EXPERIENCE (INCLUDING FEELINGS OF INTIMACY, CONNECTEDNESS, COMMITMENT, AND SO FORTH) WITH PEOPLE WHO ARE NOT IN ONE'S ACTUAL FAMILY" (HARTRICK & LINDSEY, 1 9 9 5 , P. 154). THROUGHOUT THIS DISSERTATION, I HAVE USED THE TERM "CAREGIVER" TO REFER TO A PERSON (I.E., FAMILY M E M B E R ) WHO IS PROVIDING CARE AT H O M E AND WHO IS NOT A PAID HEALTH CARE WORKER. BARRIS ( 1 9 9 8 ) MAINTAINS THAT FAMILY CAREGIVING IS EXPERIENCING A "COMING OUT" PROCESS. T H E BURGEONING INDUSTRY OF SERVICES, BOOKS, AND MEDIA ATTENTION IMPLY THAT FAMILY CAREGIVING IS BECOMING A MAINSTREAM CONCERN FOR CANADIANS (BELL, 1994; CHISHOM, 2 0 0 0 ; COMMUNITY HOSPICE ASSOCIATION OF ONTARIO, 1 9 9 4 ; JOHNSON, 1994; MCLLROY & PICARD, 1 9 9 9 ; PICARD, 1 9 9 9 ; TIBBETTS, 1 9 9 7 ; VICTORIA A I D S RESPITE CARE SOCIETY, 1 9 9 5 ; VOLKART, 1998). INTERNET WEB SITES DEVOTED TO FAMILY CAREGIVER ISSUES ARE IN PLACE (SEE WWW.CAREGIVER.COM), ONLINE DISCUSSION FORUMS HAVE BEEN INITIATED (SEE WWW.GLOBEANDMAIL.COM), AND A SERIES OF DOCUMENTARY FILMS FOCUSING ON FAMILY CAREGIVING AND PALLIATIVE CARE HAVE BEEN PRODUCED BY THE NATIONAL FILM BOARD OF C A N A D A (CURTIS & S Y M A N S K Y , 1 9 9 7 ; CURTIS, 1 9 9 9 ) . EVEN J U N E CALLWOOD, A CELEBRATED CANADIAN AUTHOR AND SOCIAL ACTIVIST HOSTED A WEEKLY TELEVISION SERIES ON FAMILY CAREGIVING ( M C K A Y , 1998). IN 1 9 5 1 , 7 . 8 % OF CANADA'S POPULATION WAS AGED 6 5 AND OVER (ANDERSON & PARENT, 1 9 9 9 ) . B Y 1 9 9 6 , THIS PERCENTAGE HAD INCREASED TO 1 2 . 2 % OF THE TOTAL POPULATION, INCLUDING 2 . 8 % OVER THE AGE OF 8 0 . STATISTICS C A N A D A ( 1 9 9 9 ) ESTIMATES THAT BY THE YEAR 2 0 1 6 , THE CANADIAN POPULATION WILL REACH JUST OVER 3 7 MILLION, UP FROM 1 4 MILLION REPORTED IN 1 9 5 1 . T H E PERCENTAGE OF THOSE AGED 6 5 AND OVER WILL INCREASE TO ALMOST 1 6 % (5.9 MILLION) AND THOSE AGED 8 0 AND OVER WILL COMPRISE JUST OVER 4 % OF THE TOTAL POPULATION (1.5 MILLION). ALONG WITH CANADA'S AGING POPULATION, NEW CASES OF CANCER AND OTHER CHRONIC, LIFE-THREATENING ILLNESSES ARE ON THE RISE. T H E CANADIAN CANCER SOCIETY ( 2 0 0 0 ) ESTIMATES THAT 1 3 2 , 1 0 0 NEW CASES OF CANCER AND 6 5 , 0 0 0 DEATHS FROM CANCER WILL OCCUR IN C A N A D A IN 2 0 0 0 , WITH 8 2 % OF DEATHS OCCURRING AMONG THOSE WHO ARE AT LEAST 6 0 YEARS OF AGE. IN 1 9 9 7 , STATISTICS CANADA 3  4  5  5 the reduction of health care budgets and the closure of hospital beds have placed increased pressure on community-based care. Canadians are becoming increasingly reliant on home care as more people die at home or spend a large portion of their illness in the home setting (Alcock, Danbrook, Walker, & Hunt, 1998; McWhinney & Stewart, 1994). Calls to enhance home care services for the dying have been echoed across the country (e.g., Canadian Senate Subcommittee on End-of-Life Care, 2000; Expert Panel on Palliative Care to the Cancer 2000 Task Force - Home Care, 1991; Senate of Canada, 1995) and the federal government has voiced its support for expansion of these services (Chretien, 1997). At a national conference on home care, the federal Health Minister asserted that there is an urgent need to augment home care programs (McLean, 1998). Yet, Anderson and Parent (1999) argue that the social reality of home care in Canada is very different from the political rhetoric. Many claim that the health care system has been largely unresponsive to the needs of home-based family caregivers (Armstrong & Armstrong, 1996; Canadian Nurses Association, 1999; Mcllroy & Picard, 1999). According to a recent study conducted by the Queen's Health Policy Research Unit, the burdens associated with caregiving greatly exceed family members' capacity to cope (Anderson & Parent, 1999). Research focusing on home care for the dying suggests that there are physical, psychosocial, and financial risks to assuming the family caregiver role (Addington-Hall, MacDonald, Anderson, & Freeling, 1991; Davis, Cowley, & Ryland, 1996; Hinton, 1994a; Martens & Davies,  ( 2 0 0 0 ) REPORTED 2 1 5 , 0 0 0 DEATHS, WITH 2 7 . 2 % CAUSED BY CANCER, 2 6 . 6 % BY HEART DISEASE AND 4 6 . 2 % BY OTHER DISEASES SUCH AS CHRONIC ILLNESS, INJURIES/ACCIDENTS, SUICIDES, AND A I D S .  6 1990; Muurinen, 1986; Rando, 1988; Steele & Fitch, 1996; Wodinsky, 1992). The social context of our modern society also produces unique challenges for families in palliative care. Migration of family members away from the home, smaller families, and changing household structures mean that there are fewer people to support those dying at home (Thorpe, 1993; Sankar, 1993; Shapiro, 1994). Changes to the dynamics and structure of the Canadian family do and will continue to have an impact on home care for the dying (Vanier Institute of the Family, 1997). While the intent of palliative care is to support families with members in the terminal phase, recent media attention and debate focusing on euthanasia and endof-life decision making also urge us to consider the implications of home-based palliative care and its influence on family members. Balfour Mount, dubbed the Canadian "father of palliative care" claims that the appeal of euthanasia and assisted suicide as a compassionate alternative to palliative care is unlikely to lessen due, in part, to increasing family caregiver burden (Kennedy, 2001, p. A2). Public calls for the decriminalization of euthanasia and assisted suicide bring forward questions of whether we, in health care, have created a social context that may overburden some patients and family members. The impetus to provide home-based palliative care is growing in an effort to curtail health care costs and to provide a more therapeutic milieu for dying and death to occur (Seaton, 1991). An expanding body of research focusing on health service utilization, satisfaction with care, demographic characteristics, preferences for location of death, and on health care providers' perceptions of home care for the dying suggests that palliative care services provide specific benefits to family  7 members (Brown, Davies, & Martens, 1990; Cawley & Gerdts, 1988; Costantini, Camoirano, Madeddu, Bruzzi, Verganelli, & Henriquet, 1993; Devery, Lennie, & Cooney, 1999; Gilbar, 1996; Gomas, 1993; Keizer, Kozak, & Scott, 1992; Kristjanson, Sloan, Dudgeon, & Adaskin, 1996; Kristjanson, Leis, Koop, Carriere, & Mueller, 1997; Lecouturier, Jacoby, Bradshaw, Lovel, & Eccles, 1999; Lewis, 1986; McWhinney & Stewart, 1994; Porta, Busquet, & Jariod, 1997; Sims, Radford, Doran, & Page, 1997). Yet, in studies that examine family caregiver experiences, there has been a tendency to focus on individual descriptions with little or no attention paid to the historical, social, economic, political, and ideological structures and processes that play a role in shaping the caregivers' experiences. Research focusing on the individual experience is important and has provided a foundation for the provision of comprehensive and family centered palliative care. But, attempts to explain and understand individual-level experiences in isolation of how they are constructed and influenced by health care relationships and encounters with the larger health care system, for example, are no longer sufficient to fully understand how family members experience and participate in the provision of palliative care at home. In light of this, the central problem addressed in this study was the interaction between family caregiver experiences, and the larger structures and processes that influence the home setting where palliative care is provided. Purpose of the Study The overall purpose of this study was to uncover how larger structures and processes influence family caregiver experiences with providing home-based palliative care. To address this problem, an ethnography, informed by critical  8 perspectives (Stewart, 1998; Thomas, 1993) was conducted. The study began by focusing on individual family caregiver experiences and then moved to focus on the social context in which this experience was shaped. My interest was to describe the experiences of family members, and to examine the dialectic between these experiences and the interactions with the larger health care system. It was anticipated that such a critical inquiry would provide a deeper understanding of the social context of home-based palliative caregiving and, thus, would provide the basis for health care decision-making and policy development in the area of home care for the dying. The Research Questions The overall research question guiding this study was: What is the experience of family members who provide palliative care at home and how does the social context influence these experiences? In order to conduct the multi-layered analysis required to understand the social context of home-based palliative caregiving, I required a framework by which to organize and conceptualize the various layers of investigation. Therefore, I began my study at a micro level, seeking to understand how family members came to the decision to provide care at home and how this decision influenced their lives. From there, I focused on meso-level influences, such as the community context in which home care was taking place. This emphasis allowed me to explore how the organization of health care and regional health care reforms had shaped individual caregiver experiences. In turn, my examination moved to a macro level with the goal of exploring how ideologies influence homebased palliative care and how the sociopolitical structures and processes inherent  9  within the Canadian health care system impact on family caregiving issues. My aim was to illustrate the interactions between the micro, meso, and macro levels of health care provision. Definition of Terms In order to clarify my key area of study, the following terms were defined: Social Context: The social context is comprised of the interactions between various dimensions of the palliative home caregiving experience. These interactions include individual caregiver experiences, and how they are influenced and shaped by community health care practices and reforms, and the processes and structures within the larger health care system and society. The Dying Period: The end of living with a life-threatening illness; when a person is expected to die within one to three months . This study focused on caregivers who 6  had provided care at home to a dying family member or who were doing so at the time of the study. Home: Any dwelling in which family caregiving takes place including traditional "homes" as defined by Western society (e.g., a house, apartment), and nontraditional dwellings such as hot-plate hotels and shelters. Family Caregiver: Any individual who has significant ties to, and who was providing or had provided care to, a person dying at home, regardless of kinship ties, common household, biological relationships, or marital status.  1 USED THE EXPECTED TIME PERIOD OF ONE TO THREE MONTHS AS A GUIDELINE FOR SAMPLING CAREGIVERS. T H E RATIONALE FOR THIS APPROACH IS FULLY EXPLAINED IN CHAPTER THREE.  6  10  Organization of the Thesis Having provided an introduction to the study in this first chapter, I now turn to a review of the literature, with particular attention to what is currently known about family members' experiences in providing palliative care at home. In Chapter Three, I detail both my theoretical positioning and the methodological approaches and decisions that I made throughout this study. Chapters Four, Five and Six outline the key findings of this research. Specifically, Chapter Four will highlight the individual experiences of family caregivers, describing how and why they came to make the decision to engage in home-based caregiving and how this decision had influenced their lives. In Chapter Five, I focus on the community context, examining how the organization of health care and regional health care reforms shaped caregiver experiences. Chapter Six extends the analysis by locating the home caregiving experience within the sociopolitical context. In Chapter Seven, I offer a discussion of key themes resulting from the study in an effort to extend our understanding of the social context of home-based palliative caregiving. Chapter Eight concludes the thesis with a presentation of key conclusions, and a discussion of further implications for practice, education, administration, research, and health policy development.  11 CHAPTER TWO: REVIEW OF THE LITERATURE It is increasingly evident that family members play an important role in the provision of home health care (Douglass, 1999). A recent survey found that more than one in eight Canadians over the age of 15, or about 2.8 million adult men and women, provided some sort of care in 1996 to people at home with chronic health conditions or life-threatening illnesses (Statistics Canada, 1997). In a study examining the status of home care in Canada, 78% of survey respondents identified family caregiving as an issue in their community (Anderson & Parent, 1999). Interest in the provision of home care for the dying has led to greater coverage of this topic by the popular media in Canada and has, in part, led to the 7  recommendation that home care become a nationally funded health service for all Canadians (National Forum on Health, 1997). While it is clear that family caregiving and home care are both issues of concern for many Canadians, there has been little systematic examination of how the larger structures and processes within the health care system influence family members' experiences with home-based palliative care. The research problem directed me to explore a range of literature. In this chapter, I review both empirical and theoretical literature in order to provide an overview of what is known about home-based palliative care, the relationship between family caregiving and homebased palliative care, and the social context of this relationship. As a starting point, I focus on the possible consequences of caregiving on family members, paying  FOR EXAMPLE, S E E DALY, 1999; MCLLROY AND PICARD, 1999; MACKINLAY, 1996; METELLA, 1998; MURPHY, 1999; PATERSON, 1999; PICARD, 1999; SCHAAL, 1999; UBELACKER, 1998; WIGOD, 1 9 9 8 .  7  12 particular attention to the physical, psychosocial, and economic burdens. The literature review moves to a theoretical discussion on the ethical implications of home care for the dying. Having established what is known about the burden of caregiving, I then review current epidemiologic and demographic trends in homebased palliative care and highlight research and review articles that detail the factors that influence the location of dying. Within this discussion, I offer a critique of the research that has focused on preferences for the location of death, and provide an overview of the theoretical arguments that support the idea that the home setting is the preferred place for palliative care to occur. Finally, I present further background information to provide context to this study, as I briefly examine some of the macro issues that influence the provision of palliative care in the home setting. The Burden of Family Caregiving There is a large and expanding body of research on family caregiving, particularly as it relates to seniors (Baumgarten, et al., 1992; Clarke, 1999; Harris, 1993; Hillman & Chappell, 2000; Lindgren, 1993; Parks & Pilisuk, 1991; Rabins, Fitting, Eastham, & Fetting, 1990; Schumacher, Stewart, & Archbold, 1998; Szabo & Strang, 1999; Wuest, Ericson, & Stern, 1994). A comprehensive review of the research literature on caregiving for the elderly exemplifies the adverse impact on both the physical and mental health of caregivers (Browning & Schwirian, 1994). While the difficult nature of caregiving in the senior population has been well established (Archbold, Stewart, Greenlick, & Harvath, 1990; Baumgarten, 1989; Boland & Sims, 1996; Cohen & Eisdorfer, 1988; Given, Collins, & Given, 1988; George & Gwyther, 1986; Zarit, Reever, & Bach-Peterson, 1980; Zarit, Todd, & Zarit,  13 1986), studies focusing on caregiver experiences in palliative care are a more recent body of work. As with the senior population, this research attests to the burdens experienced by family members providing palliative care. Physical, Psychosocial, and Financial Burdens Much of the research to date suggests that family members are at serious risk for physical, psychosocial, and financial burdens as a result of caregiving at home (Blank, Clark, Longman, & Atwood, 1989; Brabant, 1994; Davies, Reimer, & Martens, 1994; Kristjanson & Ashcroft, 1994; Lewis, 1986; Lynam, 1995; Stajduhar & Davies, 1998a; Stetz & Hanson, 1992). Physical burdens have been attributed to the sheer magnitude of work required by family members. A descriptive study with 124 bereaved caregivers demonstrated that they provided, on average, 10.8 hours per day of direct care and 8.9 hours per day of companionship to their dying family member (Wyatt, Friedman, Given, & Given, 1999). Studies have shown that family members provide a range of services such as bathing, dressing, toileting, and feeding; household chores such as laundry and house cleaning; rudimentary medical care such as injections and medications; and, general supervision and coordination of care (Folkman, Chesney, Cooke, Boccellari, & Collette, 1994; Martens & Davies, 1990; Stajduhar, 1995; Stetz, 1987; Ward & Brown, 1994; Wrubel & Folkman, 1997). In a descriptive exploratory study examining the demands made on caregivers who were spouses of terminally ill cancer patients, Stetz found that 69% of family members found the physical demands of caregiving to be the most  14 difficult . Indeed, managing the physical aspects of care has been found to be one 8  of the most demanding tasks for palliative caregivers, marked by constant monitoring, vigilant assessment and management of symptoms, and relentless personal and nursing care (Folkman, Chesney, & Christopher-Richards, 1994; Holing, 1986; Martens & Davies; Powell-Cope, 1995; Stajduhar, 1995). Schachter (1992) maintains that caregivers frequently do not anticipate the day-to-day physical labours associated with caregiving and are unprepared for the demands placed on them. While it is clear that the physical strains of home caregiving are substantial, they are often rivaled by psychosocial burdens (Neale, 1993). Studies report that family caregivers commonly express feelings of emptiness, fear, helplessness, anger, anxiety, and guilt in their caregiving roles (Atkins & Amenta, 1991; Brown & Powell-Cope, 1991; Goldberg, Wool, Glicksman, & Tull, 1985; Hileman & Lackey, 1990; Hull, 1990; Irvin, Bor, & Catalan, 1995; Jensen & Given, 1991; Kissane, Bloch, & McKenzie, 1997; Perry & Roades de Menses, 1989). Uncertainty about the timing of death and the way it might occur combined with difficulties with decision making are also commonly reported in research studies (Brown & Powell-Cope; Jensen & Given; Stajduhar & Davies, 1998a). Social burdens, including restrictions on time  IN CONTRAST TO MUCH OF THE RESEARCH TO DATE, CHAN AND CHANG (1999), IN A QUANTITATIVE STUDY EXAMINING TASKS AMONG FAMILY CAREGIVERS OF CANCER PATIENTS, FOUND THAT 7 2 % OF CAREGIVERS did not HAVE DIFFICULTY WITH PHYSICAL CAREGIVING TASKS. THIS STUDY WAS CONDUCTED WITH A SMALL CONVENIENCE SAMPLE OF 2 9 ASIAN CAREGIVERS AND THE INFLUENCES OF CULTURAL VARIATIONS TO CAREGIVING EXPECTATIONS WERE NOT ACCOUNTED FOR. HOWEVER, AS TANG ( 2 0 0 0 ) ARGUES, DYING AT H O M E HAS DISTINCT CULTURAL MEANINGS TO PEOPLE OF ASIAN DECENT AND THEIR FAMILY M E M B E R S . DYING AT H O M E IS VIEWED AS A "GLORIOUS AND FORTUNATE WAY OF DEATH" (P. 3 6 8 ) WHEREAS DYING IN HOSPITAL GOES AGAINST M A N Y ASIAN CUSTOMS. IT WOULD S E E M , THEN, THAT THE PHYSICAL TASKS OF CAREGIVING MIGHT NOT B E SEEN AS BURDENSOME BY S O M E MEMBERS OF ASIAN BACKGROUND, ALTHOUGH I RECOGNIZE THAT THERE WOULD B E SIGNIFICANT VARIATION DEPENDING ON THE ASIAN COUNTRY OR TERRITORY WHERE PALLIATIVE CARE WAS OCCURRING.  15 and freedom, and disruption of personal routines and leisure activities have also been documented (Martens & Davies, 1990; Rose, 1998; Steele & Fitch, 1996). For example, Brown and Stetz (1999) conducted in-depth interviews with 26 people caring for an adult with either AIDS or cancer and found that many caregivers put their lives on hold and felt an overwhelming sense of being alone in their caregiving role. Feelings of uncertainty, helplessness, anxiety, and fear have been attributed to relationship and role changes that occur in the family when a member becomes terminally ill (Kristjanson & Ashcroft, 1994; Lewis, 1986; Schachter& Holland, 1995). Caregivers are often confronted with perplexing and changing problems of identity in the course of caring for a dying family member (Davies, Chekryn-Reimer, & Martens, 1990; Hull, 1989). In a grounded theory study examining the impact on families having a member with advanced cancer, Davies and colleagues interviewed 66 family members, both individually and as a family unit, in order to determine how they managed the illness situation and the patient's care, and their perceptions of care at home and in hospital. Constant comparative analysis revealed that family functioning influences the palliative care experience and that many family members need to redefine or reconceptualize themselves to adjust to a new reality and to accommodate to the change in their loved one's health status. Although the study provides a detailed picture of how families function when a member has advanced cancer, it did not explicitly focus on how caregiving in the home setting had influenced family functioning.  16 The literature suggests that family caregivers also encounter financial hardships as a result of in-home caregiving. Studies have documented the subjective sense of economic burden and have found variations in the amount of burden reported. For example, one survey found that approximately 25% of all caregivers experienced financial difficulties (Davis, et al., 1996); another study reported 44.9% (Emanuel, Fairclough, Slutsman, & Emanuel, 2000); and another, 60% (Muurinen, 1986). Financial burden has been attributed to several factors. Caregiving can disrupt work performance resulting in missed time from work and unemployment (Addington-Hall, MacDonald, Anderson, & Freeling, 1991; Devery, Lennie, & Cooney, 1999; Muurinen; Sykes, Pearson, & Chell, 1992). The costs incurred by caregivers in the home setting can also be substantial (Armstrong & Armstrong, 1996; Fierlbeck, 1997). The cumulative effect of buying goods and services, including medications, equipment, and home care services, can drain family resources. This is the case even in Canada where government health plans ensure a basic level of coverage. Steele (1999) maintains that "people assume that all costs are covered by insurance. Most insurance plans, however, do not pay for incidental costs. The cumulative effect of these costs may take a toll on family caregivers" (p. 48-49). Studies have reported that families caring for a terminally ill member have had to spend their life savings, take out loans or mortgages, or obtain additional jobs to ease their financial burden (Emanuel, et al.). The U.S. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment (SUPPORT) reported that families of seriously ill patients experienced substantial  17 economic losses. In 20% of families, a family member had to stop working; 31 % of families lost most of their savings (Covinsky, et al., 1994). While there is little research that documents the actual costs incurred by family members providing at-home palliative care, the costs are believed to be significant. Some authors argue that the widespread and substantial move away from long stays in hospital facilities has had the effect of pushing sizeable costs onto family caregivers (Armstrong & Armstrong, 1996). Others maintain that 9  these costs can be so substantial that they limit the ability of family members to provide home care (Arno, Bonuck, & Padgug, 1995; Wodinsky, 1992). Lost opportunity costs, such as forsaking chances for employment outside of home, are rarely considered in determining the cost-effectiveness of home-based palliative care (Arras & Neveloff-Dubler, 1995; Whynes, 1997). Peters (1995) maintains that more often than not, Canadians choose family reliance over state intervention in the care of family members. At the same time, the National Forum on Health (1997) reports that Canadians are concerned about resource allocation decisions being made on this assumption without consideration given to caregiver compensation . 10  ARMSTRONG AND ARMSTRONG ARGUE THAT COST SAVINGS TO THE CANADIAN GOVERNMENT ARISE BECAUSE COSTS ARE TRANSFERRED (I.E., THEY DO NOT DISAPPEAR) TO INDIVIDUALS AND FAMILY M E M B E R S . THEY STATE THAT "ALTHOUGH THERE ARE DIFFERENCES AMONG PROVINCES IN THE PROVISION OF COMMUNITY SERVICES, NONE ENTIRELY COVER ALL THE REQUIRED HOURS OF CARE. INDIVIDUALS AND FAMILY M E M B E R S MUST M A K E UP THE DIFFERENCE EITHER BY HIRING H O M E CARE SERVICES, USUALLY FROM THE PRIVATE SECTOR, OR BY DOING ALL THE CARE THEMSELVES" ( 1 9 9 6 , P. 185). T H E C P C A AND PROVINCIAL HOSPICE/PALLIATIVE CARE ASSOCIATIONS ARE BECOMING ACTIVE IN THE AREA OF FAMILY CAREGIVER PROTECTION (SEE WWW.CPCA.NET/FCP.HTM). IT IS THE POSITION OF C P C A THAT JOB PROTECTION AND PAID LEAVE FOR A FAMILY M E M B E R WISHING TO CARE FOR A DYING LOVED ONE IS A SIGNIFICANT SOCIAL POLICY ISSUE FACING CANADIANS. MOST PROVINCES PROVIDE LITTLE OR NO JOB PROTECTION FOR THOSE TAKING LEAVE TO PROVIDE CARE AND NONE PROVIDE INCOME SECURITY. INCOME PROTECTION IS WITHIN FEDERAL JURISDICTION. FEDERAL PROGRAMS SUCH AS EMPLOYMENT INSURANCE AND THE CANADA PENSION PLAN DO NOT PROVIDE INCOME SECURITY FOR CAREGIVERS. C P C A IS NOW IN THE PROCESS OF ACTIVELY PURSUING THE ISSUE OF JOB PROTECTION AND INCOME SECURITY FOR FAMILY CAREGIVERS PROVIDING END-OF-LIFE CARE. 9  1 0  18 Health Outcomes The physical, psychosocial, and economic demands of home-based caregiving have been linked to negative health outcomes for caregivers including chronic fatigue, depression, prolonged stress, physical exhaustion, sleeplessness, and burnout (Axelsson & Sjoden, 1998; Davis, etal., 1996; Hinton, 1994a; Hull, 1990; Jarrett, Payne, & Wiles, 1999; Schachter, 1992; Vachon, 1998; Wyatt, et al., 1999). A recent survey of 1,300 Canadians reported that over 60% of family members found caregiving difficult and that providing care to an ill person had a substantial impact on their own physical and mental well-being (Barris, 1998). Similarly, a British study investigating the effects of terminal illness on patients and their caregivers found that almost 50% of family members experienced ill health because of caregiving (Davis, et al.). In both of these studies, however, ill health was not well described. In the U.S., family members providing end-of-life care have three times as many stress symptoms as the general population and take more prescription drugs to treat depression and stress (Mezey, Miller, & Nelson-Linton, 1999). In a longitudinal study assessing stress and coping in 253 caregivers of men with HIV/AIDS, caregiving partners were found to be at high risk for clinical depression resulting from the adjustments needed to cope with the ill person's unpredictable disease course (Folkman, Chesney, & Christopher-Richards, 1994). Studies conducted with spouses of terminally ill cancer patients also report that up to one third of caregivers have depressive symptoms as a result of caregiving  19 (Emanuel, et al., 2000; Greer, et al., 1986; Kissane, Block, Burns, McKenzie, & Posterino, 1994). In a large population-based study, Emanuel and colleagues conducted in-person survey interviews with 988 terminally ill patients and 893 caregivers in six randomly selected U.S. states to gain an understanding of the burdens associated with terminal illness. The caregiver survey contained 118 questions focusing on health status and symptoms, social supports, communication with health care providers, personal and spiritual meaning, care needs, end-of-life care plans, economic burdens, sociodemographic characteristics, euthanasia and assisted suicide, and interview-related stress. While the study report provides detailed findings on terminally ill patients, the authors did not provide detailed findings on caregiver experiences and did not define or systematically report on economic and "other" burdens associated with terminal illness. Nevertheless, some of the statistical analyses revealed caregivers were more likely to have depressive symptoms and to report that caregiving substantially interfered with their lives. The study also demonstrated that, as the disease progressed and the amount of care for the terminally ill person increased, the level of burden also increased. Similarly, Hinton (1994a) found that family caregivers suffered more emotional and physical distress as the disease progressed with 17% of caregivers being rated as seriously depressed and 14% very anxious. Caregiving places considerable strain on family members regardless of the country in which it is occurring. Although the process of caregiving has been shown to have negative health outcomes, there is some evidence to suggest that caregivers neglect to seek medical attention or support for their health concerns  20 (Stetz & Hanson, 1992). Stetz and Hanson conducted a follow-up study with 31 bereaved spouses to determine if their perceptions of caregiving demands had changed over time. A mailed survey with seven open-ended questions was used to collect data and the responses were subjected to content analysis. Survey findings indicated that over half of all caregivers had suffered personal health problems and that they regretted not seeking help with caregiving. Research supports these findings and indicates that family members often think only of the needs of the dying person, at the expense of their own needs (Ryan, 1992; Skorupka & Bohnet, 1982). Some family members use avoidance as a coping mechanism to deal with the demands of caregiving. A phenomenological study with six home-based family caregivers found that avoidance was the predominant method used by family members to cope with the stresses of caregiving. These behaviors included not thinking about it, staying busy, deferring from asking questions about the dying process, and taking medications to relax (Brinson & Brunk, 2000). Even though this study provides some useful information on how family members cope with the stress of caregiving, the small sample size limits its generalizability. It is only recently that researchers have begun to investigate the long-term effects on caregivers when they provide palliative care at home. Addington-Hall and Karlsen (2000) conducted a secondary analysis of data from a large retrospective population-based survey of a random sample of people who died in 20 English health districts. Caregiver respondents were asked about the extent to which they missed the deceased, whether they could look forward to things, whether they felt they had come to terms with the death, and how well they considered things to be  21 going for them. Using the General Health Questionnaire, a validated measure of psychological morbidity, the researchers found that 10 months after the death, bereaved caregivers of cancer patients who had died at home were found to have significantly higher levels of psychological distress than caregivers of patients who had died elsewhere. Caregivers were also more likely to report that they missed the deceased a great deal, and less likely to report that they could look forward to things, that they had come to terms with the death, or that things were going well for them. These findings, however, are preliminary and the authors caution that replication and longitudinal studies are required to explore the long-term effects of providing home-based palliative care on caregivers. These unexpected findings led the authors to conclude that palliative caregiving at home may provide benefits to dying patients but may cause considerable distress for caregivers. The study supports earlier conclusions reached by Hinton (1994b) who reported that bereaved caregivers often had serious misgivings about caring for their dying family member at home. The literature presented on caregiver burden clearly indicates that there are significant risks to assuming the family caregiver role. However, with few exceptions, most studies focus on the description of tasks and do not fully explain the context of caregiver burden or provide detailed accounts of the actual causes of burden in palliative care. For the most part, burdens are mentioned almost in passing. Yet, without a broader understanding of the context of caregiver burden and those factors that might influence the home setting where palliative care is occurring, it is difficult to identify mechanisms that could meet the care needs of  22 home-based caregivers without imposing additional hardships on them. Further, most of the available research has been conducted in Britain, the U.S., or Australia. While these studies provide important information about family members' experiences, they may not be entirely applicable to the Canadian context because of differences in the organization of palliative care services. Ethical Implications of Home Care for the Dying The physical, psychosocial, and economic burdens associated with caregiving have prompted scholars to write about the ethical and moral problems associated with dying at home (Arras, 1995; Arras & Neveloff-Dubler, 1995; Coyle, 1997; Di Mola, 1997; Ruddick, 1995; Sankar, 1993). These writings have focused primarily on pointing out the possible implications of such care and have served to critique ideological perspectives that (sometimes) indiscriminately promote the "home death movement" . While it is not my intent here to present all of the 11  possible ethical and moral concerns in detail, I will highlight some of the predominant arguments that have arisen in the literature as a way to provide further background 12  and context to this study. The Home as the Ideal Care Setting for Dying The modern palliative care movement has strongly embraced the idea that the home setting represents the most appropriate place for dying and death to occur. In Canada, as in many Western countries, one of the primary goals of palliative care  I USE THE TERM " H O M E DEATH MOVEMENT" HERE TO EXEMPLIFY THE CURRENT PUSH, BOTH AT A PHILOSOPHICAL LEVEL AND A MORE PRACTICE LEVEL (E.G., ECONOMICS), TO ENABLE MORE DYING PEOPLE TO REMAIN AT HOME. JOHN D . ARRAS (1995), A PROFESSOR IN THE PHILOSOPHY DEPARTMENT AT THE UNIVERSITY OF VIRGINIA, HAS WRITTEN AN EXCELLENT BOOK ENTITLED BRINGING THE HOSPITAL HOME, WHICH FOCUSES EXPLICITLY ON THE ETHICAL AND SOCIAL IMPLICATIONS OF HIGH-TECH H O M E CARE. MANY OF THE CHAPTERS OFFER COMPREHENSIVE DISCUSSIONS ABOUT THE ETHICAL IMPLICATIONS OF H O M E CARE FOR THE DYING. 1 1  1 2  23 is to support people to die in their own home (Fraser, 1990; Mor & Hiris, 1983; Ramsay, 1992). In an article entitled, Enabling more dying people to remain at home, Thorpe (1993) contends that, "when it comes to dying there is no place like home" (p. 915). Indeed, "in a view reflected in law, poetry, and cliches, home is commonly taken to be a refuge, castle, haven, or nest" (Ruddick, 1995, p. 167). It is argued that home is where people feel most comfortable. Home represents a place where we can be at ease, away from the pressures to conform to social norms. As Ruddick explains: For many people, home is where they feel most truly themselves, the very center of their lives, closest relationships, and most intense emotions. As such, home is their natural or, in Aristotelian terms, their "telic place" where mature selves are realized and revealed (p. 167). Current conceptualizations of "home" provide an attractive alternative to institutional-based palliative care. Aries (1981) describes an institutionalized death as an unnatural experience: "the death of the patient in hospital, covered with tubes, is becoming a more popular image than the transi or skeleton of macabre rhetoric" (p. 614). These views have advanced the home death movement and have led some people to believe that the care provided at home (by family members) will be far superior to that of a depersonalized and paternalistic hospital setting (Coyle, 1997; Ruddick, 1995). However, when caregivers are led to believe that dying at home is a measure of success, and dying in hospital, a measure of failure, there is a risk that they will think they have failed their loved one when they cannot cope with the demands of home caregiving. Coyle argues that when patients and families are unsupported in the home setting, the dying period can seem unending. As such, "a rapid death becomes the hoped-for goal, while quality-of-life until death and the  24 ability to get pleasure from a moment or a day is lost. What could have been a lifefulfilling experience becomes instead a life-draining one" (p. 41). Sankar (1993) contends that our current impressions of dying at home are constructed upon romantic notions of the past where the dying were surrounded by loyal family members who had ample time to devote to caregiving. Sankar argues that technological and social changes, however, have influenced the dying at home experience and have resulted in outdated images. These images raise false expectations and inappropriate assumptions about what is involved in caregiving. Thus, despite all of the best of intentions, dying at home is not always ideal (Arras & Neveloff-Dubler, 1995). Transformations of the Home Ethicists argue that the home environment is often transformed when the provision of care is shifted from the institution to the home. Physical changes to the home in order to accommodate caregiving can transform what was once a restful and comforting environment into one that is confining and repugnant (Kapp, 1995; Ruddick, 1995). Collopy, Dubler, and Zuckerman (1990) claim that it is tempting to assume that the home provides families with greater control to self determine than do hospitals, but this is not always the case. Multiple care providers coming into the home and complicated technological mechanisms to support symptom management at home can exert a significant cost in terms of important social values associated 13  These examples (i.e., multiplicity of health care personnel and complicated technological procedures) are used to illustrate how home life can be influenced when one makes a decision to provide at-home palliative care. I do, however, acknowledge that there may be other examples that could also influence home and family life. 13  25 with the home environment. Arras and Neveloff-Dubler (1995) question how someone can truly be "at home" with the extension of health care provision into the private sphere of patients and families. As they point out: Dying at home may be initially attractive in an abstract fashion to both patients and family members, but the reality is rarely benign. Agonal breathing, incontinence of urine and feces, and fear of impending death are facts about the dying process that have been largely hidden from Americans in an era when 80 percent of people die in hospitals and nursing homes (p. 3). Some authors have expressed concern over the extension of medical care into the home, warning that hypermedicalization can seriously influence family systems 14  (Arras & Neveloff-Dubler; Di Mola, 1997). These authors also caution that the dyingat-home experience has the potential to become bureaucratized , falling into all of 15  the medical trappings of a hospital-based system. Choice in Dying and in Caregiving Another ethical issue that arises is whether or not people truly have a choice; choice for the patient in terms of where they die and choice for the family member in terms of whether they become caregivers. In a theoretical paper, Dudgeon and Kristjanson (1995) raise some key questions regarding the realization of one's preference for dying at home. On the basis of findings from a study conducted by McWhinney and colleagues (1995), the authors argue that the financial resources  "Hypermedicalization" is a term used to denote the extension of medical care into the private sphere of family and friends (Arras & Neveloff-Dubler, 1995). In the mid-1970s social critic Ivan lllich (1976) warned of the dangers of "social iatrogenesis" or an inability to cope with our surroundings engendered in part by the medicalization of everyday life. More broadly, Sawicki (1991) explains that the term "medicalization" usually "implies the negative phenomenon of reducing political, personal and social issues to medical problems thereby giving scientific experts the power to 'solve' them within the constraints of medical practice" (p. 119). James and Field (1992) offer a provocative discussion about the routinization and bureaucratization of hospice and palliative care suggesting that the hospice movement may be in danger of being unable to sustain its founding ideals. 14  15  26 required to support dying at home may influence whether or not people actually have a choice in where they die. Others suggest that having a choice in where one dies is more illusory than real. Arras and Neveloff-Dubler (1995) emphasize that there may not actually be a "choice" when considering care options. They suggest that there is no choice between an often highly regimented, rule-based institutional system versus the fragile, yet comfortable home environment. Patients and families invariably select the most appealing option (i.e., the home) because the alternative looks so bad. Some authors have hypothesized that the choice to become a caregiver is often not a choice per se but, rather, an obligation particularly if the caregiver is a woman. With the exception of husbands caring for their terminally ill wives or an increasing number of gay men caring for their partners, numerous research studies support the belief that the majority of caregiving work is provided by women (Aronson, 1992; Bunting, 1992; Graham, 1985; Rutman, 1996; Wuest, 2000). Furthermore, notwithstanding the fact that many women value their caregiving roles and may come to regard them as opportunities for reciprocity and gratitude, ethicists have argued that "our rosy picture of 'home' is often parasitic on widespread but socially unjust roles for women caretakers" (Arras & Neveloff-Dubler, 1995, p. 8). Some women may have to forsake opportunities for professional or personal satisfaction outside of the home or abandon opportunities for employment in order to provide care in the home and may come to see their lives reduced to little more than a maid (Tedlock, 1995).  27 The ethical arguments presented here uncover some of the contextual issues that may influence the home caregiving experience. With few exceptions, these arguments have not been supported by research but have stayed within the realm of theoretical and philosophical discussion. Studies focusing on the contextual factors that may influence dying at home are noticeably absent in palliative care discourse and represent a knowledge gap in the area of home care for the dying. In other words, while the arguments provide context and substance to the discourse on home care for the dying, few scientific studies examine this topic. Dying at Home The ethical arguments, together with a review of the caregiver burden literature, depict a rather dismal existence for family members who provide palliative care at home. However, many research studies have documented the benefits to family members who are also caregivers. Current evidence verifies that some family members prefer home-based palliative care to care provided in an institution, particularly if adequate health services are in place to support them. In this section, I will review what is known about dying at home. To begin, I summarize the reasonably limited epidemiological data on dying in Canada. I then provide a synopsis of the demographic characteristics of those dying at home and highlight research and review articles that detail the factors that influence the location of dying. In concluding this section, I offer a critique of the research on preferences for location of death, including a review of why the home is often the preferred location for palliative care.  28 Dying in Canada It is estimated that over 220,000 Canadians die each year (Senate of Canada, 2000). Although some studies indicate that people generally prefer to die at home, very little is known about where people actually die . In an effort to 16  understand the epidemiology of dying in Canada, Heyland, Lavery, Tranmer, Shortt, and Taylor (2000) conducted a cross-sectional analysis of death records in 1997. The primary purpose of the study was to determine the proportion of all deaths in Canada that occurred in hospitals. Data were obtained from provincial registries of vital statistics and from the Canadian Institute for Health Information (CIHI). Provincial death records were compared and then these data were aggregated to determine the proportion of total deaths . Of the 201,892 deaths that were reported 17  by the provincial registries in 1997, 73% occurred in a hospital. Hospital deaths for each province were also reported with Quebec having the highest proportion  18  (87%)  and the Northwest Territories having the lowest (52%). The province of British Columbia (B.C.) had a relatively high proportion of hospital deaths (78%) in comparison to other provinces where the average was approximately 67%. The  RESEARCHERS IN THE U . S . AND THE UNITED KINGDOM ( U . K . ) HAVE CONDUCTED EPIDEMIOLOGICAL STUDIES ON THIS TOPIC. FINDINGS INDICATE THAT MOST AMERICANS DIE IN EITHER HOSPITALS OR LONG-TERM CARE FACILITIES, AND STUDIES FROM THE U . K . SHOW SIMILAR RESULTS (SAGER, EASTERLING, KINDIG, & ANDERSON, 1989; SEALE & CARTWRIGHT, 1 9 9 4 ) . STUDIES FOCUSING ON WHERE CANADIANS DIE ARE VIRTUALLY NON-EXISTENT IN THE LITERATURE, DESPITE THE RELEVANCE THAT THIS KIND OF INFORMATION HAS FOR GUIDING PALLIATIVE CARE PRACTICE, AND FOR THE DEVELOPMENT AND EVALUATION OF EVOLVING HEALTH POLICY IN THIS AREA. EACH PROVINCIAL GOVERNMENT IS REQUIRED TO MAINTAIN A REGISTRY OF DEATH STATISTICS UNDER THE FEDERAL VITAL STATISTICS ACT. IN MOST PROVINCES, HOSPITALS ARE INCLUDED IN RECORDING LOCATION OF DEATH WITH THE EXCEPTION OF NEWFOUNDLAND AND MANITOBA WHERE HOSPITAL-BASED DEATHS ARE NOT SPECIFIED IN THEIR PROVINCIAL TOTALS. THEREFORE, DATA FROM NEWFOUNDLAND AND MANITOBA WERE EXCLUDED FROM THE ANALYSIS IN THE STUDY (HEYLAND, ET AL., 2 0 0 0 ) . MOREOVER, HOSPITALS DO NOT INCLUDE NURSING HOMES, HOMES FOR THE AGED, OR HOSPICES, EXCEPT IN Q U E B E C . CONSEQUENTLY, NOT ALL INSTITUTION-BASED DEATHS ARE CAPTURED IN THE ANALYSIS, BUT ONLY THOSE RELATED TO ACUTE CARE HOSPITAL DEATHS. T H E HIGHER PROPORTION OF HOSPITAL DEATHS RECORDED IN Q U E B E C IS DUE TO THE FACT THAT THE RESEARCHERS WERE UNABLE TO DISAGGREGATE THE DATA SO THAT NURSING HOMES, HOMES FOR THE AGED, OR HOSPICES WERE INCLUDED IN THE OVERALL PROVINCIAL ANALYSIS. 1 6  1 7  1 8  29 study was descriptive in nature and the authors offered little explanation as to why the proportions might differ among provinces. Heyland and colleagues (2000) theorize that health-system factors such as the availability of hospital and nursing home beds, and not patient preferences or patient characteristics, may be a strong determinant of the location of death,  19  although these factors were not examined in the study. Clearly, there is a need to gain a better understanding of why hospitals function as the major providers of endof-life care in Canada. There is also a pressing need to determine the circumstances that may inhibit home-based palliative care. The researchers concluded by posing a number of questions that have not been examined within the Canadian context. They suggest that investigations are needed to understand the factors that influence the provision of home-based palliative care and to help explain why 73% of Canadians currently die in hospital. I was unable to find any data that provide accurate estimates on the total number of Canadians who die at home. While death record data is available from provincial registries of vital statistics and from the CIHI, to my knowledge, they have not been subjected to any systematic analysis. Many palliative care programs across the country keep their own records that document location of death. However, they typically only capture this data on patients who are registered with the program. Some provincial home care programs or regional palliative care services keep this type of data, but this practice varies by province and the accuracy of such  A STUDY REPORTED ON IN THE MEDICAL TRIBUNE, A MEDICAL NEWS SOURCE FOR THE N E W YORK T I M E S SYNDICATE, ALSO SUGGESTS THAT THE LOCATION OF DEATH IS MORE STRONGLY INFLUENCED BY THE NUMBER OF HOSPITAL BEDS IN A COMMUNITY THAN BY THE PATIENT'S PREVIOUSLY EXPRESSED WISHES (CHRISTENSEN, 1998).  30 information is dependent on how data systems are organized and how the data is entered. Only a few Canadian studies could be found that document the proportion of people dying at home in their local health authorities. In 1992, Lubin conducted a retrospective study using the charts of 96 patients admitted to a palliative care service in B.C. These patients were followed until death and the location of death was recorded. Only five percent of patients in the study died in their own homes. Subsequent studies have shown that the number of people dying at home may be increasing. A retrospective chart review of patients referred to the palliative home care support team in London, Ontario showed that 28% of all patients on the program died at home (McWhinney, Bass, & Orr, 1995) and 68.3% of patients were able to die at home with support from a palliative care at-home team in Scarborough, Ontario (Gardner-Nix, et al., 1995). Variations in the number of people dying at home may be attributed to the type and level of health care services that are available in the particular community where palliative care is occurring. These variations were recently described from an international perspective (Grande, Addington-Hall, Todd, 1998). Despite this, Canadian researchers contend that local studies must be carried out to better understand the characteristics and circumstances that influence home care for the dying (Fainsinger, Demoissac, Cole, Mead-Wood, & Lee, 2000). In this way, solutions that account for the Canadian context can be derived in order to improve the care provided to patients and their family members in palliative care.  31 The Demography of Dying at Home Many research studies that focus on the provision of palliative home care record demographic characteristics of patients to describe those people who are most likely to receive palliative care in the home setting or die at home . Relatively 20  few studies focus purely on sociodemographic characteristics but more frequently report on age, gender, diagnosis, and social class as part of a larger research project. I will not be providing a detailed analysis of what has been reported on within all of these studies but, rather, will offer a summary of what is known about the demography of dying at home. Most studies have found that younger people die at home more frequently than the elderly (Higginson, Astin, & Dolan, 1998; Hunt, Roder, & MacHarper, 21  1989; Karlsen & Addington-Hall, 1998; Moinpour & Polissar, 1989; Seale & Cartwright, 1994). However, multivariate analysis from a population-based study of 12,343 cancer deaths in Italy from 1986 to 1990 demonstrated that the probability of home death increased with advancing age (Costantini, et al., 1993). In contrast, a correlation study by Axelsson and Christensen (1996) found no effect of age on location of death, although the number of home deaths in their sample was small (n=24).  SEVERAL RESEARCHERS HAVE INVESTIGATED PLACE OF DEATH FOR BOTH CANCER AND NON-CANCER PATIENTS AND HAVE REPORTED ON THEIR DEMOGRAPHIC CHARACTERISTICS. SEALE ( 1 9 9 1 B ) NOTES THAT CANCER PATIENTS DIFFER FROM PATIENTS IN OTHER DISEASE GROUPINGS: THEY TEND TO DIE YOUNGER AND HAVE MORE SOCIAL SUPPORTS; MENTAL CONFUSION AND LONG TERM DISABILITY ARE LESS COMMON; AND, THE INCIDENCE, DURATION, INTENSITY, AND TYPE OF SYMPTOMS FOLLOW A DIFFERENT TRAJECTORY. CHRISTAKIS AND ESARCE ( 1 9 9 6 ) REPORT THAT THOSE MOST LIKELY TO DIE AT H O M E HAVE CANCER. BECAUSE THE DEMOGRAPHIC VARIABLES RELATED TO PLACE OF DEATH MIGHT B E DIFFERENT TO THOSE OF THE REST OF THE POPULATION, I HAVE ONLY REVIEWED LITERATURE PERTAINING TO CANCER PATIENTS. WITH FEW EXCEPTIONS, AGE DEMARCATIONS ARE NOT OFTEN REPORTED IN THESE STUDIES. THUS, HOW THE RESEARCHERS DEFINE "YOUNGER" PATIENTS AND THE "ELDERLY" IS NOT ALWAYS CLEAR. 2 1  32 Some of the research literature suggests that males are more likely to receive palliative home care, or die at home than their female counterparts (Higginson, et al., 1998; Hunt, et al., 1989; Seale & Cartwright, 1994). Yet, other evidence indicates that a higher proportion of women die at home (Costantini, et al., 1993; Gilbar & Steiner, 1996). And like age, some studies suggest that gender is not significantly associated with dying at home (Axelsson & Christensen, 1996). In a comprehensive literature review investigating the relationship between patient characteristics and home deaths, Grande, et al. (1998) suggest that the variation in study findings related to age and gender are likely related to the cultural and family context in which patients live. These authors advise that further research is needed to better understand these contexts and how they may influence the provision of palliative care at home. Studies with cancer patients suggest that the type of cancer is associated with whether or not the provision of palliative care or death occurs at home. Genitourinary and gastrointestinal cancers have been shown to be more strongly associated with dying at home (Costantini, et al, 1993; Johnson & Oliver, 1991; McCusker, 1983) than have haematological cancers, head and neck cancers, or lung cancers (Costantini, et al.; McCuster; Polissar, Severson, & Brown, 1987). These deviations have been attributed to the specific treatments and symptoms that result from different types of cancer. Axelsson and Christensen (1996) did not find that the place of dying was associated with a specific type of cancer diagnosis but, again, the sample size for home-based patients in this study was small and thus, the findings should be viewed with caution.  33 Studies investigating the relationship between socioeconomic status, education levels, and home-based palliative care have shown that cancer patients who have higher education or who live in higher socioeconomic neighborhoods are most likely to die at home (Costantini, et al, 1993; Gilbar & Steiner, 1996; Seale, Addington-Hall, & McCarthy, 1997). Contrariwise, data from a study analyzing the place of death for 820 cancer patients in a British health authority reported significant variations in the place of death according to social class (Sims, Radford, Doran, & Page, 1997). The study examined data that were categorized according to occupation and found that those in skilled and semi-skilled occupations were more likely to die at home than those in professional, managerial, and technical occupations. Even though these findings show that people with skilled and semiskilled occupations are more likely to die at home, this group made up more than half of the sample population. Therefore, the authors' conclusion that people of lower socioeconomic status are more likely to die at home may have had more to do with the sample size of those employed in skilled and semi-skilled occupations than with providing an accurate representation of the entire sample. The authors' postulated that the variations in the place of death according to social class may be influenced by access to services and the availability of social support within different sections of the health authority. Despite the fact that researchers report significant variations in age, gender, diagnoses, and socioeconomic status, this information helps to describe those people who are more likely to receive palliative care in the home setting. With few exceptions, studies on the demography of dying, however, do not account for  possible contextual factors that might influence why some patients receive home palliative care over others. Most of the researchers theorize about what these influences might be, but do not include these variables in their analyses. Some of the work that has been done to delineate the factors influencing the location of dying provides some context to the home-based palliative care experience. Factors Influencing the Location of Dying Regardless of where palliative care is provided, dying Canadians and their family members have identified several outcomes that they associate with quality end-of-life care: (a) adequate pain and symptom control; (b) avoiding inappropriate prolongation of dying; (c) achieving a sense of control; (d) relieving the burden of others; and, (e) strengthening relationships with loved ones (Singer, Martin, & Kelner, 1999). Providing quality end-of-life care, particularly in the home setting, is a challenging undertaking (Cantwell, Turco, Brenneis, Hanson, Neumann, & Bruera, 2000). A review of the literature suggests that there are a number of conditions that need to be in place to effectively support dying at home. These include: 1. A desire on the part of the patient to be cared for at home (Beck-Friis & Strang, 1993; McCorkle, 1988; Stajduhar & Davies, 1998b); 2. A desire on the part of the caregiver to provide care at home (Beck-Friis & Strang; DeConno, et al., 1996); 3. The availability of skilled medical support on a 24-hour basis (Beck-Friis & Strang; DeConno, et al; McCorkle); 4. The availability of more than one caregiver, especially if the primary caregiver is elderly and the patient has been sick for several weeks or months (Beck-  35 Friis & Strang; Thorpe, 1993; Dudgeon & Kristjanson, 1995; Wilson, 2000); and, 5. Having the financial resources to allow caregiving to take place at home (Dudgeon & Kristjanson; Thorpe). Based on these factors, a group of Canadian researchers conducted a prospective study to describe the relative role of these factors in predicting home death in a cohort of palliative care patients with advanced cancer (Cantwell, et al., 2000). The investigators developed a five-item questionnaire with sub-sections and pilot-tested it with 10 patients, making only minor revisions. Ninety questionnaires were administered by trained home care coordinators and a follow-up questionnaire was carried out to record place of death. Data were analyzed using multivariate techniques. The desire for a home death by both the patient and the caregiver, support of a family physician, and the presence of more than one caregiver were all found to be significantly associated with a home death. Logistic regression analysis identified the desire for a home death by both the patient and family caregiver as the main predictive factor for a home death. While the study signifies one of the first attempts to prospectively examine the factors that need to be in place for a successful home death, the authors acknowledge that there are some limitations. The questionnaire concepts were not well defined. Because of this, some of the caregivers that participated in the study were family members who visited the patient but who were not necessarily the primary care provider at home. Furthermore, some evidence suggests that patients and family caregivers change their minds about their desire for receiving palliative care at home or for a home death (Hinton,  36 1994a). Accordingly, making predictions based on a one-time response may not provide an entirely accurate picture of the factors that predict home death. Longitudinal studies are required, particularly with caregivers, to determine if their perspectives change over time. Functional Status Research reports have indicated that there are other circumstances that influence the ability to care for a dying person at home. In a grounded theory study of 23 families (71 family members), Davies and colleagues (1995) found that the patient's physical condition determined whether home-based caregiving was feasible. Findings suggest that family caregivers were concerned about their ability to manage bed-bound and/or incontinent patients. These findings are supported by research that confirms that patients are more likely to be admitted to, or maintained on, an acute care hospital ward if they have significant functional disability (Fainsinger, et al., 2000). An examination of the discharge characteristics of palliative care patients in an acute care hospital found that patients with better functional ability, as measured on the Karnofsky Performance Status Scale and the Palliative Performance Scale, were far more likely to return home than those with significant functional losses (Fainsinger, et al.). Similarly, a report by Karlsen and Addington-Hall (1998) on a survey of 229 cancer deaths in London, England concluded that high levels of dependence inhibited the likelihood of patients dying at home. A person's impaired functional status has also been linked to reduced quality of life in home hospice family caregivers (Weitzner, McMillan, & Jacobson, 1999).  37 Symptom Management Managing pain and other symptoms, often with complicated technological devices, also produces considerable distress for family caregivers and can discourage home care for the dying (Davies, et al., 1995). Lubin's (1992) study, aimed at identifying some of the obstacles to home care, showed that pain, weakness, and shortness of breath were three of the top five reasons why people were admitted from home to hospital for palliative care. Beck-Friis and Strang (1993) also suggest that symptoms must be controlled in order for terminal home care to be satisfactorily achieved; however, this conclusion is not supported by their research findings. Participants in a study aimed at describing the components of a "good death" said that the proper management of pain and other symptoms was important (Steinhauser, Clipp, McNeilly, Christakis, Mclntyre, & Tulsky, 2000). The study sample, however, was drawn primarily from in-patient settings and the findings derived mostly from health care providers' perspectives. In a comparative study examining perceptions of the good death, Payne, Langley-Evans, and Hillier (1996) discovered that patients and health care providers often differ in their conceptualizations of a good death. Therefore, study findings based on the perceptions of health care providers should be viewed with caution when applying them to patient and family situations. Access to Palliative Care Services Some studies also found that people who are more likely to spend their final days at home are supported by palliative care services (Brown, Davies, & Martens, 1990; Grande, et al., 1998). For example, Costantini, et al. (1993) found that the  38 proportion of cancer deaths at home were twice as frequent among users of palliative care home services (60.8%) than among nonusers (29.3%). Interviews with 80 caregivers following the death of a family member from cancer revealed that 90% of those dying at home had contact with home care nurses (HCNs) and 52% with palliative care specialist nurses (Addington-Hall, et al., 1991). Evaluative studies have also indicated that specialized home care services for the dying can promote home death. In an evaluation of a newly created community-based hospice rapid-response service, 82% of all referred patients were able to die at home (King, Mackenzie, Smith, & Clark, 2000). However, the study sample size was small (n=17) and no comparisons were provided. Therefore, it is difficult to ascertain whether this high proportion of home deaths was related to service implementation or other factors. Bruera and associates (1999) conducted a retrospective study comparing the pattern of care and site of deaths before establishment of a regional palliative care program in Edmonton and during its second year of operation. Post hoc comparisons showed that significantly more cancer-related deaths occurred in acute care hospitals prior to program implementation (86% versus 49%). However, this does not mean that dying was occurring at home. On careful examination, study findings demonstrated that increased numbers of patients were dying in in-patient hospices and that the percentage of patients dying at home had increased only 10% after initiation of the regional palliative care program. Similarly, a randomized control trial was conducted in the U.K. to determine if "hospital at home" facilitated dying at 22  IN THE U . K . , HOSPITAL AT H O M E WAS A PROGRAM SET UP WITH THE A I M OF IMPROVING THE PROVISION OF CARE FOR TERMINALLY ILL PATIENTS (PARTICULARLY AT NIGHT) AND TO INCREASE THEIR CHOICE OF PLACE OF CARE. HOSPITAL AT  39 home (Grande, Todd, Barclay, & Farquhar, 1999). Patients were randomized to a control group (offering standard care) and to a hospital at home group. The study did not find that hospital at home significantly increased the number of people dying at home. The researchers suggest that this might be because the study was conducted in a locality with a good provision of standard community-based palliative care. Hinton (1994a) conducted a prospective study to examine whether home care sufficiently addressed the comfort needs and adjustment of cancer patients and their caregiving relatives. A random sample of 77 adults and their relatives participated in the study with the ultimate aim of determining the success of home-based palliative care. Semi-structured interviews were administered with patients and relatives, each week over an eight-week period, and the Spritzer Quality of Life Index and several linear analogue scales were used to determine quality of life. Study findings indicated that community palliative care services can sufficiently meet the needs of patients and their caregiving relatives. However, they also suggest that perceptions about home care change over time as the patient's disease progresses. As patients become weaker and more debilitated, hospital care is favoured, particularly by family caregivers. The author concludes that, while home care services do much to enhance the comfort and adjustment of patients and their relatives, it should be complemented with inpatient care. Hinton's (1994b) follow-up study, examining cancer patients who are admitted to a hospice unit from home, concludes with an interesting discussion about the  HOME WAS AVAILABLE FOR TERMINAL PATIENTS WITH ANY DIAGNOSIS WHOSE PROGNOSIS WAS LESS THAN TWO WEEKS (GRANDE, ET AL., 1 9 9 9 ) .  40 feasibility of dying at home. The researcher asks whether there are practical limits to the amount of palliative care that can be provided in the home. These questions are raised because several caregivers in this study disclosed that they had considerable misgivings about home-based palliative care. Based on this research, the author argues that patients will likely continue to require inpatient care at several points throughout their disease trajectory unless full-time home nursing care is provided. The studies presented here indicate that access to home-based palliative care services may be an important component of facilitating the dying-at-home experience but it is unclear whether they increase the proportion of people who are able to die at home. A systematic literature review that synthesized 83 research papers on patient and caregiver preference for, and satisfaction with, specialized models of palliative care revealed that patients and caregivers were more satisfied with all types of palliative care, regardless of where it occurred (e.g., hospice or home) (Wilkinson, et al., 1999). At the same time, the reviewers argued that the evidence is too patchy to determine whether any particular model of care is preferred over the other and whether this has any influence on the location of dying. Furthermore, with the exception of Hinton's work (1994a; 1994b), very few of the studies detailed here took into account how family situations or the context of dying at home might influence the location of dying. Clearly, further research is required in this area to determine those factors that might influence the provision of palliative care at home, and to better understand how access to palliative care services may or may not enhance the ability for people to die at home.  The Nature of Support^  4  The nature of support provided to patients and family caregivers has also been found to be a factor that can influence the location of dying and the experience of family members providing end-of-life care at home. Many authors claim that having a team of skilled palliative care professionals can facilitate access to a range of services that are required to support home-based palliative care (Gomas, 1993; Mount, 1997; Roe, 1992). While these claims are not always supported by research findings, it is generally accepted that committed and skilled family physicians, HCNs, home support workers (HSWs), counsellors, and palliative care volunteers are essential to supporting family members (McCorkle, Robinson, Nuamah, Lev, & Benoliel, 1998; Moore, 1993; Wight, LeBlanc, & Aneshensel, 1995). Studies have found, for instance, that access to specialized nursing services can maintain dying people in their homes and significantly reduce psychological distress in caregivers (McCorkle, et al.). Maintaining contact, and providing explanations and support to caregivers, are reported to help as often as do physical treatments and practical nursing tasks in maintaining people at home (Hinton, 1996). Although patients and family caregivers access support from a variety of sources, research has demonstrated that the nature of support received can be problematic and can seriously inhibit caregivers' abilities to provide care at home. In a grounded theory study of family members providing home care to people with AIDS, Stajduhar (1995) found that the nature of support received greatly influenced  SUPPORT IS A CONCEPT THAT HAS BEEN USED LOOSELY IN HEALTH CARE LITERATURE TO DENOTE THE EXISTENCE, QUANTITY, AND TYPE OF INTERPERSONAL RELATIONSHIPS (NORBECK, LINDSEY, & CARRIERI, 1 9 8 1 ) , THE FUNCTIONAL CONTEXT OF THESE RELATIONSHIPS (KAHN, 1979), AND THE PERCEIVED QUALITY OF THIS SUPPORT (WEISS, CITED IN DIMOND & JONES, 1 9 8 3 ) . FOR M Y PURPOSES HERE, I A M USING THE TERM "SUPPORT" IN ITS BROADEST SENSE. 2 3  42 the caregiving experience. When caregivers received "good" support, they were better able to cope with their caregiving demands and feel a sense of personal worth in their role; they were able to carry on with their caregiving irrespective of the demands placed on them. But, when the nature of support provided by health care providers was unfavourable, caregivers felt isolated, angry, and disillusioned; they were more likely to have difficulties with caregiving at home. Similar studies have found that family members generate a passionate disrespect for the health care community when they are unsatisfied with the support they receive (Geis, Fuller, & Rush, 1986; Grief & Porembski, 1988). These unsupportive experiences have been linked to prolonged grief reactions following the death of a loved one (Broadhead, et al., 1983; Vachon, Sheldon, Lancee, Lyall, Rogers, & Freeman, 1982). It is not only family caregivers who have spoken of the difficulties that result from a lack of good support. In a study investigating the attitudes of 151 physicians and nurses towards cancer patients dying at home, Porta, Busquet and Jariod (1997) discovered that despite high motivation, physicians and nurses reported widespread frustration with the quality of care provided to terminally ill cancer patients. While the majority of participants reported "the system" as the main cause of frustration, they indicated that there are few appropriately skilled health care professionals in palliative care to support patients and family caregivers. Other studies are just beginning to reveal that families may be reluctant to access formal community-based health care services, due in part to the unsupportive nature of the services (Davis, et al., 1996; Grande, Todd, Barclay, & Doyle, 1996). However,  43 none of these studies provides detailed explanations of the context that might lead to unfavorable support. Family Caregiver Needs  Research studies, systematic literature reviews, and case reports on the needs of family caregivers - particularly those that address patient and caregiver needs during the cancer illness experience -- are abundant (Hampe, 1975; Hileman & Lackey, 1990; Hinds, 1985; Meissner, Anderson, & Odenkirchen, 1990; Steele & Fitch, 1996; Wingate & Lackey, 1989; Wright & Dyck, 1984). Some studies focus on the needs of patients and their family members in hospital settings, whereas others are directed at determining home care needs. The literature in this area is extensive and, while it is not my intent here to review the plethora of needs-based studies, I will provide a short synthesis of the studies focused on home care needs . 24  Almost two decades ago, Grobe and colleagues (1982) concluded that family members might have more needs that those who are dying. Caregiver needs have tended to be categorized in the literature as physical, psychosocial, and informational (Edlund, 1995). In an early study centered on the needs of families who care for patients with cancer at home, Hinds (1985) found that as many as 31% of all caregivers had difficulties coping with physical care, especially giving treatments and managing symptoms such as nausea and pain. Five years later, Brown and her colleagues (1990) found similar results in a qualitative study with family members who revealed the need for support services, such as personal care,  IT IS IMPORTANT TO UNDERSTAND THE NEEDS OF HOME CAREGIVERS IN PALLIATIVE CARE. STUDIES HAVE DEMONSTRATED THAT UNMET CAREGIVER NEEDS MIGHT INFLUENCE WHETHER DYING AT H O M E IS A REALISTIC OPTION (BECK-FRIIS & STRANG, 1 9 9 3 ) . 2 4  44 respite, and homemaking. These supports were especially important when there was only one caregiver in the home. Respondents emphasized the need for skilled health care providers to support symptom control and a need to learn practical caregiving skills. Later studies (e.g., Grande, et al., 1997; Silveira & Winstead-Fry, 1997; Steele & Fitch, 1996; Wyatt, et al., 1999) do not differ remarkably from these earlier findings, although Grande and colleagues found that the caregivers' need for assistance is sometimes in conflict with the patients' need to preserve independence. They recommended that caregivers' needs should be studied separately from patients' needs so that caregiver perspectives can be better understood with the aim of developing caregiver-focused interventions. Home caregivers also have a number of psychosocial needs and some studies have found that they are more important than both physical and informational needs (Hileman & Lackey, 1990; Wingate & Lackey, 1989). Wright and Dyck (1984) reported that caregivers need encouragement and assurance that they are doing a good job keeping their loved one comfortable; for many, this is a top priority (Cherney, Coyle, & Foley, 1994; Ferrell, Ferrell, Rhiner, & Grant, 1991; McGinnis, 1986; Stajduhar, 1995). Lev (1991) maintains that unless the comfort needs of patients are managed, family members will often neglect their own concerns. Yet, in a descriptive study of 20 family members providing care at home to cancer patients, Steele and Fitch (1996) found that caregivers reported the need to have time away for oneself as one of their most frequent needs. Communicating with patients and having an awareness of the dying process has also been found to be important for caregivers (Seale, 1991a), as has the need to develop trusting and  45 cooperative relationships with home health care providers (Andershed & Ternestedt, 1998; Cherney, Coyle, & Foley, 1994; Kristjanson, 1989; Wright & Dyck). The need for information is also an important component in facilitating the dying-at-home process. Lewandowski and Jones (1988) found that in the home setting, six of the 11 highest ranked nursing interventions involved nurses giving information to family members. Indeed, some studies have indicated that the more that family members understand the patient's care, the more satisfied they are with the care (Medigovich, Porock, Kirstjanson, & Smith, 1999). Conley and Burman (1997) conducted a qualitative study to determine the informational needs of homebased caregivers. Fourteen family members were interviewed and transcript data were subjected to constant comparative analysis. While the majority of caregivers indicated that they were satisfied with the information available to them, they often had difficulty in articulating what type of information was most needed. On careful examination, the researchers discovered that information on the disease process and on health care services that were available to them were most important in supporting home-based caregivers. Although caregivers in the study were vigilant about accessing information, the information was not always provided in a way that was most helpful. The authors recommended that written information must be reinforced with informal discussions. Steele and Fitch (1996) postulate that the complexity of health care systems may make it difficult for family caregivers to understand the services that are available, and suggest that information must be provided in a variety of ways to support them at home. Systems such as the U.S. Cancer Information Service have proved to be particularly helpful for caregivers,  46 although further research is required to determine the best ways to meet the informational needs of caregivers (Meissner, Anderson, & Odenkirchen, 1990). Preferences for Place of Dying and Death Dying at home is considered ideal when defining a good death (McNeil, 1998). In fact, with few exceptions, it is difficult to find any author who challenges this assumption. Some authors base their beliefs on their own personal experiences (Collett, 1997; Duda, 1987) while others draw on their professional knowledge of caring for patients with life-threatening illnesses (Bowling, 1983; Gomas, 1993; Moore, 1993; Thorpe, 1993). Health professionals who write about this topic invariably maintain that "death like birth, is a family affair" and should occur in the home, given that there is adequate support (Bowling, p. 158). Providing palliative care at home embraces what Gomas says "is most noble in medicine: sometimes curing, always relieving, supporting right to the end!" (p. 45). Research reports suggest that, in general, people would prefer to die at home (Dunlop, Davies, & Hockley, 1989; Gilbar & Steiner, 1996; Townsend, et al., 1990), and this preference is often supported by physicians and nurses (Cartwright, 1991; McWhinney & Stewart, 1994; Seamark, et al., 1995). In one of the first studies that looked at patient preferences for location of dying, Dunlop and colleagues recorded the preferred place of death of 160 patients who were consecutively referred to a hospital palliative support team in the U.K. The information was recorded during routine interviews with patients when they were asked where they wanted to be cared for during the final stages of their illness. All of the interviews took place while patients were in hospital and did not include any in-depth discussion about the  47 reasons for patient preferences, but simply recorded the patient's response. Overall, 56% of the study participants expressed a preference for dying at home. However, the researchers were able only to record patient preference on 90 of the 160 participants. In addition, although they mention that family members were considered in the assessment, no details are provided about family perspectives. In another prospective study of terminally ill cancer patients in the U.K., Townsend and others (1990) randomly sampled 84 patients to determine their preference for location of dying. A structured questionnaire was used and patients were asked where they would like to be if their disease progressed. Interviews took place more than once to assess changes in patient preferences. Of the 84 patients who were initially interviewed, 58% expressed a wish to be at home, 20% wished to be in hospital, 20% at a hospice, and 2% elsewhere. At the time of the final interview, 49% of patients wished to be at home and the remainder, either in hospital or hospice. The researchers also interviewed 30 caregivers three months after the death of their relative. Of the 11 whose relatives had died at home, seven were satisfied; three accepted the patient's wish to die at home although they thought they would have been better in hospital; and one would have preferred hospital but there had been no time to arrange for admission. Based on these findings, the researchers conclude that patients and their caregivers prefer palliative care to be provided at home. However, both this study and the study conducted by Dunlop and colleagues (1989) suffer from small sample sizes making it difficult to generalize patient preferences to other countries or patient populations.  48 Gilbar and Steiner's (1996) study of terminally ill cancer patients in Israel also concluded that the majority of patients (88.4%) preferred palliative care to be provided at home. Demographic, medical, and psychosocial variables gathered from 171 patient files were included in a multiple classification analysis. The data were collected throughout the period of illness and during interviews with family caregivers. The study reports on a number of factors that may influence the location of dying, factors which have been previously reviewed. While the study found that a large percentage of patients would prefer to remain at home for the last part of life, there were differences in patient and family preferences. Demographic variables were found to have the strongest association with patient preference, whereas family support systems were strongly associated with caregiver preference. Although no percentages were reported for caregiver preferences, when asked, they stated that they only wanted to fulfil the patient's wishes, even if these wishes did not always conform to their own. The researchers explain that these differences were likely due to the timing of the interviews and to the fact that patient preference data were collected by family caregivers who did the interviewing. The study report is somewhat confusing, however, as the conclusions reached by the researchers are not always supported by the data presented. Furthermore, having family members interview their dying relatives may have substantially biased the results and, thus, the results should be viewed cautiously. Acknowledging that the family unit is affected by the need for palliative care in one family member, Brown, Davies, and Martens (1990) conducted a qualitative study to find out from both family members and patients with advanced cancer which  49 location of care (i.e., home or hospital) they preferred and why. This Canadian study focused on the preferred place of care rather than the preferred place of death. Content analysis of in-depth interviews with eight families (24 members) revealed that patients and their families preferred care at home over care in the hospital or care in an institution, such as a nursing home. Moore (1993) also suggests that patients and their family members preferred home care to care in an institution; however, her contention is not supported by research findings. Other Canadian studies have focused on preferences for location of dying. McWhinney, Bass, and Orr (1995) conducted a retrospective case-control chart review on 150 patients referred to a home palliative care support program (75 patients who had died at home and 75 who had died in hospital). Of the 150 patient records that were reviewed, about 38% of patients expressed a strong preference for dying at home and 16% expressed a preference for dying at home conditional on the family's ability to cope. Conversely, an earlier study with 125 patients admitted to a palliative care unit pointed out that 90% of patients did not want to die at home (Bruera, Kuehn, Emery, MacMillan, & Hanson, 1990), citing a lack of family support and home care services. The patients in this study also had severe medical problems that were being treated in an acute care setting and this may have influenced their perspectives. Although both of the preceding studies allude to the importance of the family's perspective, inclusion of family members' preferences were not included in these studies. However, as Keizer, Kozak, and Scott (1992) found, patients and their caregivers often vary in their preferences for location of palliative care. They  50 conducted a retrospective study involving 45 randomly selected family caregivers who were known to the regional palliative care unit in Ottawa. Caregivers were interviewed up to 12 months following the death of a loved one and were asked about preferences for location of dying. Unexpectedly, the large majority (80%) of caregivers believed that inpatient palliative care offered the best quality of life and death, even for those patients who had resisted admission to the palliative care setting. Only 16% reported that the home was the preferred location for palliative care to occur and four percent indicated acute care as the preferred location. Many researchers and practitioners who write about the topic of home care for the dying state that most patients and their family members, if given a choice, would prefer that dying take place at home. These statements are surprising given that the research evidence in this area is both inconclusive and conflicting. Yet, many authors theorize or have found in research studies that there are specific benefits to patients and family members when palliative care is provided in the home setting. Most common is the belief that dying at home enhances the overall quality of life afforded the terminally ill and their family members (Roe, 1992). It is widely held, for example, that dying at home facilitates a sense of normalcy, providing comfort, security, intimacy, and familiarity to those coming to the end-of-life (Duda, 1987; Fraser, 1990). Research findings suggest that it promotes a sense of freedom and control that is not accorded in an institutional setting (Davies, et al., 1995). Being in the home environment also helps to sustain relationships with family and friends and contributes to reciprocity between the ill person and family members (Brown, et al., 1990).  51 Widespread criticisms of institutional-based care have also been linked to the belief that dying at home is preferable to dying in an institution. While attitudes toward death and dying among health care providers have changed considerably over the past 20 years, some authors contend that current hospital systems and those who work in them continue to value life above living productively (Cowley, et al., 1992; Gomas, 1993). As such, the predominant influence of biomedicine, with its focus on intervention and control of disease, has served to limit attention to the personal needs of the dying. Other less tangible needs (e.g., social, psychological, and spiritual) that are important to providing effective palliative care are often ignored in the hospital setting (Field & James, 1993). Even when attention to physical problems is required, the level of skill needed to control the symptoms associated with a terminal illness varies among general hospital staff (Ferrell, Eberts, McCaffery, & Grant, 1991; Hamilton & Edgar, 1992). Communication between disciplines and among patients, family members, and health care providers is also fraught with difficulties in the hospital system (Seale, 1991b). In a study of people dying from cancer and their family members, more than 50% of caregivers reported dissatisfaction with hospital care citing that health care providers were too busy to attend to the needs of the dying (Addington-Hall, Macdonald, Anderson, & Freeling, 1991). A comparative study of hospice and hospital care for people who die found that hospitals still have a long way to go in improving communications skills and the psychosocial climate, in spite of the provision of palliative care support teams in all of the hospitals in the study (Seale & Kelly, 1997). These difficulties  52 have led providers to advocate for better home care for the dying, as it may be better coordinated and inclusive of patients and family members (Moore, 1993). There is little debate that dying at home can significantly contribute to the overall quality of life of patients and their family members. Although there are many good reasons why patients and their caregivers might prefer palliative care to occur at home, the research evidence suggests that this may not always be the case and that preferences can change over time. With few exceptions, the studies that have been presented on patient and family preferences suffer from methodological flaws or small sample sizes that make the findings difficult to generalize. Even then, these studies are often used as benchmarks for other authors who uphold and make assumptions that patients and families prefer palliative care to be provided in the home setting. Many of the studies that have been conducted focus primarily on patient preferences or do not detail family perspectives. Yet, as Dudgeon and Kristjanson (1995) argue, the family's perspective is clearly important given the interactive nature of decisions regarding place of palliative care. While a patient's preference for dying at home is undoubtedly important, this decision has implications for family members who will provide the majority of intense physical and emotional support. The assessment of preferences is not a simple matter. "Whose point of view should be included in the assessment? How should preferences by measured? When and how often should an assessment be conducted?" (Dudgeon & Kristjanson, p. 338). These questions point to the need for an in-depth investigation aimed at gaining a more complete picture of the home setting in which palliative care occurs. While ensuring patients' preferences and increasing the numbers of people  53 dying at home might be seen as a laudable aim, it is possible that this aim might be in conflict with another stated goal of palliative care: that of supporting families and improving bereavement outcomes for surviving family members (Addington-Hall & Karlsen, 2000). Home-Based Palliative Care: Mapping Out the Macro Level Influences By almost all indicators, the Canadian health care system functions impressively in relation to most western states (Evans, 1986). Yet, in Canada, there has been little move to implement the policy recommendations put forward in relation to home care in general, and palliative home care specifically (Scott, 1992). While a fundamental principle of palliative care is that the patient and family together comprise the unit of care (CPCA, 1995), the complexity of our health care system has greatly influenced and, at times, interfered with the ability to provide appropriate home-based palliative care (Scott; Stajduhar & Davies, 1998b; Wodinsky, 1992). One could argue that the issue of home-based palliative care is embedded within the larger contexts of Canadian health care reform, most notably the home care and self-care movements . Indeed, in 1991, Anderson, Blue, and Lau argued that self25  care was one of the most powerful ideologies underscoring the organization of health care delivery services, and this likely continues to be true today . 26  Provincial and federal governments expound the benefits of self-care and home care (Armstrong & Armstrong, 1996). This is often based on taken-for-granted  S E E ANDERSON ( 1 9 9 0 ) FOR AN EXCELLENT OVERVIEW OF THE IDEOLOGIES UNDERPINNING THE H O M E CARE AND SELF-CARE MOVEMENTS. POLICY ADVANCES ARE CONTINUING TO SHIFT THE CONTEXT OF CARE PROVISION FROM HOSPITAL TO AMBULATORY AND COMMUNITY SETTINGS. HEALTH CARE POLICY IN CANADA AND B . C . , FOR EXAMPLE, HAS ESTABLISHED GOALS TO INCREASE SELF-CARE AND TO FOSTER APPROACHES THAT BUILD UPON PRINCIPLES OF MUTUAL AID AND COMMUNITY SUPPORT ( E P P , 1 9 8 6 ; SEATON, 1 9 9 1 ) . 2 6  54 assumptions that these movements are in the best interests of patients and family members, and are ultimately cost-effective (Anderson, 1990; Fierlbeck, 1997). In 27  writing about the status of home care for example, the Honourable Elizabeth Witmer (2000), Minister of Health and Long Term Care for the province of Ontario, stated that the increasing demands and pressures placed on the health care system are demands that can often best be met by delivering care in the home, both from an economic and humane point of view. Claims such as these, however, have been sharply challenged by researchers and health advocacy groups who claim that current home care proposals are aimed at off-loading responsibilities onto unpaid family members (Armstrong & Armstrong; Canadian Health Coalition, 1999). Bjorkman and Altenstetter (1997) suggest that one of the current aims of health care reform is to shift the burdens of health care financing from the public sector to private sources . Rarely, however, are the outcomes of such actions considered. 28  Furthermore, the factors involved in providing care in the home are frequently neglected when considering home care policy. In essence, the tension between what is occurring in health care reform, and what ought to be occurring, is what Storch (1996) calls a tension between the economics and ethics of care. While I was unable to find any research that specifically addresses the influence that Canadian health care reforms have had on family caregivers providing palliative care, there are issues that should be considered in contexualizing the  T H E FEDERAL LIBERAL GOVERNMENT HAS RECOGNIZED THE COST BURDEN THAT H O M E CARE M A Y ENTAIL FOR FAMILY M E M B E R S WHO ARE CARING FOR ELDERLY OR DISABLED LOVED ONES BY INSTITUTING A NEW CAREGIVER TAX CREDIT. THIS TAX CREDIT C A M E INTO EFFECT IN THE 1 9 9 8 FISCAL YEAR. T H E M A X I M U M PERSONAL AMOUNT OF THE CLAIM IS A $ 4 0 0 TAX CREDIT OR A COMBINED FEDERAL AND PROVINCIAL CREDIT OF ABOUT $ 6 0 0 PER YEAR FOR BRITISH COLUMBIANS. B Y THIS, I A M REFERRING TO THE SHIFT IN COSTS FROM THE STATE TO THE INDIVIDUAL OR FAMILY AND TO THE SHIFT IN PUBLIC FUNDS TO THE PRIVATE FOR-PROFIT SYSTEM. 2 7  2 8  55 home care picture. One of the primary issues is the way in which home care is administered across the country. Currently, Canada has a patchwork of home care programs run by the provinces and territories. Typically, programs are communitybased and administered by provincial health and/or social service departments or community/regional health boards. According to the Canadian Home Care Association (cited in Canadian Health Coalition, 1999), about 85% of home care is paid for by provincial governments with the remainder coming from user fees and third-party payers. As such, funding for home care takes up only a small part of Canada's overall health care budget. Because financing for home health care is not an integral part of the Canadian Medicare system (i.e., not a part of the Canada Health Act), there are large variations in provincial home care programs. National standards for home care are nonexistent, meaning that the services provided, the methods of payment, and the criteria for eligibility vary widely from one jurisdiction to another and from one province to another. Consequently, home care plays a secondary and largely supportive role to the primary health care system. This model, however, has significant implications for families providing palliative care at home. As the provision of palliative care continues to expand to the home, there is no obligation on the part of the provinces to ensure that home health care services are provided in accordance with the principles of the Canada Health Act . Armstrong and 29  Armstrong (1996) argue that the provinces are more likely to respond to the lure of  1 A M IMPLYING HERE THAT THE PRINCIPLES OF THE CANADA HEALTH ACT (I.E., PUBLIC ADMINISTRATION, PORTABILITY, ACCESSIBILITY, COMPREHENSIVENESS, AND UNIVERSALITY), THAT ARE MEANT TO ENSURE QUALITY HEALTH CARE TO ALL CANADIANS, ARE IN JEOPARDY WHEN PROGRAMS SUCH AS H O M E CARE ARE NOT PART OF A NATIONAL HEALTH CARE STRATEGY.  2 9  56 privatization to reduce fiscal responsibility. However, overlooking the potential outcomes of privatization is dangerous (Baumgart, 1988). An extensive study of Manitoba's home care program, for example, found that privatization resulted in lower wages, increased staff turnover, and an inclination to reduce staff training in order to maintain profit levels (Shapiro, 1996). These outcomes resulted in poorer and sometimes detrimental quality of service to clients and their family members, and did not produce substantial cost savings to the health care system. Whether home care will result in substantial savings to the health care system is a contentious issue (Arno, Bonuck, & Padgug, 1995; Whynes, 1997). Studies conducted in the U.S. have demonstrated that home care leads to higher spending and an over-reliance on family members (Weissert, 1991). A series of studies done by Weissert and others has gained such credence in the U.S. that requests for home care research proposals by the Robert Wood Johnson Foundation (1996) stated that, "the old rationale that increasing home care benefits pays for itself by keeping people out of nursing homes is no longer tenable, given research findings to the contrary" (p. 4). Canadian writers have begun to challenge the idea that home care is not cost-effective, at least compared to care in a hospital or long-term care facility (Hollander, 1994; Jacobs, Hall, Henderson, & Nichols, 1995; Saskatchewan Health Services Utilization & Research Commission, 1998). A large scale, multi-phase project examining the cost-effectiveness of home care in Canada is currently  57 underway to determine whether long-term-care substitution models are cost30  effective (Hollander, 1999). Preliminary findings have suggested that home care is a cost-effective alternative to long-term care, particularly for stable clients. However, as Hollander attests, "the issue of stability is important because it also reflects the converse, that home care may not be particularly cost-effective for those who change their type and level of care" (p. 104). According to the study's author, for those who are dying, home care may be more costly than care in an institutional setting. Whynes (1997) suggests that the home care costs associated with palliative care are not usually accurate because a large part of the burden falls upon family caregivers whose efforts are not usually costed in a formal sense. Studies conducted both in Canada and in European cities that have shown that home care is cost-effective, have done so in part because the 'hotel' costs are borne, not by the institution, but by the patients themselves or by their family members (Ferris, et al., 1991; Beck-Friis, Norberg, & Strang, 1991). Even though Hollander's (1999) study demonstrated that home care costs for the dying may be higher than that of longterm care costs, his study did not account for costs incurred by family caregivers in palliative care. In sum, it is unclear whether home care will result in substantial savings to the health care system and, overall, there is a lack of convincing scientific evidence to  IN CANADA, THERE ARE THREE MAJOR MODELS OF HOME CARE: (A) AN ACUTE-CARE SUBSTITUTION MODEL, WHERE HOME CARE MEETS THE NEEDS OF PEOPLE WHO WOULD OTHERWISE BY CARED FOR IN ACUTE CARE FACILITIES; (B) A LONG-TERM CARE SUBSTITUTION MODEL, WHERE H O M E CARE MEETS THE NEEDS OF PEOPLE WHO WOULD OTHERWISE REQUIRE INSTITUTIONALIZATION; AND, (C) THE MAINTENANCE AND PREVENTATIVE MODEL, WHICH SERVES PEOPLE WHO ARE ABLE TO LIVE AT HOME, BUT WHO HAVE FUNCTIONAL DEFICITS AND M A Y NEED SUPPORT TO CONTINUE TO LIVE INDEPENDENTLY.  58 suggest precisely when home care is cost-effective (Arno, Bonuck, & Padgug, 1995). Clearly further research is required in this area in order to make meaningful assessments of cost-effectiveness and the relative well being of patients who are hospitalized and those who are receiving home care. Summary of the State of Knowledge The shortcomings of our health care system in meeting the needs to support people dying at home have been echoed across the country (Dudgeon & Kristjanson, 1995; Johnson, 1995; Roe, 1992; Scott, 1992). However, it is only recently that there has been widespread public and government recognition of the important role that caregivers play in supporting home-based care for the dying. It is clear that a significant number of Canadians are actively engaged in providing care at home. The need for home care will increase substantially over the next several years as our health care system struggles to face demographic changes and demands resulting from a growing number of people with life-threatening illnesses who are in need of care. While it is generally assumed that home-based palliative care is preferred, the research evidence does not always support this assumption. The research does suggest, however, that there are benefits to patients and their family members when palliative care is provided in the home setting. This research has greatly contributed to our understanding of the dying-at-home experience but, with few exceptions, this research provides only a cursory overview of the situation for family members when care is provided in the home setting. Most studies focus on death distribution patterns, service utilization, and demographic characteristics of home care patients.  59 In general, these studies have relied on quantitative measurements to explain patient and family experiences. However instructive, these studies neither provide the depth nor breadth of knowledge needed to fully understand the context of caregiving at home. It has been argued that the quantitative emphasis in the body of family caregiving research has constrained our understanding of the multiple ways in which the experience of family caregiving is important to family members (Kellett & Mannion, 1999). I was able to find only one study, an ethnography, that involved observing patients and families in their natural settings (i.e., home), but this study focused on developing a conceptual framework for describing dying at home as a status passage and did not examine the family caregiving experience (Wilson, 1989). Indeed, there are very few studies that have focused exclusively on home care for the dying and much of the evidence that is available arises from studies done in the U.K., and to a lesser extent, the U.S. and Australia. Although there is a growing number of Canadian studies that have helped to "set the stage" for my current project, they also do not focus on some of the larger issues that may influence the caregiving experience at home. Most of the studies that have been done, regardless of the country of origin, have focused on individual experiences in isolation of the larger social contexts that may influence these experiences. Although the research conducted to date provides an important beginning to understanding home-based palliative care, the dialectic between the social context of dying at home and family caregiving has failed to attract much analysis or systematic inquiry. Yet, palliative caregiving at home is a complex social  phenomenon that may improve life for many while, for others, may threaten to erode the conditions that tend to foster important social goods and opportunities. Indeed, the threat that such care poses to the conditions of a robust home life ought to inform how health care decisions and policies are made. Thus, this study was undertaken with the goals of enhancing our understanding of what Thome (1993) calls the "grand and small pictures" of caregivers' experiences. Through the explorations that follow, I will endeavour to provide a contextualized portrait of the experiences of those who care for people who are dying at home.  61 CHAPTER THREE: RESEARCH DESIGN AND IMPLEMENTATION A survey of the empirical and theoretical literature directed me to explore a range of possible methodologies to address the research problem. One of my central concerns in conceptualizing this study was the lack of attention given to the broader social contexts in which home-based palliative care occurs. As previously mentioned, much of the current knowledge has described caregiver experiences in isolation of how they are shaped and influenced by larger social structures and processes. In this study, I wanted to situate caregiver experiences within a broader social context in order to gain a deeper understanding of the provision of palliative care in the home setting. My secondary aim was to move beyond individual descriptions in order to create understandings that might influence how health care decisions and policies are made in the area of home care for the dying. Addressing these aims meant that I required a methodology that would allow me to understand caregivers' experiences within the context of the whole. In other words, I needed a methodology that would direct me to probe the intersecting elements of the home experience in order to create a multi-layered account of the social context of dying at home in relation to home-based family caregiving. As such, I conducted an ethnography informed by critical perspectives. In this chapter, I elaborate on the actual "steps" that I took in carrying out this research. I begin by exploring ethnographic and critical perspectives, explaining how I combined the two to provide theoretical and methodological guidance for the study. I then describe the more "practical" aspects of this project as I detail how  62 participants became involved in the study and how I negotiated entry into observation sites. I go on to present the ways in which I collected information as the study was implemented, commenting on interviewing, participant observation, writing field notes, and reviewing documents. The approach to data analysis is outlined and I conclude with a discussion of the strategies that I employed to enhance scientific integrity and maintain ethical standards. The Intersection of Ethnographic and Critical Perspectives Ethnographic Characteristics Scrutinizing the ethnographic terrain quickly became perplexing as I endeavoured to wade through the sometimes ambiguous and paradoxical positions put forward by ethnographic scholars. At this time in history, ethnography is at a crossroads and, as Denzin (1997) argues, "it is no longer possible to take for granted what is meant by ethnography" (p. xiii). Ethnography has had a long and controversial tradition in the social sciences and has evolved over the past century as philosophies of science have developed and various disciplines have adopted and adapted ethnographic techniques (Hammersley & Atkinson, 1995). Founded on realist beliefs that offer authoritative accounts of the processes being examined (Malinowski, 1922/1961), ethnography has grown to include such diverse streams as ethnographic feminisms (Behar & Gordon, 1995; Cole & Phillips, 1996; Visweswaran 1994), autoethnography and ethnographic short stories (Ellis, 1995; Ellis & Bochner, 2000), ethnographic poetics (Marcus & Fischer, 1986), and performance texts (Richardson, 1993). Such developments arose from discontent with classic realist ethnographies. The classic tradition has been sharply criticized, mainly because it  63 continues to betray the influence of positivism and scientism, and claims to be value neutral (Anderson, 1989; Carspecken, 1996; Marcus, 1998; Quantz, 1992). Given the various forms of ethnographic research, no universally adopted definition of ethnography exists (Boyle, 1994; Stewart, 1998; Wolcott, 1992; Wolcott, 1999). Some refer to ethnography as a philosophical paradigm requiring a total commitment to its tenets, whereas others define it as a method to be used when and where appropriate (Atkinson & Hammersley, 1994). At its core, however, ethnography is a theoretically-based methodology used for constructing knowledge aimed at the study of social contexts (Hammersley & Atkinson, 1995). Even so, this delineation was not particularly instructive in helping me to understand the "gist" of ethnography as a research methodology. Atkinson and Hammersley provided some guidance in this area by highlighting that ethnography has a number of the following features: 1. a strong emphasis on exploring the nature of particular social phenomena, rather than setting out to test hypotheses about them; 2. a tendency to work primarily with "unstructured" data, that is, data that have not been coded at the point of data collection in terms of a closed set of analytic categories; 3. investigation of a small number of cases, perhaps just one case, in detail; and 4. analysis of data that involves explicit interpretation of the meanings and functions of human actions, the product of which mainly takes the form of verbal descriptions and explanations, with quantification and statistical analysis playing a subordinate role at most (p. 248).  64 Certainly, many of these features characterize my current project in that I was exploring a social circumstance and working with unstructured data to interpret verbal descriptions and explanations. Yet, these features alone did not seem entirely different from other research methodologies in the qualitative domain . I 31  took further guidance from Stewart (1998) who proposes that ethnography has at least four widely accepted characteristics . First, ethnography involves participation 32  in the "everyday" setting to be studied. Therefore, immersion in the 33  GROUNDED THEORY AND ETHNOGRAPHY, FOR EXAMPLE, SHARE S O M E COMMONALITIES. AGAR ( 1 9 8 6 ) CLAIMS THAT BOTH USE SIMILAR PROCESSES FOR GENERATING UNDERSTANDINGS BY COMPARING DATA AND THEORY. FURTHERMORE, "ETHNOGRAPHIC THEORY IS 'GROUNDED', AS ETHNOGRAPHERS PRIORITIZE TRYING TO THINK THROUGH OBSERVATIONS IN TERMS OF THEIR IMMANENT CONTEXT, BEFORE TRYING TO CONSTRUE THE DATA IN THE LIGHT OF EXISTING THEORIES" (STEWART, 1998, P. 9). NEVERTHELESS, GROUNDED THEORY AND ETHNOGRAPHY DIFFER IN AT LEAST THREE IMPORTANT WAYS. UNLIKE ETHNOGRAPHY, THE METHODOLOGICAL PROCEDURES FOR GROUNDED THEORY ARE DOCUMENTED IN DETAIL (FOR EXAMPLE, CHENITZ & SWANSON, 1986; STRAUSS & CORBIN, 1 9 9 0 ) . FURTHER, GROUNDED THEORY DOES NOT REQUIRE THE S A M E TYPE OF IMMERSION IN THE FIELD SETTING THAT ETHNOGRAPHY DOES. A S STRAUSS ( 1 9 8 7 ) ASSERTED, "ONE DOESN'T HAVE TO SPEND A LOT OF TIME AND ENERGY IN THE TYPICAL FIELDWORKER'S FASHION, BECAUSE THEORETICAL SAMPLING [FOR GROUNDED THEORY] ALLOWS FOR MORE EFFICIENT, SHORT-TIME OBSERVATION AND INTERVIEWING" (P. 275). THIS DISTINCTION LIES IN THE CLAIM THAT THE main FOCUS OF GROUNDED THEORISTS IS ON THE DEVELOPMENT OF CONCEPTS FOR THEORY DEVELOPMENT RATHER THAN, IN ETHNOGRAPHY, WHERE THE FOCUS IS ON THE CONTEXT OF SOCIAL SITUATIONS (STEWART). UNDERSTANDING THIS CONTEXT REQUIRES IMMERSION IN THE FIELD SETTING. FINALLY, AND IN A SIMILAR VEIN, THE "DISCOVERY" AND MODIFICATION OF THEORY IS THE PRIMARY GOAL OF GROUNDED THEORY (STRAUSS). IN ETHNOGRAPHY, HOWEVER, THEORY DOES NOT SIMPLY AWAIT REFINEMENT AS ANALYSTS TEST CONCEPTS ONE BY ONE AGAINST EVENTS IN THE SOCIAL WORLD ... RATHER, THE ETHNOGRAPHER'S ASSUMPTIONS, INTERESTS, AND THEORETICAL COMMITMENTS ENTER INTO EVERY PHASE OF WRITING AN ETHNOGRAPHY AND INFLUENCE DECISIONS THAT RANGE FROM SELECTING WHICH EVENTS TO WRITE ABOUT TO THOSE THAT ENTAIL EMPHASIZING ONE MEMBER'S PERSPECTIVE ON AN EVENT OVER THOSE OF OTHERS. T H E PROCESS IS THUS ONE OF REFLEXIVE OR DIALECTICAL INTERPLAY BETWEEN THEORY AND DATA WHEREBY THEORY ENTERS IN AT EVERY POINT, SHAPING NOT ONLY ANALYSIS BUT HOW SOCIAL EVENTS C O M E TO B E PERCEIVED AND WRITTEN UP AS DATA IN THE FIRST PLACE (EMERSON, FRETZ, & SHAW, 1995, P. 167). THEORY MATTERS IN ETHNOGRAPHY, THEN, BUT ETHNOGRAPHY IS NOT FOCUSED ON THEORY OR CONCEPTS AS IN GROUNDED THEORY (STEWART). STEWART ACKNOWLEDGES THAT ANY ATTEMPT TO DEFINE ETHNOGRAPHY RUNS THE RISK OF BEING ESSENTIALIST AS IT IS DIFFICULT TO ASSIGN A PARTICULAR STUDY TO A CATEGORY OR TAXON. H E ALSO ADMITS THAT MORE SPECIALIZED FORMS OF ETHNOGRAPHY WILL HAVE THEIR OWN GUIDING ELEMENTS. NEVERTHELESS, I DID FIND STEWART'S ARTICULATION TO B E HELPFUL IN CONCEPTUALIZING THIS STUDY BUT C A M E TO S E E THAT ETHNOGRAPHY M A Y NOT BE "DEFINED" IN THE CONVENTIONAL SENSE BUT, RATHER, COULD B E VIEWED AS MORE OF A MODE OF RESEARCHING AND PRESENTING KNOWLEDGE. 2  WHILE THE AIMS OF ETHNOGRAPHIC FIELDWORK ARE SIMILAR FOR ANTHROPOLOGISTS AND SOCIOLOGISTS, V A N MAANEN ( 1 9 8 8 ) ASSERTS THAT THEY DIFFER IN THEIR VIEW OF ETHNOGRAPHIC FIELD IMMERSION. ETHNOGRAPHIES ARISING FROM THE ANTHROPOLOGICAL TRADITIONS TYPICALLY GO ELSEWHERE TO STUDY CULTURE. IN CONTRAST, ETHNOGRAPHIES ARISING FROM SOCIOLOGICAL TRADITIONS USUALLY STAY "CLOSE TO H O M E " (P. 2 1 ) AND THE CULTURE OF INTEREST IS OFTEN FAMILIAR TO THE RESEARCHER. BASED ON V A N MAANEN'S CONTENTION, I WOULD S E E M Y WORK AS ALIGNED WITH SOCIOLOGICAL TRADITIONS IN THAT I STAYED WITHIN A LOCALE THAT WAS KNOWN TO M E AND 3 3  65 field through participant observation becomes a primary instrument (though not the only instrument) for data collection. Second, ethnography has a holistic mandate. As such, individual level experiences are understood as being a part of a larger whole and the range of attention to elements in the setting is comprehensive. In other words, the data obtained through ethnographic inquiry has "breadth" (Becker, 1996, p. 65). Third, ethnography involves providing a contextualized explanation for social events. Similar to the holistic nature of ethnography, Agar (1986) maintains that ethnographers seek to connect their observations to each other and then to a broader social context. Lastly, ethnography is concerned with describing, in a detailed way, the point of view of the main actors in a particular culture . 34  As I understand it then, operationalizing ethnography involves a period of immersion in the field setting in order to better understand the emic or "insider's" point of view (Spradley, 1979). Working from data in the form of observations, field notes, interviews, and other documents, such as policy statements and records (Germain, 1986; Morse & Field, 1995), the ethnographer seeks to understand a particular "culture" within a broader social context . Typically, ethnography involves 35  "thick description" whereby the analysis moves beyond simple descriptions to  STUDIED A CULTURE THAT I WAS FAMILIAR WITH. NONETHELESS, I ACKNOWLEDGE THAT I DREW HEAVILY FROM BOTH TRADITIONS IN SEEKING GUIDANCE FOR THIS STUDY. IN ETHNOGRAPHY, CULTURE CAN BE USED IN THE BROADEST SENSE (MORSE & FIELD, 1 9 9 5 ) FROM EXAMINING THE BELIEFS OF ETHNIC GROUPS (FOR EXAMPLE, LIPSON, 1 9 9 1 ) TO INVESTIGATING BELIEFS AND PRACTICES OF PARTICULAR COMMUNITIES, SUCH AS THE OPERATING ROOM (FOR EXAMPLE, FISHER & PETERSON, 1 9 9 3 ) TO INQUIRING ABOUT GROUPS OF INDIVIDUALS THAT SHARE A COMMON EXPERIENCE (FOR EXAMPLE, HAGGSTROM, AXELSSON, & NORBERG, 1 9 9 4 ) . I RECOGNIZE THAT THIS MIGHT B E CONSTRUED AS AN OVERSIMPLIFIED VIEW OF ETHNOGRAPHY. I ONLY USE IT HEURISTICALLY TO OUTLINE S O M E OF THE MAIN FEATURES OF ETHNOGRAPHY. 3 4  3 5  66 explain the meanings of behavior and the cultural context in which it occurs (Geertz, 1973). All of these features were included as a part of this study. Critical Perspectives While I found that literature about ethnography was helpful in conceptualizing this study, I came to realize that ethnography alone was insufficient to direct me to examine and critique how social and political structures and processes shape caregivers' experiences; it also did not provide the philosophical and theoretical guidance that I believe is important to the scientific process. Indeed, it is now widely held that the worldview or paradigm that one brings to the research enterprise is an 36  integral part of sound scientific inquiry. More specifically, a researcher's ontology, epistemology, and methodology influence all phases of the research process from conception to completion (Anderson, 1991). Denzin and Lincoln (1994) assert that ontology, epistemology, and methodology are interconnected elements, which not only influence the research process, but also reflect the encompassing personal biography of the researcher. As such, an understanding of the philosophical and theoretical base from which inquiry paradigms have arisen is central to engaging in thoughtful, reflective research. Thinking through my own commitments to the research enterprise helped to clarify my philosophical and theoretical positioning. I found Guba and Lincoln's (1994) work to be particularly useful in making distinctions about the various inquiry paradigms. Guba and Lincoln use the term critical theory to denote a set of  FOR THE PURPOSE OF THIS STUDY, I UNDERSTAND A PARADIGM TO B E "A BASIC BELIEF SYSTEM OR WORLDVIEW THAT GUIDES THE INVESTIGATOR, NOT ONLY IN CHOICES OF METHOD BUT IN ONTOLOGICALLY AND EPISTEMOLOGICALLY FUNDAMENTAL WAYS" ( G U B A & LINCOLN, P. 105).  67 alternative paradigms including feminisms . Using ontologic, epistemologic, and methodologic questions to guide my thinking, I came to understand that critical 38  theoretical perspectives could provide direction for my methodology. Thus, the study design that I employed is an ethnography informed by critical theoretical perspectives . Accordingly, I proceeded from a moderate realist position (the 39  ontological question), believing that reality is historically and socially created, and is influenced by social, political, historical, economic, cultural, and gender values. Thus, local meanings are embedded within larger societal meanings. Epistemologically, I hold that knowledge is socially constituted, historically situated, and valuationally based; it is intersubjective and transactional, and, therefore, both the researcher and the researched shape what knowledge is constructed (Cook & Fonow, 1990; Henderson, 1995; Rutledge-Shields & Dervin, 1993). Yet, I concur with Reimer Kirkham (2000) in that I do not believe that "all knowledge is local or subjective -- there are certain shared realities and common meanings that transcend the individual and the context" (p. 86). Methodologically, I take a dialogic and  APPROACHES TO ETHNOGRAPHIC RESEARCH INFORMED BY FEMINISMS ARE CONCERNED PRIMARILY WITH GENDER INEQUITIES AND HAVE BOTH CONTRIBUTED TO (ANDERSON, 1989) AND BEEN SEEN AS CONTRADICTORY TO CRITICAL APPROACHES (STACEY, 1988; STRATHERN, 1987). I HAVE CHOSEN TO USE G U B A AND LINCOLN'S ( 1 9 9 4 ) CONCEPTUALIZATION OF CRITICAL THEORY AS ENCOMPASSING FEMINIST PERSPECTIVES AS AN OVERARCHING FRAMEWORK FOR DISCUSSION AND UNDERSTOOD THAT AN ETHNOGRAPHY INFORMED BY CRITICAL PERSPECTIVES WOULD ALLOW M E TO CONDUCT AN IN-DEPTH ANALYSIS OF THE SOCIAL CONTEXT WITHOUT DIMINISHING OR IGNORING GENDER ISSUES (CAMPBELL & BUNTING, 1991). T H E ONTOLOGICAL QUESTIONS (I.E., WHAT IS THE NATURE OF REALITY? WHAT CAN B E KNOWN ABOUT REALITY?), EPISTEMOLOGICAL QUESTIONS (I.E., WHAT IS THE RELATIONSHIP BETWEEN THE KNOWER AND THE WOULD-BEKNOWER? WHAT CAN B E KNOWN?), AND METHODOLOGICAL QUESTIONS (I.E., HOW CAN THE WOULD-BE-KNOWER GO ABOUT FINDING OUT WHATEVER SHE/HE BELIEVES TO B E KNOW?), HIGHLIGHTED BY G U B A AND LINCOLN (1994), WERE PARTICULARLY HELPFUL IN GUIDING M E TO ARTICULATE M Y THEORETICAL POSITIONING. 1 A M DELIBERATELY MAKING A DISTINCTION BETWEEN CRITICAL ETHNOGRAPHY AND ETHNOGRAPHY INFORMED BY CRITICAL THEORETICAL PERSPECTIVES. A S I LATER ARGUE, THE EMANCIPATORY AIMS IMBUED IN CRITICAL SOCIAL THEORY ARE PROBLEMATIC FOR THIS PROJECT. INSTEAD OF ALIGNING MYSELF WITH THE OVERT IDEOLOGICAL GOALS OF NEO-MARXIST CRITICAL ETHNOGRAPHY OR THE CRITICAL SOCIAL ACTION PROJECTS ADVOCATED BY S O M E CRITICAL ETHNOGRAPHERS (CARSPECKEN, 1996; LATHER, 1 9 8 6 ) , I CHOSE TO FOCUS ON THE PHILOSOPHICAL AND THEORETIC DIRECTIONS THAT CRITICAL PERSPECTIVES MIGHT OFFER FOR M Y CURRENT PROJECT. 3 8  3 9  68 dialectical approach in seeking what there is to be known. Thus, while I believe that relationships between the researcher and the researched are inevitably permeated with unequal power differentials, I believe that through the process of reflexivity, these differences can be minimized to create nonhierarchical, free, uncoerced, and undistorted communications between the researcher and researched. Having articulated my critical "interpretive lens", I then considered those assumptions that underpin critical theoretical perspectives. This examination proved difficult as critical science has many forms (e.g., critical theory, critical social science, new paradigm research, praxis-oriented research, critical inquiry, emancipatory social theory, emancipatory research, and so on). I took direction from a number of sources to "piece together" those elements that were instructive for my intended purposes, although I did so with full awareness that some critical scholars might criticize this approach. Nevertheless, I chose these perspectives for what they offered, taking a pragmatic approach to decision making. As previously mentioned, I wanted to situate caregiver experiences within a broader social context and move beyond individual description to develop knowledge with which to influence health care decision making and policy development in the area of home care for the dying. This meant that I would need to examine and critique some of the larger social, political, cultural, and historical structures and processes that shape caregivers' experiences. These aims imply that I am seeking not only to describe caregiver experiences, but also to develop knowledge to influence change . Thomas (1993) maintains that ethnographers 40  RESEARCH INFORMED BY CRITICAL PERSPECTIVES IS MORE LIKELY TO BE PRAXIS-ORIENTED (LATHER, 1 9 8 6 , 1 9 9 1 ; MILLER, 1 9 9 7 ) , AIMING TO UNCOVER HEGEMONIC AND RESTRICTIVE HEALTH CARE PRACTICES. THUS, INDIVIDUAL  69 working within the critical paradigm study social structures and processes not only to describe them, but also to critique and change them. As such, ethnography informed by critical perspectives is "conventional" ethnography with a political purpose (Thomas). Thus, the critical part of the ethnography pushes the analysis beyond mere description of a particular experience, situation, or process to make sense of it in relation to both the broader social and organizational context (Bent, 1993; Forester, 1993; Wells, 1997). Since one of my explicit intentions was to create knowledge that may improve care for family members in the future, I believed that using knowledge derived from this study may be able to correct inequities, thereby contributing to the greater social good. As Rabinow and Sullivan (1987) point out, "the great strength of... critical theory has been continually to urge that the human sciences cannot be detached from the greater problems of living" (p. 15). Certainly, many nursing scholars have articulated the need for nursing science to be directed at generating knowledge to meet its social and moral mandates (Chopoorian, 1986; Moccia, 1988a; Starzomski & Rodney, 1997; Thorne, Canam, Dahinten, Hall, Henderson, & Reimer Kirkham, 1998; Williams, 1991). While I recognized the benefits of using critical theoretical perspectives to guide my interpretations, I was concerned about the focus on oppression and the emancipatory intents inherent in critical social orientations . Although there is utility 41  SITUATIONS CAN B E TRANSCENDED TO IDENTIFY LARGER SOCIETAL STRUCTURES AND IDEOLOGIES THAT MAINTAIN INEFFECTIVE AND UNEQUAL HEALTH CARE PRACTICES. A S SUCH, A HALLMARK OF ETHNOGRAPHY INFORMED BY CRITICAL PERSPECTIVES IS ITS EMPHASIS ON DEVELOPING KNOWLEDGE FOR CHANGE (CARSPECKEN, 1 9 9 6 ) . O N E OF THE OVERALL GOALS OF CRITICAL SOCIAL THEORY IS TO LIBERATE INDIVIDUALS FROM CONSCIOUS AND UNCONSCIOUS CONSTRAINTS BY NULLIFYING THE EFFECTS OF DOGMA AND IDEOLOGY (ALLEN, 1 9 9 0 ; BRENT, 1993; WILSON-THOMAS, 1 9 9 5 ) . PEOPLE'S PERCEPTIONS AND EXPERIENCES ARE THUS FREED OR EMANCIPATED AND THEY ARE ABLE TO S E E BEYOND THE TAKEN-FOR-GRANTED REALITIES THAT OPPRESS THEM ( R A Y , 1 9 9 2 ) . A S SUCH, THE FOCUS OF CRITICAL THEORY IS TO LIBERATE PEOPLE FROM UNACKNOWLEDGED CIRCUMSTANCES OF DOMINATION AND TO TRANSFORM CONSTRAINING CONDITIONS (STEVENS, 1 9 8 9 ) . T H E EMPHASIS IS ON ELIMINATING FALSE  4 1  70 to the emancipatory aims of critical social theory (see, for example, Brown, 2000; McCormick & Roussy, 1997; Thompson, 1987; Thorne, 1997), I was not intending to conduct a critical social action project nor was I intending to free individuals from oppressive conditions. A significant concern of mine regarded how oppression is defined and what constitutes an oppressive condition. If a goal of critical social theory is to enlighten agents about their oppression so that they can realize their own best interests, how do we best go about doing this in a way that is not paternalistic? Campbell and Bunting (1991) warn that, "the assumption can be made that agents (the oppressed) begin the process as less emancipated than the theorists" (p. 6) and suggest that this stance can result in a paternalistic position. The goal of enlightenment seemed antithetical to what I was attempting to achieve. That is, there are inherent dangers and implications in labeling people as "oppressed" since this label may influence how people are perceived and how others relate to them. The notion of false consciousness was also problematic. While perhaps appealing from a theoretical perspective, false consciousness undermines subjective knowledge as valuable and legitimate (Brown, 2000). For example, do we support caregivers in their use of avoidance to cope with the demands of caregiving or do we characterize it as false consciousness? Characterizing this experience as false consciousness has the potential to minimize caregivers' experiences and dismiss their legitimate ways of coping. I concur with Brown who suggests that,"... nursing CONSCIOUSNESS IN ORDER TO REDRESS POWER IMBALANCES AND END OPPRESSION (FREIRE, 1 9 9 0 ) . EMANCIPATION IS A CENTRAL CONCEPT FOR CRITICAL THEORISTS WHO ARGUE THAT BELIEF SYSTEMS THAT ARE PRESENT AND TREATED AS FACTS BY THE RULING CLASS, ACT AS BARRIERS TO CONSCIOUS ACTION AND FREEDOM (ACKER, BARRY, & ESSEVELD, 1 9 8 3 ; SCHWANDT, 1 9 9 0 ) .  71 science may benefit most from using false consciousness as a self-reflection strategy but refrain from applying the concept to individual patient-care situations" (p. 48). In this study, I aimed to critique larger structures and processes in order to develop knowledge to influence change. However, I did not propose to emancipate family caregivers. The emancipatory aims imbued in critical social action approaches are problematic for family members in palliative care . Family 42  members are already overwhelmed in their caregiving roles. Engaging them in dialogue to "free them from oppression" would be inconsistent with the mandate of palliative care. Similarly, caregivers might be challenged to fully participate in the research process (a requirement for many critical social action projects) without adding to their already substantial burden of caregiving. As Thorne and Varcoe (1998) argue, "focusing emancipatory intent on the participants of the research itself rather than on the larger objective of the research ... fails to recognize the social inequities that might render full participation impossible..." (p. 488). Furthermore, overreliance on critical social theory may privilege the collective over the individual  STUDIES DERIVED FROM CRITICAL SOCIAL THEORY OFFER THE POSSIBILITY THAT PARTICIPANTS WILL BE EMANCIPATED. EMANCIPATION M A Y M E A N ENGAGING WITH PEOPLE IN CRITICAL THOUGHT AND ANALYSIS OR ENABLING THEM TO REFLECT ON THEIR SITUATIONS AND GAIN NEW MEANINGS FROM THEIR EXPERIENCES. HOWEVER, I THOUGHT THAT THESE AIMS PRESENTED DIFFICULTIES FOR THE POPULATION I WAS STUDYING. WHILE S O M E MIGHT ARGUE THAT THE CAREGIVING EXPERIENCE ITSELF IS OPPRESSIVE, AND THAT CAREGIVERS ARE IN NEED OF BEING "TRANSFORMED" AND EMANCIPATED, M A N Y CAREGIVERS HAVE AN INTENSE EMOTIONAL INVESTMENT IN THE CARE OF THE DYING. ENGAGING CAREGIVERS IN A PROCESS TO CRITICALLY EXAMINE THEIR POSITION AS A CAREGIVER AND THE POSSIBLE INEQUITIES THAT THEY M A Y FACE OR M A Y HAVE FACED WOULD INEVITABLY BRING FORTH PAINFUL ISSUES THAT MIGHT BE BEST LEFT ALONE (AT LEAST FROM A RESEARCH PERSPECTIVE). I WAS NOT PREPARED TO INSTIGATE AN EXPRESSION OF DEEPLY E M B E D D E D FEELINGS WITH THE PARTICIPANTS IN THIS STUDY. THIS DECISION WAS INFORMED BY M Y EXTENSIVE CLINICAL EXPERIENCE IN PALLIATIVE CARE AND SUPPORTED BY O P I E ( 1 9 9 2 ) WHO EXPLAINS THAT "THE TENSIONS ARISING FROM THE DUALITY OF POSITIONING (WILLING/EXPLOITED) HAS TO B E SIMULTANEOUSLY MAINTAINED SO THAT WOMEN (AND M E N ) WHO DERIVE SIGNIFICANT POSITIVE AFFECTIONAL AND PERSONAL MEANING FROM THEIR ROLE ARE NOT DEFINED AS IMBUED WITH FALSE CONSCIOUSNESS" (P. 55).  72 (Brown, 2000) and can "permit representative appropriation of voice by more powerful members of a community over others in a misguided attempt to avoid researcher misuse of power" (Thorne & Varcoe, p. 488). At the same time, I did not want to reject altogether the idea that emancipation could be a useful concept to direct my thinking. Therefore, instead of expecting change to occur based on the research process, I came to understand that emancipation could direct me to the possibilities of change (Lather, 1991). I saw these possibilities for change, however, at a systems level rather than at an individual one. I also came to believe that the knowledge developed through the interpretive process was emancipatory in itself (Thorne & Varcoe, 1998) as the knowledge derived from the study findings could be used to create an expanded repertoire of options for supporting caregivers. In summary, ethnography informed by critical perspectives provided the substantive guidance for studying the social context of home-based palliative caregiving. This methodology directed me to: 1. Use a range of data collection strategies, including participant observations, interviewing, field notes, and a review of documents; 2. Situate caregivers' experiences within a broader context, taking into account how they were shaped by social, political, historical, economic, cultural, and gender values and contexts; 3. Intersubjectively construct knowledge with the main actors in the research process;  73 4. Engage in a reflexive process, always paying attention to the relationships between myself and the study participants ; 43  5. Employ a critical stance in examining how larger structures and processes (inherent in the social context) influenced caregivers' experiences; 6. Explore the possibilities for change at a systems level and for formulating how the knowledge derived from this study could be used to influence health care decision making and policy development; and 7. Consider how the knowledge generated from nursing science could be used to meets its moral and social mandates. On a final note, I believe that one of the major strengths of critical theoretical perspectives is in the ability for researchers to account for their biases through critical analysis. While many qualitative researchers endeavour to suspend or "bracket" their own experiences to enhance objectivity (Oiler, 1982; Ornery, 1983), criticalists practice "strong objectivity" (Alcoff, 1991; Harding, 1991; Opie, 1992; Roman, 1993) by acknowledging their own experiences and locating themselves in  OAKLEY ( 1 9 8 1 ) HAS ARGUED THAT RESEARCHING OTHERS M A K E S THOSE BEING STUDIED