Show Summary Details
Page of

Defining a ‘good’ death 

Defining a ‘good’ death
Defining a ‘good’ death

Karen E. Steinhauser

and James A. Tulsky

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2022. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 16 May 2022

Introduction to defining a ‘good’ death

Woody Allen often commented that he was not afraid to die, but did not want to be there when it happened (Allen, 1983). This is emblematic of our society’s ambivalence towards the notion of a ‘good death’ (Rousseau, 1997; Steinhauser et al., 2001). Its inherent irony is instructive for palliative care clinicians. While death may be inevitable, for patients, it is rarely the goal. Therefore, naming deaths ‘good’ or ‘bad’ should be met with caution.

There is a long historical and literary tradition discussing the ‘good death’. In some of the best known work on evolution in Western attitudes towards death, social historian Philippe Aries uses cemetery iconography, notary records, wills, art, and literature to explore changing patterns in cultural norms of dying and death over the last 1500 years (Aries, 1980). For example, we learn that contemporary preferences for sudden death or death during sleep stand in contrast to previous eras in which populations literally prayed, in the Anglican Great Litany, not to die, ‘suddenly and unprepared’ (Aries, 1980; Vig and Pearlman, 2004). Aries’ work demonstrates the plasticity of how we view circumstances of death and the relativism of the terminology ‘good’ and ‘bad’.

Extraordinary variation exists in the social construction of the meaning of dying and death and the social organization and cultural norms surrounding end of life (Aries, 1980; Hart et al., 1998; Seale, 1998; Kim and Lee, 2003; DelVecchio Good et al., 2004; Long, 2004; Radley, 2004; Hirai et al., 2006). Recent work describing end-of-life activities in diverse regions such as Japan, North America, the Netherlands and Papau New Guinea, and spanning contemporary, classical, and biblical eras (Seale and van der Geest, 2004) finds that humans create and enact cultural scripts in ways that describe death as either ‘good’ or ‘bad’ (Seale and van der Geest, 2004).

While these cited works provide in-depth sociocultural and historical exploration of how we define the meaning of death, the focus of this chapter is on the reappearance in the last 40 years of attempts to define a good death in the medical context, the empirical investigation of the construct, the clinical implications of using the terminology ‘good death’, and an alternative framework for language defining preferences at end of life.

Context of contemporary exploration of ‘good death’

In 1908, William Osler conducted a study of 486 deaths at Johns Hopkins Hospital reporting that 90 patients experienced pain, 11 anxiety, and for the majority, death was ‘nothing more than falling asleep’ (Kring, 2006). Deaths in this era occurred all across the life course spectrum and were the result of either old age or catastrophic illness with limited medical intervention capable of extending life. Despite the site of Osler’s investigation, the majority of deaths at that time did not happen in the hospital.

Rather, deaths in the nineteenth and early twentieth centuries occurred primarily at home with support of family, church, and community (Sudnow, 1967; Aries, 1980; Hart et al., 1998). However, by the second half of the twentieth century the primary site of death had shifted to the hospital. Moreover, by the 1960s, the landscape of death in Western culture had changed dramatically as medicine had experienced a variety of therapeutic and technological revolutions resulting in the capacity to extend life, including antibiotic therapies, artificial nutrition and cardiopulmonary resuscitation. In this latter twentieth-century hospital setting, death’s meaning was narrowed primarily to a physiological event. As such, death was defined less as an expected and natural part of the life course and more as a failure of medical technology and intervention (Byock, 1996).

By the late 1950s, social reformers, such as Cicely Saunders, began to critique conventional medical care for dying patients, arguing that hospitals lacked both the specific expertise in palliation of symptoms as well as the multidisciplinary perspective that attended to social, psychological, and spiritual aspects of care (Saunders, 1978). In 1967, after returning to medical school to supplement her nursing, social work, and divinity training, she opened St. Christopher’s hospice as a multidisciplinary care centre that emphasized palliative versus curative therapies and promoted quality of life over quantity of life.

Amidst this social reform movement, in-depth inquiry of death and dying began to appear in the medical, nursing, and social science literatures. Seminal qualitative accounts were published in the few years following the opening of St. Christopher’s. Prominent sociologists Glaser, Strauss, and Sudnow were among the first to refocus study on end of life and to conceptualize trajectories of dying (Glaser and Strauss, 1965; Sudnow, 1967). Emerging from a tradition of grounded theory, Glaser and Strauss became participant observers in hospital settings and described four ‘contexts of awareness surrounding the dying experience: closed awareness, suspicious awareness, mutual pretense, and open awareness’ (Glaser and Strauss, 1965). Their conceptualization reflected an era during which patients often were not informed of terminal diagnoses or poor prognoses. Their sociological critiques were heavily counter-cultural in their scrutiny of the power dynamics of the paternalistic hospital culture.

A few years later, Elizabeth Kubler-Ross called further attention to the unmet needs of patients and the personal evolution of people coming to terms with a terminal diagnosis (Kubler-Ross, 1969). Though the stages of grief—denial, anger, bargaining, depression, and acceptance—were never tested empirically, her theory of dying and death is perhaps the best known and most frequently cited to this day. In fact, after her work, investigation of death and dying would receive scant attention until the 1990s.

The subtext of many of these investigations was that conventional medical care settings often played host to ‘bad’ deaths, typified by excessive use of technology, with patient and family wishes ignored, lack of patient knowledge and autonomy in decision-making, the patient reduced to a physiological system versus whole person, and quality of life devalued.

The hospice movement, both in its British foundations and its importation to the United States, arose within this context and was part of a larger ‘death with dignity’ movement of the 1960s and 1970s. Those involved in hospice and early palliative care worked to reclaim the experience of dying and death beyond a biomedical event. In this context, a ‘good’ death was the obverse of the previously described situations. The goals were to increase awareness of end of life as a part of a natural life course and to acknowledge dying patients as whole persons in the context of fuller lives lived, as well as family and community nexus. ‘Good death’ connoted a model of care more closely matching patient and family preferences, with the terminology designed to serve as a vision of improved experience for dying persons.

Empirical investigations of a ‘good death’

Despite the popularity and growth of the hospice movement, widespread, systematic attempts to define a ‘good death’ empirically did not appear in the medical literature until after the publication of the Study to Understand Prognosis and Preferences for the Outcomes and Risks of Treatment (SUPPORT) findings. This large multi-site study documented poor care of dying patients and their families in five top US medical centres and redefined the landscape of end-of-life and palliative care (The SUPPORT Principal Investigators, 1995). SUPPORT described hospitalized patients dying in pain, without their wishes known, and in isolation. The results provided empirical evidence of what was wrong with hospital deaths.

Care of dying patients became a priority in the United States and organizations such as the American Medical Association, the Veterans Health Administration, and The Robert Wood Johnson Foundation committed funding to improving education and quality of care (Field and Cassel, 1997). Efforts to develop and evaluate hospice and palliative care became expanded in US and international journals. If clinicians and administrators were to provide quality of care and quality of life at the end of life, they must first define quality. If SUPPORT had given empirical evidence of ‘bad’ deaths, what was the empirical evidence regarding the definition of a ‘good’ death?

In the 17 ensuing years, over 400 Medline articles include the construct ‘good death’. A smaller subset of independent studies have attempted to define the construct through analyses of qualitative and quantitative data gathered from patients, family members, and health-care providers. These have been reviewed systematically in both the medical and nursing literatures (Morris et al., 1986; King and Bushwick, 1994; McNamara et al., 1994; Asch et al., 1995; Layde et al., 1995; Low and Payne, 1996; Payne et al., 1996; Lynn et al., 1997; Ellington and Fuller, 1998; Emanuel and Emanuel, 1998; Singer, Martin, and Kelner, 1999; Steinhauser et al., 2000a, 2000b; Kristjanson et al., 2001; Mak, 2001; Cohen and Leis, 2002; Curtis et al., 2002; Hanson et al., 2002; Hopkinson and Hallett, 2002; Masson, 2002; Pierson et al., 2002; Ganzini et al., 2003; Kim and Lee, 2003; Tong et al., 2003; DelVecchio Good et al., 2004; Leichtentritt, 2004; Long, 2004; McNamara, 2004; Vig and Pearlman, 2004; Borbasi et al., 2005, 2006; Ferrell, 2005; Goldsteen et al., 2006; Hirai et al., 2006; Kring, 2006; Rietjens et al., 2006; Miyashita et al., 2007a; Hales et al., 2008). While each study lends a unique population or conceptual nuance, some common features exist among attempts to define a ‘good death’.


The social reformers propelling the ‘good death’ movement of the 1960s and 1970s were responding to the narrowing of the patient experience to the biomedical realm. Recent empirical investigation suggests that perhaps the most important feature to recognize about a ‘good death’ is its multidimensional nature.

In 2008, Hales et al. reviewed 17 studies defining quality at end of life (Hales et al., 2008). Five of the studies were conducted with health-care providers only; five were conducted among only patient populations, three were studies of non-patient, non-health-care provider populations; and the remaining included both patients and or families, and providers. Seven common broad domains were found: physical experience, psychological experience, social experience, spiritual or existential experience, the nature of health care, life closure and death preparation, and circumstances of death.

Expectedly, pain and symptom management was the most commonly identified theme. Yet, variation exists in how individuals prefer this domain to be addressed. For example, some patients may wish to balance analgesia with lucidity to allow meaningful personal interactions, and thus may tolerate some pain to achieve a higher state of alertness, if necessary. More recent studies suggest that physical dimensions include not only pain and symptoms but attention to functional status, which is highly correlated with continued independence and quality of life (Walke et al., 2007).

This body of work confirms that the physiological aspects of end-of-life experience are only a point of departure in overall definitions of a ‘good death’ or quality at end of life (Steinhauser et al., 2000a). Attention to emotional or psychological and social well-being are crucial. Patients experience their illness living a variety of roles and inter-relationships that need to be sustained as part of whole-person care. Interestingly, earlier sociological theory proposed that dying patients, and older adults, in general, experienced a natural ‘disengagement’ as end of life loomed. However, this theory has been debunked by empirical evidence showing the desire for and power of continued role engagement. Although dying patients usually experience physical decline or limitation, they may experience growth in social and emotional areas.

Similarly, attention to spiritual or transcendent aspects of experience are reported as central to quality experience and hold opportunity for growth (Daaleman and Nease, 1994; Byock, 1996; Cohen and Leis, 2002). This domain may be expressed in traditional religious terms, via connection with nature or as overall sense of meaning and purpose in life. There is some evidence of it increasing in importance as death nears (Byock, 1996). Of note, the absence of this domain in traditional quality of life measures was a main factor limiting their reliability and validity when applied to the context of dying.

While physical, social, psychological, and spiritual domains had been predicted by some of the pre-empirical literature, several new domains emerged from empirical investigation. These include preparation for death or end of life, nature of health care, and life completion. In contrast to settings of ‘closed awareness’ reported as common in the 1960s, ‘good death’ investigations of the 1990s and early 2000s suggested many patients with advanced serious illness wanted an opportunity to know what to expect about the course of their illness, to put personal affairs in order, to make financial arrangements and personal business, to not be a burden to family, to prepare their families for the future, and, for some, to plan one’s own funeral. It is important to note the contradiction found in the literature whereby populations within the same study report valuing ‘dying in their sleep’, ‘dying suddenly’, and ‘being prepared’ (Vig and Pearlman, 2004).

These studies show that preparation is not limited to patients. Families also need to be prepared for what to expect about the course of illness and decision-making.

I can’t tell you how many times, working in the emergency room, [that I saw] families [take a patient home]; this patient was going to die at home. And, when the last breath came, the families panicked. They brought the patient into the emergency room and went through the whole process [resuscitation]. Preparing the family, assessing what they actually know, and figuring out what you have to teach them is essential. [Nurse] (Steinhauser et al., 2000b)

Finally, some have also described the importance of provider preparation, coming to terms with their own fears about mortality and the emotions generated in caring for those who die (Steinhauser et al., 2000a).

Another less expected domain revealed in the ‘good death’ studies was the nature of health care. This domain focused on issues such as the appropriateness of level of technological intervention—levels in keeping with patient and family wishes, as well as communication with health-care providers, knowing how and where to get answers to questions, and overall relationship with the provider (Steinhauser et al., 2000a, 2000b; Perkins et al., 2008). The latter issue of relationship included maintaining patient dignity and treating patients as whole persons rather than as diseases.

Finally, the domain of life completion has been central to many investigations conceptualizing a ‘good’ death (Steinhauser et al., 2000b; Vig and Pearlman, 2004; Hirai et al., 2006; Rietjens et al., 2006). Attributes of completion include life review, closure, coming to peace, resolving conflicts, contributing to others, spending time with family and friends, and saying good-bye. Completion may involve personal reflection or individual spiritual practice, or may be more explicitly communal including family or a wider social circle. Of course many organized religions denote particular rites of spiritual completion for both the dying and the mourner. As with all domains, cues regarding specific expression should come from the patient or family. Within this domain, the attribute of contributing to others reminds family members and providers the importance not only of what patients may need to receive but also what they need to give to experience wholeness as they face the end of their lives.

Importance of role

Studies attempting to define a positive end-of-life experience reveal the importance of role in perception of what constitutes ‘good’. For example, studies show that physician perspectives tend to be more narrowly biomedical. And the data suggest a discrepancy in physician versus family and patient ratings of the importance of spirituality and completion attributes such as prayer (Byock, 1996; Cohen and Leis, 2002). In one survey asking participants to rank order nine attributes of end of life, families’ and patients’ rankings of being at peace and freedom from pain were statistically equal in importance (Steinhauser et al., 2000a). In contrast, physicians rated coming to peace as a distant third. Patients also were more likely to rate higher the importance of mental alertness and a desire not to be a burden to family or society.

In this same study, non-physician providers were more likely than patients to rate as important, ‘talking about the meaning of dying’. Family members were more likely than patients to rate the importance of discussing personal fears or meeting with clergy. Again, it is instructive to take cues from patients about how they want to discuss this issue. Qualitative findings suggest patients may wish to discuss purpose and life more than meaning of death.

In a survey of hospice nurses definitions of a ‘good death’ McNamara and colleagues found that definitions of ‘bad’ deaths included those in which the patient did not internalize hospice philosophy, leaving staff frustrated (McNamara et al., 1994). Non-internalization included not accepting the imminence of death, allowing non-palliative therapies to continue, and the family wanting ‘everything done’ despite terminal diagnoses. Such circumstances were thought to compromise a peaceful ‘natural’ death.

Importance of culture

In addition to individual and family variation in preferences for end of life, cultural scripts also predominate (see Chapter 2.5). While this is expanded upon more fully elsewhere in this textbook, studies relating to minority and majority population variation suggest a desire for clinician awareness of cultural issues, particularly heightened attention to the role spiritual beliefs play in decision-making, and attention to individual interpretation of minority culture scripts (Tong et al., 2003). Within the US population, a growing literature has demonstrated increased preference for life-sustaining therapies among Latino and African American populations (Tulsky et al., 1997). However, in one study African Americans were more likely to ‘want all available treatments’ but less likely than Caucasian participants to want to be ‘connected to machines’ (Steinhauser et al., 2000a). These varying responses among groups sensitized providers to patient and family differential interpretation of medical jargon, and true variation in preferences for treatment. As many sources have noted, when working with populations of patients traditionally denied access to care, withdrawing and withholding treatments are met with understandable apprehension.

Much of the empirical work defining a ‘good death’ has involved English-speaking Western populations in which individual decision-making and autonomy is culturally rewarded. However, more recently, a number of studies have explored the meaning of a ‘good death’ in a variety of cultural contexts (Munn and Zimmerman, 2006; Rietjens et al., 2006; Miyashita et al., 2007, 2008; Sanjo et al., 2007; Yao et al., 2007; Spathis and Booth, 2008; Murakawa and Nihei, 2009; Nelson et al., 2009; Iranmanesh et al., 2011; Wilches-Gutierrez et al., 2012; Wilson et al., 2009a, 2009b).

For example, Asian cultures with normative scripts including notions of filial piety will display markedly different preferences for treatment and communication on the part of patients and families. A noteworthy study of Korean attitudes towards patient autonomy showed that while 42% of respondents knew of their terminal condition, approximately 22% made treatment decisions primarily on their own with approximately 36% leaving their treatment decisions to others. Higher individual decision-making was associated with lower quality of life and quality of death scores, including assessments of physical and psychological comfort (Mo et al., 2012). More familial or communal decision-making models, though prevalent, are less well represented in the research literature, and therefore deserving of future study.

A study of Mexican culture found increased mortality rates around holidays, such as Christmas and All Saints Day, and accompanying higher ratings of quality of dying at these times (Wilches-Gutierrez et al., 2012). Respondents discussed the belief in a greater presence of religious deities on such holidays, ensuring a more ‘beautiful death’. Such interpretations stand in contrast to other cultural scripts which consider holiday deaths to be imbued with a greater sense of loss. Social construction of the meaning of death is varied and often in direct opposition to the palliative care’s normative views of grief and loss.

Importance of timing

In addition to cultural variation, clinicians must take note of the importance of timeframe in patient and family preferences. The literature reveals little consensus on what time frame constitutes the end of life. Furthermore, preferences will likely differ by stage of illness and evolve over time with definitions of ‘good experience’ living with advanced serious illness being distinct from ‘good dying’, as distinct from a ‘good death’. For example, immediately after a diagnosis of metastatic disease preparation may include discussions of possible courses of treatment, the combination of curative and palliative therapies, and helping patients remain integrated with normal work and social roles. As illness progresses, preparation may include discussions of decreasing the use of curative therapies and increasing palliative approaches, discussion of hospice, and increased attention to issues of completion. As dying becomes imminent, preparation may involve working with the family about expectations of care, location of care, and education regarding the very end of life.

Developmental stage

Most ‘good death’ studies have been done within adult populations. The Initiative for Pediatric Palliative Care and the National Hospice and Palliative Care Organization, among others, have spearheaded efforts to improve care of children at the end of life (Feudtner, 2004; Welch, 2008). Towards that end, investigators have gathered perspectives of families and providers to identify whether childhood death can ever be considered ‘good’, what elements might be associated with better and worse scenarios, and whether these are applicable to the nearly 55 000 childhood deaths in the United States (Feudtner, 2004; Welch, 2008). Full reviews of efforts to improve paediatric palliative care are covered elsewhere in the textbook. We limit our brief discussion to definitions of the construct of good death.

Some of the domains considered essential to ‘good’ end-of-life experiences in children are similar to those found in adult populations, such as pain and symptom management, emotional and spiritual support, maximizing quality of life, need for family respite, continuity of care across settings, and making care match individual values and preferences (Feudtner, 2004; Welch, 2008). Other domains are similar to adult populations but entail special considerations due to developmental and population complexities of paediatrics. For example, Feudtner outlined the importance of addressing the total population in need; in paediatrics that may include those children born with the expectation of impending death, those who acquire illnesses after birth, and those with a sudden death due to trauma (Feudtner, 2004). The leading causes of death after the age of 1 year include unintentional injuries followed by congenital abnormalities, malignant neoplasms, and intentional injuries. Research also shows the importance of collaborative decision-making and supportive decision-making which in paediatrics includes the family as well as informing and involving children to the extent to which they are developmentally capable and desiring involvement. An additional domain of ‘managing trade-offs adroitly’ refers to the potential caregiver tension between a desire to limit suffering and take therapeutic risks given a child’s age, life yet to be lived, and biological resilience (Feudtner, 2004). Additionally, investigators note the importance of maximizing safety and effectiveness, largely around polypharmacy, and finally, attending to timely introduction of bereavement services. The latter domain acknowledges that nearly all paediatric deaths are viewed by familial caregivers as untimely, adding to the complexity of grief. The sense of a death being on or off time in an expected life course applies to most deaths not experienced during old age. While investigation of the unique developmental needs are obvious in paediatric populations, the field would benefit from additional research examining social construction of the meaning of death across the lifespan and various adult developmental stages (Erikson, 1982).

Importance of diagnosis

Early hospice and palliative care populations were comprised, primarily, of patients with cancer. As such, literary, cultural, and empirical definitions of a ‘good death’ grew from the experience of living and dying from advanced cancer. Realizing the limitations to generalizability for broader end-of-life populations, investigators began seeking the perspectives of those living with a variety of life limiting illnesses (Steinhauser et al., 2000a, 2000b; Pierson et al., 2002; Walke et al., 2007; Chattoo and Ahmad, 2008; Chattoo and Atkin, 2009; Kaufman, 2011). Though sample size often precluded in-depth investigation of experiences unique to various illness types and disease trajectories, more recent work has been designed specifically to learn about the particular constructions of a ‘good death’ for those with congestive heart failure, chronic obstructive pulmonary disease, dementia, end-stage renal disease among others (Russ et al., 2007; Gott et al., 2008; Spathis and Booth, 2008). Two studies in heart disease (Gott et al., 2008; Chattoo and Ahmad, 2008; Chattoo and Atkin, 2009) demonstrated that both the heightened levels of uncertainty and increased possibility of sudden death among those with congestive heart failure, posed a challenge to conventional palliative care beliefs regarding values of patient ‘open awareness’ and even autonomy and individuality. They revealed conflict between cardiologists’ culture of ‘living with heart failure’ and palliative care’s attention to ‘dying with heart failure’ calling for different and more nuanced engagement with patient and illness uncertainty. Similarly, research in patients with chronic obstructive pulmonary disease highlighted notions of uncertainty and reported definitions of ‘good’ end-of-life care not precluding life-sustaining interventions.

Location of death

The seven domains listed earlier often are referred to under the general rubric of the biopsychosocial and spiritual model of care. The model, as just described, enjoys significant empirical support. In addition, two recent studies expand the notion of elements that contribute to the quality of end of life for patients and families. Casarett and colleagues highlighted advanced cancer patients’ preferences for ‘supportive services’ even over traditional hospice services (Casarett et al., 2008a). Supportive services included vouchers for practical assistance at home, transportation, peer support, meal delivery, case management, and family care. Perkins and colleagues emphasize additional supportive services such as emergency contacts and case management (Perkins et al., 2008). While much of the ‘good death’ literature was built on investigation into improving care of hospitalized patients, we must recognize that a majority of care in the dying trajectory occurs in the outpatient, home, and community settings. Therefore, newer models of good end-of-life experience must expand beyond the individual inpatient model of care.

The most recent research extending investigation beyond hospital end-of-life experiences has focused on defining a ‘good death’ in long-term care settings. Studies identify domains such as adequacy of staffing, facility environment and size, and the capacity of staff to ‘be there’ for residents and family (Munn and Zimmerman, 2006). Other studies identify the importance of creating ‘family-like’ bonds between staff and residents and family and improved communication. In this setting, definitions of a good death confirm previously identified interpersonal and more quality of life domains and extend those domains to include facets of resident and family experience directly related to quality of care.

In addition, a literature trend is the transition from identifying what constitutes a ‘good death’ to implementation of such definitions. One of the broadest examples is the efforts by the Department of Veterans Affairs to introduce and fund policy and clinical care initiatives to ensure quality end-of-life care for veterans across their nationwide health-care system (Edes et al., 2007; Casarett et al., 2008b). In these efforts, researchers wrestle with measuring the interconnected yet distinct concepts of ‘good death’, quality of life at end of life, and quality of care at end of life (Hales et al., 2010).

Opportunity for growth

Ira Byock has described the benefit of adopting a life cycle model when providing care at the end of life (Byock, 1996). The medical model begins assessment and treatment by generating a problem list. From the time a patient presents with symptoms, the clinical interviews and choice of diagnostic testing is determined by these problems. While this approach brings focus and efficiency to diagnostic testing and treatment, it is best suited to acute medicine and has limitations when applied to the context of care of incurably ill patients. Incurably ill patients surely are met with the daunting challenges of physical symptom exacerbation and functional decline; however, a purely problem-based approach offers less guidance in helping patients navigate areas of experience in which they may experience improvement or growth. A life cycle model assumes death is the natural end of a life course. And, building on the work of human development by Erikson, Bulter, and Cassell, Byock notes the expected developmental tasks associated with this phase of life (Cassell, 1973; Butler, 1974, 1980; Byock, 1996). These tasks include attention to life review, resolution of conflict, forgiveness, acceptance and generativity. Most importantly, this framework allows one to conceptualize end of life, like other phases of the life course, as holding opportunity for growth rather than only the decline predicted by the medical model. Growth will most likely occur in emotional and spiritual domains (and areas like preparation and completion) and is hypothesized to account for discrepancies in patient versus observer ratings of quality of life.

Measuring a ‘good death’ and quality of death and dying

Attempts to measure a ‘good death’ can be found among assessment tools specifically reflecting this name, (e.g. The Good Death Questionnaire, The Good Death Inventory) and also in a larger body of work assessing the quality of dying and death (Hales et al., 2008). A recent review showed extensive variation in measures rigor with regard to design and reliability and validity. Of the 18 measures reviewed, half reported no reliability of validity information, five were single-item measures, and fewer than half reported definitions of quality of death and dying. All measures were designed for retrospective accounts of family or professional caregivers. This is the soundest way of assessing last days of life and moment of death in populations with significant percentages unable to respond, yet requires additional efforts to understand direction and strength of proxy bias among domains as well as surrogate characteristics (Steinhauser et al., 2002; Hales et al., 2008). Additionally, most research demonstrates few measures allow for weighting response preferences to match individual circumstances of changing values over time. Measures of sensitivity to change over time also is an area in need of refinement.

Because this chapter focuses on definition and measurement of a ‘good death’, we have not discussed the vast literature on quality of life or quality of care at the end of life. The former is best suited to periods in the palliative care trajectory preceding periods of imminent dying and allows for real-time, prospective patient-rated assessments (Steinhauser, 2005). Quality of care may rely on patient, family, or provider perspective as well as administrative and organizational data of clinical benchmarks (Ferrell, 2005).

Clinical implications of the term ‘good death’

In this article we have reported on research which sought to define a ‘good’ death. Yet, it is important to discuss the clinical implications of using such language. In one of our studies, ‘In search of a good death: observations of patients, families and providers’ we concluded that there is no one ‘good death’ (Steinhauser et al., 2000b). Rather, each end-of-life experience is a process to be negotiated and renegotiated in the context of that patient’s and family’s values, preferences, and life course. We were strongly cautioned by nurses, for example, that it was important to know there was ‘no one right way to die’ and warned against implying to patients that ‘you’re not dying the right way, because you’re not dying the way we think you should’.

While early hospice founders used language of a ‘good’ death to rally reformers to a new vision of care, in recent years the language of a ‘good’ death often has taken on a denotation of specific expectations of what should occur at the end of life. The zeal driving the early movement has risked imposing a sense of a ‘right way’ to die. Its components include being free of pain, surrounded by family, free of conflict, acceptance death, stopping curative treatment, being at peace, and preferably dying at home. While those may be components many or most would define as positive, the implication is that one can define a good death, and should achieve it. Unfortunately, though propelled by positive intentions, such definition risks imposing an unintended paternalism. Furthermore, data of patient and family preferences at end of life exhibit far more nuance and variation.

For example, a national survey revealed that only about 50% of respondents ranked dying at home as important (Steinhauser et al., 2000a). In related qualitative research linked to the survey, while many valued dying at home, other patients and families described circumstances of caregiver frailty, superstition, or fear of bad memories as dissuading them from wanting the death to occur at home. Family members and non-physician health-care providers were significantly more likely than patients to identify ‘talking about the meaning of dying’ as important. And though coming to peace often is highly valued by patients, working to resolve conflicts can be complex and contain periods of great uncertainty, for patients, families, and providers hoping to guide them. And, by the time of death, everything is not always resolved. Again, there is caution to those working with dying patients that these uncertainties and lack of resolution do not represent failures on the parts of patients and families, but merely illustrate rich variation in the way people live their entire lives.

Health-care providers are most likely to be challenged by patients whose conception of a good death includes the use of medically non-beneficial treatments in the setting of clearly imminent death (e.g. parenteral nutrition or attempted cardiopulmonary resuscitation). In some cases, such preferences are reflective of unresolved loss or other emotional distress, and the best response may be empathy and support, which often leads to a change of treatment course. For other patients, these views are rooted in deeply vitalist traditions and are consonant with a broader set of values and goals. When that is true, an honest respect for individual choice compels us to honour such preferences. In the rare cases where health-care providers cannot find a way to reconcile these actions with their own sense of the goals of medicine, mediation via ethics consultation or other processes may be required.

Alternative language—defining goals at end of life

Therefore, rather than promoting a construct of a ‘good’ death, we favour the language of helping patients and families define and meet goals at the end of life. This paradigm has several advantages. First, in the medical context, we expect that goals vary between individuals. Second, we expect that they may change over time. Therefore, they require ongoing communication and negotiation. Third, the language of goals engages patients as active participants. Drawing from the self-management literature, the clinician may be the expert on the disease process, but the patient is the expert on their life. Fourth, it acknowledges a future orientation and ongoing contributions. Fifth, for the clinician, it is action oriented and moves beyond the idea (not common in palliative care) of having nothing more to offer. There are always additional goals.

Although we favour the paradigm in general, it does have some disadvantages, and points of caution. First, the language of goals may imply achievement and productivity, and risk imposing such expectations on patients. This would be counterproductive. Second, culturally we may think of goals as ‘doing’. Yet, we do not want to understate the importance of offering a sense of presence to patients and simply ‘being’ with them in their illness.

Being with patients and families may involve offering assurance, through words and actions, that the provider will be present through the course of treatment and illness. It may involve deep listening and necessary silences, in the presence of powerful emotions. And, it may involve assurance of commitment to negotiating and renegotiating preferences for care that attend to domains of physical, social, emotional, spiritual well-being as well as issues of preparation and life completion. Together, acknowledging that the quality of end-of-life experience is dependent upon attending to the multiple dimensions of whole persons facing illness in the context of an entire lifetime of values and choices and the web of family and community.


Allen, W. (1983). Without Feathers. New York: Ballantine BooksFind this resource:

Aries, P. (1980). The Hour of Our Death. New York: Knopf Publishers.Find this resource:

Butler, R.N. (1974). Successful aging and the role of the life review. Journal of the American Geriatrics Society, 22(12), 529–535.Find this resource:

Byock, I. R. (1996). The nature of suffering and the nature of opportunity at the end of life. Clinics in Geriatric Medicine, 12(2), 237–252.Find this resource:

Casarett, D., Fishman, J., O’Dwyer, P.J., Barg, F.K., Naylor, M., and Asch, D.A. (2008a). How should we design supportive cancer care? The patient’s perspective. Journal of Clinical Oncology, 26(8), 1296–1301.Find this resource:

Casarett, D., Pickard, A., Amos Bailey, F., et al. (2008b). Important aspects of end-of-life care among veterans: implications for measurement and quality improvement. Journal of Pain and Symptom Management, 35(2), 115–125.Find this resource:

Cassell, E.J. (1973). Learning to die. Bulletin of the New York Academy of Medicine, 49(12), 1110–1118.Find this resource:

Chattoo, S. and Atkin, K.M. (2009). Extending specialist palliative care to people with heart failure: semantic, historical and practical limitations to policy guidelines. Social Science & Medicine, 69(2), 147–153.Find this resource:

Curtis, J.R., Patrick, D.L., Engelberg, R.A., Norris, K., Asp, C., and Byock, I. (2002). A measure of the quality of dying and death: initial validation. Journal of Pain and Symptom Management, 24(1), 17–31.Find this resource:

Daaleman, T.P., and Nease, D.E., Jr. (1994). Patient attitudes regarding physician inquiry into spiritual and religious issues. The Journal of Family Practice, 39(6), 564–568.Find this resource:

DelVecchio Good, M.J., Gadmer, N.M., Ruopp, P., et al. (2004). Narrative nuances on good and bad deaths: internists’ tales from high-technology work places. Social Science & Medicine, 58(5), 939–953.Find this resource:

Emanuel, E.J. and Emanuel, L.L. (1998). The promise of a good death. The Lancet, 351(Suppl. 2), SII21–29.Find this resource:

Erikson, E. (1982). The Life Cycle Completed: A Review. New York: Norton.Find this resource:

Ferrell, B.R. (2005). Overview of the domains of variables relevant to end-of-life care. Journal of Palliative Medicine, 8(Suppl. 1), S22–29.Find this resource:

Field, M. and Cassel, C. (1997). Approaching Death: Improving Care at the End of Life. Washington, DC: Institute of Medicine.Find this resource:

Ganzini, L., Goy, E.R., Miller, L.L., Harvath, T.A., Jackson, A., and Delorit, M.A. (2003). Nurses’ experiences with hospice patients who refuse food and fluids to hasten death. The New England Journal of Medicine, 349(4), 359–365.Find this resource:

Glaser, B. and Strauss, A. (1965). Awareness of Dying. Chicago, IL: Aldine.Find this resource:

Gott, M., Small, N., Barnes, S., Payne, S., and Seamark, D. (2008). Older people’s views of a good death in heart failure: implications for palliative care provision. Social Science & Medicine, 67(7), 1113–1121.Find this resource:

Hales, S., Zimmermann, C., and Rodin, G. (2010). Review: the quality of dying and death: a systematic review of measures. Palliative Medicine, 24(2), 127–144.Find this resource:

Hales, S., Zimmermann, C., and Rodin, G. (2008). The quality of dying and death. Archives of Internal Medicine, 168(9), 912–918.Find this resource:

Hirai, K., Miyashita, M., Morita, T., Sanjo, M., and Uchitomi, Y. (2006). Good death in Japanese cancer care: a qualitative study. Journal of Pain and Symptom Management, 31(2), 140–147.Find this resource:

Kaufman, S. (2011). Improving quality and outcomes with alternative dialysis modalities. Nephrology News & Issues, 25(12), 8, 10.Find this resource:

Kubler-Ross, E. (1969). On Death and Dying. New York: Macmillan Publishing.Find this resource:

Leichtentritt, R.D. (2004). The meaning that young Israeli adults ascribe to the least undesirable death. Death Studies, 28(8), 733–759.Find this resource:

Long, S.O. (2004). Cultural scripts for a good death in Japan and the United States: similarities and differences. Social Science & Medicine, 58(5), 913–928.Find this resource:

Lynn, J., Teno, J.M., Phillips, R.S., et al. (1997). Perceptions by family members of the dying experience of older and seriously ill patients. SUPPORT Investigators. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments. Annals of Internal Medicine, 126(2), 97–106.Find this resource:

McNamara, B., Waddell, C., and Colvin, M. (1994). The institutionalization of the good death. Social Science & Medicine, 39(11), 1501–1508.Find this resource:

Miyashita, M., Sanjo, M., Morita, T., Hirai, K., and Uchitomi, Y. (2007). Good death in cancer care: a nationwide quantitative study. Annals of Oncology, 18(6), 1090–1097.Find this resource:

Mo, H.N., Shin, D.W., Woo, J.H., et al. (2012). Is patient autonomy a critical determinant of quality of life in Korea? End-of-life decision making from the perspective of the patient. Palliative Medicine, 26(3), 222–231.Find this resource:

Munn, J.C. and Zimmerman, S. (2006). A good death for residents of long-term care: family members speak. Journal of Social Work in End-Of-Life & Palliative Care, 2(3), 45–59.Find this resource:

Murakawa, Y. and Nihei, Y. (2009). Understanding the concept of a ‘good death’ in Japan: differences in the views of doctors, palliative and non-palliative ward nurses. International Journal of Palliative Nursing, 15(6), 282–289.Find this resource:

Pierson, C.M., Curtis, J.R., and Patrick, D.L. (2002). A good death: a qualitative study of patients with advanced AIDS. AIDS Care, 14(5), 587–598.Find this resource:

Rousseau, P. (1997). Hope and the terminally ill. Clinics in Geriatric Medicine, 5(13), 15.Find this resource:

Saunders, C. (1978). Hospice care. American Journal of Medicine, 65(5), 726–728.Find this resource:

Seale, C. and van der Geest, S. (2004). Good and bad death: introduction. Social Science & Medicine, 58(5), 883–885.Find this resource:

Singer, P.A., Martin, D.K., and Kelner, M. (1999). Quality end-of-life care: patients’ perspectives. Journal of the American Medical Association, 281(2), 163–168.Find this resource:

Spathis, A. and Booth, S. (2008). End of life care in chronic obstructive pulmonary disease: in search of a good death. International Journal of Chronic Obstructive Pulmonary Disease, 3(1), 11–29.Find this resource:

Steinhauser, K.E., Christakis, N.A., Clipp, E.C., McNeilly, M., McIntyre, L., and Tulsky, J.A. (2000a). Factors considered important at the end of life by patients, family, physicians, and other care providers. Journal of the American Medical Association, 284(19), 2476–2482.Find this resource:

Steinhauser, K.E., Clipp, E.C., McNeilly, M., Christakis, N.A., McIntyre, L.M., and Tulsky, J.A. (2000b). In search of a good death: observations of patients, families, and providers. Annals of Internal Medicine, 132(10), 825–832.Find this resource:

The SUPPORT Principal Investigators (1995). A controlled trial to improve care for seriously ill hospitalized patients: the study to understand prognosis and preferences for outcomes and risks of treatments (SUPPORT). Journal of the American Medical Association, 274, 1591–1598.Find this resource:

Tong, E., McGraw, S.A., Dobihal, E., Baggish, R., Cherlin, E., and Bradley, E.H. (2003). What is a good death? Minority and non-minority perspectives. Journal of Palliative Care, 19(3), 168–175.Find this resource:

Tulsky, J.A., Cassileth, B.R., and Bennett, C.L. (1997). The effect of ethnicity on ICU use and DNR orders in hospitalized AIDS patients. Journal of Clinical Ethics, 8(2), 150–157.Find this resource:

Vig, E.K., and Pearlman, R.A. (2004). Good and bad dying from the perspective of terminally ill men. Archives of Internal Medicine, 164(9), 977–981.Find this resource:

Welch, S.B. (2008). Can the death of a child be good? Journal of Pediatric Nursing, 23(2), 120–125.Find this resource:

Wilches-Gutierrez, J.L., Arenas-Monreal, L., Paulo-Maya, A., Pelaez-Ballestas, I., and Idrovo, A.J. (2012). A ‘beautiful death’: mortality, death, and holidays in a Mexican municipality. Social Science & Medicine, 74(5), 775–782.Find this resource:

Wilson, D.M., Fillion, L., Thomas, R., Justice, C., Bhardwaj, P.P., and Veillette, A.M. (2009a). The ‘good’ rural death: a report of an ethnographic study in Alberta, Canada. Journal of Palliative Care, 25(1), 21–29.Find this resource:

Yao, C.A., Hu, W.Y., Lai, Y.F., Cheng, S.Y., Chen, C.Y., and Chiu, T.Y. (2007). Does dying at home influence the good death of terminal cancer patients? Journal of Pain and Symptom Management, 34(5), 497–504.Find this resource:

Online references

Asch, D.A., Hansen-Flaschen, J., and Lanken, P.N. (1995). Decisions to limit or continue life-sustaining treatment by critical care physicians in the United States: conflicts between physicians’ practices and patients’ wishes. American Journal of Respiratory and Critical Care Medicine, 151(2 Pt. 1), 288–292.Find this resource:

Borbasi, S., Jones, J., Lockwood, C., and Emden, C. (2006). Health professionals’ perspectives of providing care to people with dementia in the acute setting: toward better practice. Geriatric Nursing, 27(5), 300–308.Find this resource:

Borbasi, S., Wotton, K., Redden, M., and Chapman, Y. (2005). Letting go: a qualitative study of acute care and community nurses’ perceptions of a ‘good’ versus ‘bad’ death. Australian Critical Care, 18(3), 104–113.Find this resource:

Butler, R.N. (1980). The life review: an unrecognized bonanza. International Journal of Aging and Human Development, 12(1), 35–38.Find this resource:

Chattoo, S. and Ahmad, W.I. (2008). The moral economy of selfhood and caring: negotiating boundaries of personal care as embodied moral practice. Sociology of Health & Illness, 30(4), 550–564.Find this resource:

Cohen, S. and Leis, A. (2002). What determines the quality of life in terminally ill cancer patients from their own perspectives? Journal of Palliative Care, 18, 48–58.Find this resource:

Edes, T., Shreve, S., and Casarett, D. (2007). Increasing access and quality in Department of Veterans Affairs care at the end of life: a lesson in change. Journal of the American Geriatrics Society, 55(10), 1645–1649.Find this resource:

Ellington, S. and Fuller, J. (1998). A good death?: finding a balance between the interests of patients and caregivers. Generations, 22(3), 87–91.Find this resource:

Feudtner, C. (2004). Perspectives on quality at the end of life. Archives of Pediatrics and Adolescent Medicine, 158(5), 415–418.Find this resource:

Goldsteen, M., Houtepen, R., Proot, I.M., Abu-Saad, H.H., Spreeuwenberg, C., and Widdershoven, G. (2006). What is a good death? Terminally ill patients dealing with normative expectations around death and dying. Patient Education and Counseling, 64(1–3), 378–386.Find this resource:

Hanson, L.C., Henderson, M., and Menon, M. (2002). As individual as death itself: a focus group study of terminal care in nursing homes. Journal of Palliative Medicine, 5(1), 117–125.Find this resource:

Hart, B., Sainsbury, P., and Short, S. (1998). Whose dying? A sociological critique of the ‘good death.’ Mortality, 3(1), 65–77.Find this resource:

Hopkinson, J. and Hallett, C. (2002). Good death? An exploration of newly qualified nurses’ understanding of good death. International Journal of Palliative Nursing, 8(11), 532–539.Find this resource:

Iranmanesh, S., Hosseini, H., and Esmaili, M. (2011). Evaluating the ‘good death’ concept from Iranian bereaved family members’ perspective. The Journal of Supportive Oncology, 9(2), 59–63.Find this resource:

Kim, S. and Lee, Y. (2003). Korean nurses’ attitudes to good and bad death, life-sustaining treatment and advance directives. Nursing Ethics, 10(6), 624–637.Find this resource:

King, D. and Bushwick, B. (1994). Beliefs and attitudes of hospital inpatients about faith healing and prayer. The Journal of Family Practice, 39(4), 349–352.Find this resource:

Kring, D.L. (2006). An exploration of the good death. Advances in Nursing Science, 29(3), E12–24.Find this resource:

Kristjanson, L.J., McPhee, I., Pickstock, S., Wilson, D., Oldham, L., and Martin, K. (2001). Palliative care nurses’ perceptions of good and bad deaths and care expectations: a qualitative analysis. International Journal of Palliative Nursing, 7(3), 129–139.Find this resource:

Layde, P.M., Beam, C.A., Broste, S.K., et al. (1995). Surrogates’ predictions of seriously ill patients’ resuscitation preferences. Archives of Family Medicine, 4(6), 518–523.Find this resource:

Low, J.T. and Payne, S. (1996). The good and bad death perceptions of health professionals working in palliative care. European Journal of Cancer Care (England), 5(4), 237–241.Find this resource:

Mak, M.H. (2001). Awareness of dying: an experience of Chinese patients with terminal cancer. Omega (Westport), 43(3), 259–279.Find this resource:

Masson, J. (2002). Non-professional perceptions of ‘good death’: a study of the views of hospice care patients and relatives of deceased hospice care patients. Mortality, 7(2), 191–209.Find this resource:

McNamara, B. (2004). Good enough death: autonomy and choice in Australian palliative care. Social Science & Medicine, 58(5), 929–938.Find this resource:

Miyashita, M., Morita, T., Sato, K., Hirai, K., Shima, Y., and Uchitomi, Y. (2008). Factors contributing to evaluation of a good death from the bereaved family member’s perspective. Psycho-Oncology, 17(6), 612–20.Find this resource:

Morris, J., Suissa, S., Sherwood, S., Wright, S., and Greer, D. (1986). Last days: a study of the quality of life of terminally ill cancer patients. Journal of Chronic Diseases, 39, 47–62.Find this resource:

Nelson, M.L., Schrader, S.L., and Eidsness, L.M. (2009). ‘South Dakota’s Dying to Know’: personal experiences with end-of-life care. Journal of Palliative Medicine, 12(10), 905–913.Find this resource:

Payne, S.A., Langley-Evans, A., and Hillier, R. (1996). Perceptions of a ‘good’ death: a comparative study of the views of hospice staff and patients. Palliative Medicine, 10(4), 307–312.Find this resource:

Perkins, P., Booth, S., Vowler, S.L., and Barclay, S. (2008). What are patients’ priorities for palliative care research?—a questionnaire study. Palliative Medicine, 22(1), 7–12.Find this resource:

Pierson, C.M., Curtis, J.R., and Patrick, D.L. (2002). A good death: a qualitative study of patients with advanced AIDS. AIDS Care, 14(5), 587–598.Find this resource:

Radley, A. (2004). Pity, modernity and the spectacle of suffering. Journal of Palliative Care, 20(3), 179–184.Find this resource:

Rietjens, J.A., van der Heide, A., Onwuteaka-Philipsen, B.D., van der Maas, P.J., and van der Wal, G. (2006). Preferences of the Dutch general public for a good death and associations with attitudes towards end-of-life decision-making. Palliative Medicine, 20(7), 685–692.Find this resource:

Russ, A.J., Shim, J.K., and Kaufman, S.R. (2007). The value of ‘life at any cost’: talk about stopping kidney dialysis. Social Science & Medicine, 64(11), 2236–2247.Find this resource:

Sanjo, M., Miyashita, M., Morita, T., et al. (2007). Preferences regarding end-of-life cancer care and associations with good-death concepts: a population-based survey in Japan. Annals of Oncology, 18(9), 1539–1547.Find this resource:

Seale, C. (1998). Theories in health care and research: theories and studying the care of dying people. BMJ, 317(7171), 1518–1520.Find this resource:

Steinhauser, K.E. (2005). Measuring end-of-life care outcomes prospectively. Journal of Palliative Medicine, 8(Suppl. 1), S30–41.Find this resource:

Steinhauser, K.E., Christakis, N.A., Clipp, E.C., et al. (2001). Preparing for the end of life: preferences of patients, families, physicians, and other care providers. Journal of Pain and Symptom Management, 22(3), 727–737.Find this resource:

Steinhauser, K.E., Clipp, E.C., and Tulsky, J.A. (2002). Evolution in measuring the quality of dying. Journal of Palliative Medicine, 5(3), 407–414.Find this resource:

Sudnow, D. (1967). The Social Organization of Dying. Englewood Cliffs, NJ: Prentice Hall.Find this resource:

Walke, L.M., Byers, A.L., Gallo, W.T., Endrass, J., and Fried, T.R. (2007). The association of symptoms with health outcomes in chronically ill adults. Journal of Pain and Symptom Management, 33(1), 58–66.Find this resource:

Wilson, D.M., Fillion, L., Thomas, R., Justice, C., Veillette, A.M., and Bhardwaj, P. (2009b). Planning and providing for a good death using rural French-Canadian and English-Canadian insights. Reviews in the Neurosciences, 20(3–4), 313–319.Find this resource: