10

Conclusions and Recommendations

Our lives are time travel, moving in one direction only. We accompany one another as long as we can; as long as time grants us.

We go toward something that is not yet.

Dying is at once a fact of life and a profound mystery. Death comes to all, yet each person experiences it in ways that are only partly accessible to the family member or physician, the researcher or philosopher. In principle, humane and skillful care for the dying is a social obligation as well as a personal offering from those directly involved. In reality, both society and individuals fall short of what is reasonably, if not simply, achievable.

Evidence and experience both indicate that much dying is far harder than it should be. Like the mythic king Tantalus, who reached for fruit and water just beyond his grasp, the dying person has too often been destined to seek but not find a pain-free and peaceful death. Such a fate need not be; better deaths are possible now.

This is not to claim that dying and death can be made easy, although certain illnesses tend to bring less-difficult deaths than others; nor is it to suggest that caring well for those approaching death can be made an exact science or flawless art. Caring well involves a frequently stressful, imperfect effort to balance science with sentiment, honesty with hope, patterns of care with room for exceptions.

This report has identified ways in which clinicians, educators, and communities do not care adequately for those approaching death. It has pointed to steps that can improve care at the end of life and sustain improvements through difficult times. It has also highlighted the reasons for believing that professionals, policymakers, and the public are ready to support such steps. These reasons range from the examples of well-known men and women facing death with grace to the focus on deficiencies in care stimulated by the debate over assisted suicide. In sum, the timing appears right to press for a vigorous societal commitment to improve care for the dying. Such a commitment would motivate and sustain local and national efforts to strengthen and apply the existing knowledge base, reform procedures and policies that impede good care for the dying, and stress the caring functions of health delivery systems.




TODAY AND TOMORROW

The preceding chapters have profiled important aspects of dying in America. They have noted that three-fourths of those who die are elderly and that most people who die are covered by Medicare or other federal, state, or local programs for older, disabled, or impoverished people. Compared to earlier times, death now comes more often in old age from a chronic or progressive disease. Most deaths occur in institutions, mainly hospitals and, to a lesser extent, nursing homes. The proportion of people who die at home is, however, increasing. Only a minority of all deaths involve people cared for by formal hospice programs, and the majority of these involve cancer diagnoses. All patients can potentially benefit from good palliative care, but hospice programs as organized and financed in the United States are most applicable to those with relatively predictable prognoses.

Although specifics may vary according to the nature of a person's disease and his or her personal circumstances, care for dying people and those close to them has several broad dimensions: physical, psychological, spiritual, and practical. Each of these dimensions is intertwined in the key processes of care, which include determining and communicating diagnosis and prognosis, establishing goals and plans, and fitting palliative and other care to these goals. The way care is organized, financed, monitored, and regulated affects how these processes are carried out in different settings for different kinds of patients.

The twenty-first century will bring new realities as well as continuing problems and opportunities in care at the end of life. It will undoubtedly deliver improvements in what medical science can do to prevent and relieve distress for those approaching death, but demographic, economic, and other trends will strain systems that already find it difficult to deliver what clinical knowledge currently allows—and what compassion should grant.

The committee focused primarily on trends and expectations relevant to care for dying patients during the next 10 to 20 years. One of these expectations is that policymakers and others will need to prepare during this period for the final demographic consequences of the post-World War II baby boom. The oldest members of the baby boom generation will reach age 65 in the year 2011, and the youngest members will do so nearly 20 years later. Barring dramatic unexpected developments, older adults will constitute a larger proportion of the population than today, and the absolute numbers of dying patients will be much larger. In the face of an ever-increasing population with chronic conditions, the current system, with its emphasis on acute curative care, will fail to provide needed long-term chronic care and palliative and support services. Although health care and social service providers have a long lead time compared with the educators and communities who had to scramble to provide schooling for the baby boom generation, the difficulties that policymakers are having with Social Security and Medicare do not bode well for the nation's ability to cope with an aging population.

In addition, continuing increases in longevity will mean that the proportion of those dying past age 75 will increase. The impact on the health care system and society on larger numbers of people dying in older old age will depend in part on their health status in the years before death. The "compression of morbidity hypothesis," for which the evidence is ambiguous, posits that an increase in healthier life styles means that people will experience disabling conditions for a smaller proportion of the years before death than today. Even if this does occur, the overall impact of the larger proportion of older people and their longer life span will still pose enormous challenges for social and health care systems.

Demographic politics are likely to be complex in coming years. On the one hand, the so-called senior lobby should be strengthened by the "baby boom" cohort. The "baby boomers," throughout their childhood, youth, and early middle age, have drawn public attention to their needs and concerns. On the other hand, the burden on the smaller working-age population of supporting its elders raises the specter of a generational backlash (even though younger people could also be burdened if the social support available to their aging parents is reduced). Such a backlash could take many forms, including means-testing Social Security benefits or less sympathy for supportive services for frail elderly people, those with serious disabilities, and people who are dying.

Cultural trends add further complexity to demographic politics. Immigrants, who tend to be young and have more children than native-born Americans, have been a major source of growth in the total U.S. population (NRC, 1997). Nonhispanic Caucasians—now a majority—are a decreasing proportion of the U.S. population. Thus, the large elderly population in the first decades of the next century will be dependent on a smaller and more culturally diverse population of younger people.

As described in preceding chapters, health care in the United States is undergoing major changes. The use of hospital services has dropped significantly, and further declines are likely, particularly in those areas with continued relatively high rates of use. Hospital care will likely be less available to dying patients, potentially including some who could benefit from such care as well as many who will do better with alternatives.

Despite indications that health care spending increases have moderated from years past, the pressure to cut health care costs does not appear to be abating. In particular, given the bleak financial picture for Medicare and the significant portion of spending accounted for by beneficiaries in their last year of life, these constraints may be manifested in some degree of age-based rationing. That is, older people could be denied services simply because of their age, without regard to their life expectancies, function, or expected benefit from care. Decreased access to advanced technologies intended to prolong life could be a highly visible issue, although it is relatively unimportant for the great majority of those dying at an advanced age.

If successful, increased efforts to shift Medicare beneficiaries into various forms of managed care will test these entities, whose enrollment now is overwhelmingly concentrated in younger, healthier age groups. Potential problem areas include payment, contracting, and utilization review mechanisms that limit access to clinicians and care teams experienced in palliative care, patient scheduling norms that limit time for careful clinician-patient communication, marketing strategies intended to discourage enrollment or encourage disenrollment by seriously ill people, drug formularies that exclude or restrict important medications, and bureaucratic hurdles that discourage people from seeking care or preclude them from receiving requested services. One characteristic of many managed care organizations is an emphasis on moving more responsibility for diagnosis and patient management to primary care clinicians, including nurse practitioners, nurses, and other nonphysicians. How well this will work for seriously ill and elderly patients is not yet clear. A major role for nurses and other health care professionals is, however, already well tested in hospice, social health maintenance organizations, and other settings that care for patients dying of advanced progressive diseases, particularly cancer, and for frail elderly people.

As well as focusing attention on competing ethical and policy values, this prospect of constrained resources underscores the need for more valid and reliable tools for assessing health status, measuring outcomes, and linking them to health interventions. Considerable progress is being made in this arena. In addition, improved computer-based information and decision support systems promise several benefits. At the level of the individual, they will help patients, families, and clinicians in making more informed decisions about care at the end of life. At the system or societal level, they will strengthen structures and processes of accountability involving health care providers, purchasers, patient/consumers, and public officials.

Moreover, advanced information and communications technologies may make health care more accessible for some who face geographic or other barriers to care. Although the potential for practical, effective, and affordable "medicine at a distance" is more promised than proved, telemedicine could help bring palliative and other expertise and services to homes, nursing homes, and other places where they are not very available today (IOM, 1996e). People who have chronic and advanced illnesses but do not qualify for hospital inpatient care could benefit.




FINDINGS AND RECOMMENDATIONS

Although the committee found many problems, it also found much that was good and improving in care for those approaching death. It was impressed by the principles, aspirations, and practices being advanced by the field of palliative medicine and being implemented or attempted through interdisciplinary care teams in varied settings, including hospices, hospitals, nursing homes, and private homes. More effort is now being devoted to understanding differences in people's paths toward death and developing programs that are flexible enough to accommodate these differences. In addition to dedicated professionals, many volunteers contribute their time and energy to provide emotional and practical support to dying patients and those close to them. People from a variety of backgrounds have also joined together to direct the attention of policymakers and ordinary citizens to the need to remove barriers to good care created by laws, regulations, organizational practices, and lack of supportive community resources.

Care for the dying should also benefit from ongoing efforts to improve continuity of care, strengthen information systems, prevent health problems, and create accountability for the quality of care. The developing fields of outcomes measurement can make important contributions to the care of the dying by teaching us how to conceptualize and measure quality of life and well-being for those approaching death. Quality improvement strategies are helping to identify and remedy system problems that impede good care. The committee is encouraged by the growing interest in end-of-life issues as recently evidenced by the publication of books and articles for lay and professional audiences, the commitment of foundation resources to support research and education, and the organization of conferences, working groups, and other initiatives sponsored by professional societies and others.




Deficiencies in Care at the End of Life

Notwithstanding these positive features, the committee concluded that very serious problems remain. Indeed, in this committee's view, if physician and hospital performance in infection control were as poor as it is, for example, in pain management, the ensuing national outcry would create an immediate demand for responses from clinicians, managers, and educators.

The committee identified four broad deficiencies in the current care of people with life-threatening and incurable illnesses. First, too many people suffer needlessly at the end of life both from errors of omission—when caregivers fail to provide palliative and supportive care known to be effective—and from errors of commission—when caregivers do what is known to be ineffective and even harmful. As reported in Chapter 3, studies have repeatedly indicated that a significant proportion of dying patients and patients with advanced disease experience serious pain, despite research identifying a range of effective pharmacological and other options for relieving most pain. Other symptoms are less well studied, and more research on symptom prevalence and management is needed, but the information available to the committee suggested similar care problems. Deficiencies in the application of existing knowledge to prevent and manage pain and other distressing symptoms stem from an unfortunate mix of ignorance about effective pharmacological and other interventions, misplaced concern about opioid addiction, and inadequate attention to people's quality of life while dying. These problems are reinforced by care systems that are not structured to provide the clinical expertise, reliability, continuity, and emotional support needed by people approaching death. Cultural biases and fears about illness, disability, and death may also contribute to avoidance of dying patients and those close to them.

In perverse counterpoint to the problem of undertreatment, the aggressive use of ineffectual and intrusive interventions may prolong and disfigure the period of dying. Some of this care is knowingly accepted; some is provided counter to patients' wishes; much is probably provided and accepted with little knowledge or consideration of its probable benefits and burdens. Medical culture still tolerates and even rewards the misapplication of life-sustaining technologies while slighting the prevention and relief of suffering.

Second, legal, organizational, and economic obstacles conspire to obstruct reliably excellent care at the end of life. Despite some reforms, outdated and scientifically flawed drug-prescribing laws, regulations, and interpretations by state medical boards still frustrate and intimidate physicians who wish to relieve their patients' pain. Addiction to appropriately prescribed opioids is virtually nonexistent whereas their underprescription for pain is well documented. Organizational structures often interfere with the coordination and continuity of care and impede the further development and application of palliative care strategies in patient care, professional education, and research.

Financial incentives also discourage health care practitioners and providers from rearranging care so that it serves dying patients well. Traditional financing mechanisms—including arrangements based on discounted fees—provide incentives for the overuse of procedural services and the under-provision of the assessment, evaluation, management, and supportive services so important for people with serious chronic or progressive medical problems. Medicare hospice benefits have helped fill gaps for a small segment of dying patients, but many more have conditions that do not readily fit the current hospice model or requirements. Alternatives to fee-for-service financing in combination with restrictive administrative practices pose different potential hazards that are largely unstudied as they affect seriously and incurably ill people.

Third, the education and training of physicians and other health care professionals fails to provide them with the knowledge, skills, and attitudes required to care well for the dying patient. Many deficiencies in practice stem from fundamental prior failures in professional education. Despite encouraging signs of change, most clinicians-in-training experience and learn too little of the caring that helps people to live well while dying. Undergraduate, graduate, and continuing medical education do not sufficiently prepare health professionals to recognize the final phases of illnesses, construct effective strategies for care, communicate with patients and those close to them, and understand and manage their own emotional reactions to death and dying.

Fourth, current knowledge and understanding are inadequate to guide and support the consistent practice of evidence-based medicine at the end of life. In addition to existing knowledge not being well used, we still know too little about how people die; how they want to die; and how different kinds of physical, emotional, and spiritual care might better serve the dying and those close to them. Biomedical and clinical research have emphasized the development of knowledge that contributes to the prevention, detection, or cure of disease and to the prolongation of life. Research on the end stages of diseases and the physiological bases of symptoms and symptom relief has been less well supported. Epidemiological and health services research has likewise not provided a strong base for understanding the degree to which people suffer symptoms (except, perhaps, cancer pain), experience death alone rather than in the company of those who care, comprehend diagnostic and prognostic information, and achieve a dying that is reasonably consistent with their preferences, community norms, and palliative care principles. Methods development is important to define and measure outcomes other than death (including patient and family perceptions) and to monitor and improve the quality of care for those approaching death.

More generally, it seems that this nation has not yet discovered how to talk realistically but comfortably about the end of life, nor has it learned how to value the period of dying as it is now experienced by most people. For most of human history, death came fairly quickly in childhood or at adult ages that today seem relatively young. As the twentieth century ends, most people in economically advanced countries live fairly healthy lives into older age but then experience progressive disability for some time before they die. Except what can be inferred from newspaper obituary pages, this reality is largely shunned by the news, information, and entertainment media as distasteful or uninteresting. One result is an unhelpful combination of public fear, misinformation, and oversimplification that views misery as inescapable, pain as unavoidable, and public spending as misdirected for people who are approaching death.




Recommendations and Future Directions

The committee developed seven recommendations directed at different decisionmakers and different deficiencies in care at the end of life. These recommendations and a brief explanation follow. Each applies generally to people approaching death, including those for whom death is imminent and those with serious, eventually fatal illnesses who may live for some time.

RECOMMENDATION 1: People with advanced, potentially fatal illnesses and those close to them should be able to expect and receive reliable, skillful, and supportive care. Educating people about care at the end of life is a critical responsibility of physicians, hospitals, hospices, support groups, public programs, and media. Most patients and families need information not only about diagnosis and prognosis but also about what support and what outcomes they should reasonably be able to anticipate. They should, for example, not be allowed to believe that pain is inevitable or that supportive care is incompatible with continuing efforts to diagnose and treat. They should learn—before their last few days of life—that supportive services are available from hospices and elsewhere in the community and that those involved in their care will help arrange such services. Patient and family expectations and understanding will be aided by advance care planning that considers needs and goals, identifies appropriate surrogate decisionmakers, and avoids narrow preoccupation with written directives. To these ends, health care organizations and other relevant parties should adopt policies regarding information, education, and assistance related to end-of-life decisions and services. For those who seek to build public understanding of dying as a part of life and to generate public demand for reliable and effective supportive services, one model can be found in the perspectives, spirit, and strategies that have guided efforts to promote effective prenatal care and develop mother- and family-oriented arrangements for childbirth.

RECOMMENDATION 2: Physicians, nurses, social workers, and other health professionals must commit themselves to improving care for dying patients and to using existing knowledge effectively to prevent and relieve pain and other symptoms. Most patients depend on health care professionals to prevent and manage the varying physical and psychological symptoms that accompany advanced illness. To meet their obligations to their patients, practitioners must hold themselves and their colleagues responsible for using existing knowledge and available interventions to assess, prevent, and relieve physical and emotional distress. Unrelieved pain and other symptoms are the most evident problems that practitioners can readily avoid for the great majority of patients, but problems with communication, appropriate regard for patient and family wishes, and timely referral to palliative care specialists or teams are other areas in need of improvement. When good practice is hindered by organizational, financial, or legal impediments, health professionals have the responsibility as individuals and members of larger groups to advocate for system change.

RECOMMENDATION 3: Because many deficiencies in care reflect system problems, policymakers, consumer groups, and purchasers of health care should work with health care providers and researchers to

a. strengthen methods for measuring the quality of life and other outcomes of care for dying patients and those close to them;

b. develop better tools and strategies for improving the quality of care and holding health care organizations accountable for care at the end of life;

c. revise mechanisms for financing care so that they encourage rather than impede good end-of-life care and sustain rather than frustrate coordinated systems of excellent care; and

d. reform drug prescription laws, burdensome regulations, and state medical board policies and practices that impede effective use of opioids to relieve pain and suffering.

Although individuals must act to improve care at the end of life, systems of care must be changed to support such action. System change requires the involvement of public and private purchasers of care, regulators, and others whose policies and practices may create incentives for inappropriate care and barriers to excellent care.

Better information systems and tools for measuring outcomes and evaluating care are critical to the creation of effective and accountable systems of care and to the effective functioning of both internal and external systems of quality monitoring and improvement. Reliable and valid information about quality of care should be available to patients, purchasers, and accrediting organizations, but organizations also need to audit their structures and processes for problem areas and commit themselves to improving care on an ongoing basis. Problem areas include insufficient numbers and types of properly trained personnel, nonexistent or inadequate protocols for symptom assessment and management, and poor procedures for evaluating the need for patient transfers or referrals and for carrying out such transfers without harm to patients. The committee supports the development of guidelines for clinical practice that assist clinicians in preventing and relieving symptoms and in managing the end stages of specific diseases.

Policymakers and purchasers need to consider both the long-recognized deficiencies of traditional fee-for-service arrangements and the less thoroughly understood strengths and limitations of alternatives, including various kinds of capitated and per case payment methods that apply in diverse ways to individual practitioners, interdisciplinary care teams, specific institutions, networks of providers, or integrated systems of care. Particularly in need of attention are the levels of payment for home and nursing home visits by physicians, the interpretation of evaluation and management codes, the lack of diagnosis- or condition-related adjustments in hospice payments for both home and inpatient care, and other financing incentives that may discourage hospices or health plans from enrolling sicker patients. In addition, reimbursement methods and related incentives should encourage continuity of care that supports patients all the way through death and reduces disconnected and episodic care. Health care professionals and organizations also need to formulate ethical guidelines to guard against possible conflicts of interest in capitated and other care systems.

The problems with laws relating to prescribing practices are twofold. One problem is outdated, scientifically flawed laws and regulations themselves. The other problem is clinician, regulator, and public misunderstanding of the appropriate use of opioids. State medical societies, licensing boards, legislative committees, and other groups should cooperate to review state laws, regulations, board practices, and physician attitudes and practices to identify problem areas and then devise revisions in those statutes and regulations that unduly burden clinical management of pain. Regulatory review and revision should be accompanied by educational efforts to increase scientifically and clinically based knowledge and correct misperceptions about the appropriate medical use of opioids and about the biological mechanisms of opioid dependence, addiction, and pain management. Legal change should help—but cannot be relied upon alone—to correct undertreatment.

RECOMMENDATION 4: Educators and other health professionals should initiate changes in undergraduate, graduate, and continuing education to ensure that practitioners have the relevant attitudes, knowledge, and skills to care well for dying patients. Dying is too important a part of life to be left to one or two required (but poorly attended) lectures, to be considered only in ethical and not clinical terms, or to be set aside on the grounds that medical educators are already swamped with competing demands for time and resources. Every health professional who deals directly with patients and families needs a basic grounding in competent and compassionate care for seriously ill and dying patients. For clinicians and others to be held truly accountable for their care of the dying, educators must be held accountable for what they teach and what they implicitly and explicitly honor as exemplary practice. Textbooks and other materials likewise need revision to reflect the reality that people die and that dying patients are not people for whom "nothing can be done." As described in Chapter 8, a number of promising curriculum models exist, and others are being developed and tested.

RECOMMENDATION 5: Palliative care should become, if not a medical specialty, at least a defined area of expertise, education, and research. The objective is to create a cadre of palliative care experts whose numbers and talents are sufficient to (a) provide expert consultation and role models for colleagues, students, and other members of the health care team; (b) supply educational leadership and resources for scientifically based and practically useful undergraduate, graduate, and continuing medical education; and (c) organize and conduct biomedical, clinical, behavioral, and health services research. More generally, palliative care must be redefined to include prevention as well as relief of symptoms. Attention to symptoms should begin at earlier points during the trajectory of an illness because early treatment may well contribute to lessening pain at the end of life. The model for palliative care stresses interdisciplinary, comprehensive, and continuing care of patients and those close to them.

RECOMMENDATION 6: The nation's research establishment should define and implement priorities for strengthening the knowledge base for end-of-life care. The research establishment includes the National Institutes of Health, other federal agencies (e.g., the Agency for Health Care Policy and Research, the Health Care Financing Administration, the National Center for Health Statistics), academic centers, researchers in many disciplines, pharmaceutical companies, and foundations supporting health research. Their active support and involvement is necessary to advance basic and clinical research on the physiological mechanisms and treatment of symptoms common during the end of life, including neuropsychiatric problems. To extend understanding of quality-of-life issues in the treatment of advanced disease, those supporting clinical trials should encourage the collection of more information on the quality of life of those who die while enrolled in such trials. A further step is to support more research on the physiological mechanisms and treatment of symptoms common during the end of life, including neuropsychiatric problems. Pain research appears to supply a good model for this enterprise to follow. In addition, pathways need to be developed to further the dissemination and replication of proven health care interventions and programs. Demonstration projects to test new methods of financing and organizing care should be a priority for the Health Care Financing Administration. To encourage change in the attitudes and understandings of the research establishment, the committee urges the National Institutes of Health and other public agencies to take the lead in organizing a series of workshops, consensus conferences, and agenda-setting projects that focus on what is and is not known about end-stage disease and symptom prevention and treatment and that propose an agenda for further research. For the Agency for Health Care Policy and Research, the committee encourages support for the dissemination and replication of proven health care interventions and programs through clinical practice guidelines and other means.

RECOMMENDATION 7: A continuing public discussion is essential to develop a better understanding of the modern experience of dying, the options available to dying patients and families, and the obligations of communities to those approaching death. Individual conversations between practitioners and patients are important but cannot by themselves provide the supportive environment for the attitudes and actions that make it possible for most people to die free from avoidable distress and to find the peace or meaning that is significant to them. Although efforts to reduce the entertainment and news media's emphasis on violent or sensational death and unrealistic medical rescue have not been notably successful, a modicum of balance has recently been provided by thoughtful analyses; public forums; and other coverage of the clinical, emotional, and practical issues involved in end-of-life care. Regardless of how the current, highly publicized policy debate over physician-assisted suicide is resolved, the goal of improving care for those approaching death and overcoming the barriers to achieving that goal should not be allowed to fade from public consciousness. Much of the responsibility for keeping the public discussion going will rest not with the media but with public officials, professional organizations, religious leaders, and community groups.

Finally, the committee agreed that it was not prepared to take a position on the legality or morality of physician-assisted suicide. The issue should not, in any case, take precedence over those reforms to the health care system that would improve care for all dying patients. The goal of those who favor legalizing physician-assisted suicide is to promote patient autonomy, but true autonomy is not possible when significant numbers of people have the limited choice between suicide or continued suffering. If the laws permitting physician-assisted suicide are implemented at the state level, careful monitoring of their consequences for the quality of care and the public's trust in health care will be very important.




CONCLUDING THOUGHTS

Improving care at the end of life will require many changes in attitudes, policies, and actions. Such changes will involve a multitude of people and institutions that have a role in making and implementing decisions about patient care or in structuring the environments in which such decisions are reached and realized. Clearly, what patients and their families know, expect, and desire is important. Health care professionals play critical roles in diagnosis, communication, guidance and direction, treatment, negotiation, and advocacy for patients at many levels. Decisions by health plan managers, institutional administrators, and governmental officials shape and often impede the ability of patients, families, and clinicians to construct a care plan that serves the dying person well.

In general, changes in systems of care—not just individual beliefs and actions—are necessary if real gains are to be made in helping people live well despite fatal illness. Such widespread changes depend in part on a stronger social consensus on what constitutes appropriate and supportive care for those approaching death. Widely publicized—albeit not necessarily typical—instances of patient and family powerlessness to stop what they see as futile and painful treatments reflect a lack of such consensus. Paradoxically, this lack of consensus also is evident when patients or families demand treatments that practitioners see as useless, counterproductive, or even inhumane. It similarly reveals itself in a health care delivery and financing system that still rewards life-prolonging interventions (even when they will be ineffectual) and slights palliative and supportive services for those for whom life-extending treatment is neither helpful nor desired.

Freud may have been right that "our unconscious does not believe in its own death; it behaves as if immortal" (Freud, 1915, in Freud, 1959, p. 304). The committee was optimistic, nonetheless, that this society would cultivate the conscious intelligence and spirit to recognize the reality of death and the likelihood that it will bring distress. It likewise was optimistic that people would work together to create humane systems of care that assure the consistent use of existing knowledge to prevent and relieve suffering and that support efforts to provide people the right care at the right time in the right way. The analyses, conclusions, and recommendations presented here are offered with optimism that we can, individually and together, "approach" death constructively and create humane care systems that people can trust to serve them well as they die.