for it. Many pediatric patients, for example, spend years—and with better treatments, even decades—with illnesses such as cystic fibrosis, complex congenital malformations, and neurodegenerative disorders. Although families hope for a cure to emerge from scientific research, premature death (i.e., less than normal life expectancy) is a likely prospect. It may occur suddenly, for example, from a superimposed viral illness, or it may come after an extended but clearly evident physical decline. Chapters 3 and 4 further consider variations in the dying process and the implications for care at the end of life.

The diagnosis of incurable, progressive disease that is expected to prove fatal is among the most difficult and sobering judgments that physicians make. Such a determination generally signals the need for patients, families, and clinicians to reconsider clinical and personal priorities. It is also currently important because Medicare coverage for hospice services, which are designed specifically for dying patients, requires a determination that a beneficiary has a terminal illness and, perhaps more intellectually daunting, has a life expectancy of six months or less.

Moving from a clinical diagnosis of a terminal or incurable illness expected to end in death to a prognosis—a prediction of the course of the illness and remaining life expectancy—is an exercise in uncertainty. (See Appendix D and Chapter 3 for further discussion.) One approach to prognosis uses explicit clinical criteria and formulas to generate a calculated probability that a person will live some defined period of time (e.g., a 50 percent probability of living six months) (McClish and Powell, 1989; Lynn et al., 1996). Even very near the actual time of death, however, prognosis is often imprecise. Some will die more quickly than expected; others will die more slowly.

Uncertainty about prognosis and methods for establishing it may have significant policy as well as personal implications. For example, "the Medicare hospice benefit, notably, manages to give a date" but not an expected rate or probability of surviving for six months (Lynn et al., 1996, p. 315). When insurance or other benefits are contingent on the accuracy of prognoses, physicians' judgments may be reviewed or even "audited" by hospice personnel or government officials who may not understand the uncertainty inherent in projecting survival. (See Chapter 6.)

As will be discussed throughout this report, hospice and palliative care are responses to perceived inadequacies in the prevention and relief of

Front Matter (R1-R18)

Summary (1-13)

1 Introduction (14-32)

2 A Profile of Death and Dying in America (33-49)

3 Caring at the End of Life (50-86)

4 The Health Care System and The Dying Patient (87-121)

5 Accountability and Quality in End-Of-Life-Care (122-153)

6 Financial and Economic Issues in End-Of-Life Care (154-187)

7 Legal Issues (188-206)

8 Educating Clinicians and Other Professionals (207-234)

9 Directions for Research to Improve Care at the End of Life (235-258)

10 Conclusions and Recommendations (259-271)

References (272-312)

Appendix A (313-320)

Appendix B (321-326)

Appendix C (327-357)

Appendix D (358-362)

Appendix E (363-382)

Appendix F (383-399)

Appendix G (400-404)

Appendix H (405-406)

Appendix I (407-412)

Appendix J (413-418)

Index (419-437)