Good deaths, bad deaths, and preferences for the end of life: a qualitative study of geriatric outpatients.

OBJECTIVES: Patient involvement in decision-making has been advocated to improve the quality of life at the end of life. Although the size of the oldest segment of the population is growing, with greater numbers of older adults facing the end of life, little is known about their preferences for the end of life. This study aimed to explore the attitudes of older adults with medical illness about the end of life, and to investigate whether current values could be extended to end-of-life preferences.

DESIGN: Descriptive study with interviews using open- and closed-ended questions.

SETTING: Patients attending two university-affiliated geriatric clinics were interviewed in a private conference room near the clinic they attended or in their homes.

PARTICIPANTS: Sixteen older men and women identified by their physicians as having nonterminal heart disease or cancer.

MEASUREMENTS: The interview contained open-ended questions such as: "What are the most important things in your life right now?" and "What would you consider a good/bad death?" The interview also contained closed-ended questions about symptoms, quality of life, and health status. Additional questions elicited preferences for the end of life, such as location of death and the presence of others. The open-ended questions were tape-recorded, transcribed, and analyzed using qualitative methods. The closed-ended questions were analyzed using descriptive statistics.

RESULTS: Patients with heart disease and cancer provided similar responses. Participants' views about good deaths, bad deaths, and end-of-life scenarios were heterogeneous. Each participant voiced a unique combination of themes in describing good and bad deaths. Because each participant described a multifaceted view of a good death, for instance, no theme was mentioned by even half of the participants. Participants provided differing explanations for why given themes contributed to good deaths. Currently valued aspects of life were not easily translated into end-of-life preferences. For example, although the majority of participants identified their family as being important, many gave reasons why they did not want their family members present when they died.

CONCLUSIONS: Because of the heterogeneity of views and the difficulty in inferring end-of-life preferences from current values, older adults should be asked not only questions about general values, but also specific questions about their end-of-life choices and the reasons for these choices. A thorough understanding of an individual's end-of-life preferences may help health professionals working with older adults develop patient-centered care plans for the end of life.