Although much has been written about the need for advance medical planning, only 33 percent of nursing home residents had living wills or health care proxies in one 1993 survey. Although establishing a patient's wishes for do-not-resuscitate (DNR) orders is important, a better approach to the discussion might be to have the patient rank three common goals of care as a basis for discussing his or her wishes. Gillick and colleagues conducted a prospective, descriptive study that examined three goals of care and five patterns of care to help patients plan for end-of-life care.

The authors identified three principles of end-of-life care: (1) prolongation of life, (2) maintenance of function, both physical and cognitive, and (3) maximization of comfort. All clinically relevant rankings of these three goals were considered, leading to five major patterns of care—intensive, comprehensive, basic, palliative and comfort only (see accompanying table). In the intensive pattern of care, prolongation of life was ranked first, followed by maintenance of function and comfort. In the comprehensive pattern, the order was maintenance of function, prolonged life and comfort; in the basic care pattern, the order was maintenance of function, comfort and prolonged life; in the palliative pattern, the order was comfort, maintenance of function and prolonged life; and in the comfort-only pattern, the only goal was comfort.

The residents of a nursing home and their families attended a meeting to discuss these goals and patterns for end-of-life care. Within two weeks of the meeting, the residents or their surrogates were asked to determine their individual priorities for end-of-life care. Of the 38 patients who participated in the study, 21 participated directly and 17 were represented by their surrogates. Overall, 21 percent of study participants selected intensive care (33 percent of patients versus 6 percent of surrogates); 16 percent selected comprehensive care (28 percent of patients, zero percent of surrogates); 18 percent selected basic care (5 percent of patients, 35 percent of surrogates); 18 percent selected palliative care (5 percent of patients, 35 percent of surrogates); and 6 percent chose comfort care only (zero percent of patients, 12 percent of surrogates). Twenty-eight percent of patients and 12 percent of surrogates were unable to decide on a pattern of care or stated that they did not want think about these issues.

The authors conclude that while current DNR orders are important, they address only a small proportion of the medical decisions that must be made by or for a nursing home resident. The authors also acknowledge that this study did not determine whether the goals set were subsequently used to guide treatment decisions. As the nursing home population increases, further studies such as this one are clearly needed.

editor's note: A variety of devices can facilitate the discussion of advance medical care. This study, which helps patients view various treatments (or lack thereof) as supporting one of three possible goals of care (prolonging life, maintaining function or maximizing comfort), is an elegant way to conceptualize a problem that is sometimes difficult for both the physician and the patient or surrogate. Of particular note is the large discrepancy between the patient's wishes and the surrogate's wishes. Although this result could be explained by a variety of study factors, it underscores the need to ask the surrogate what the patient would have wanted, not what the surrogate wants.—g.b.h.