Medical decision-making in the last six months of life: choices about limitation of care.

OBJECTIVE: To characterize the limitation of care in routine geriatric practice in advance of and at the time of a patient's final episode of illness.
DESIGN: A descriptive study performed by retrospective chart review.
SETTING: An outpatient geriatric practice affiliated with a community teaching hospital. PATIENTS: Fifty-nine recipients of primary care who were community-dwelling and older than 65, died in the years 1988-1991, and were enrolled in the practice for at least 6 months prior to death. MEASUREMENTS: We recorded the type(s) of care patients (or, in the case of incompetence, their families) and their physicians chose to limit during the last episode of illness preceding death and during previous episodes of illness by examining those instances when therapy other than that considered "standard" was given. We also examined whether the presence of dementia, functional impairment, chronic disease, terminal illness, site of routine care (home vs hospital), and location of death were associated with the limitation of care.
RESULTS: A choice to limit diagnostic tests or treatment was made by the patient or surrogate in 40% of the 59 patients during the 6 months before the patient's final episode of illness. Most frequently limited were diagnostic tests, surgery, and hospitalization for purposes other than surgery. Terminal illness and location of death were associated with the limitation of care, but dementia, functional impairment, chronic illness, and location of care were not. By comparison, 89% of the patients had limitation of care during the final episode of illness, and more aggressive therapies such as cardiopulmonary resuscitation and intubation constituted the majority of therapies withheld.
CONCLUSIONS: In one geriatric practice, care is frequently limited before a patient's final illness in the course of routine practice. In contrast to recent discussion focusing on limitation of end-of-life interventions or interventions in the severely impaired, these results suggest that there are multiple points in the course of a community-dwelling elderly patient's illness at which choices about level of care can be made. Given this opportunity, a significant number of elderly patients of their surrogates will choose less intensive therapy.