Terminal care preferences: hospice placement and severity of disease.

National Hospice Study data for 1981-82 were used to predict the location of care for terminal cancer patients. Sites of care were conventional care in hospitals, hospital-based hospice care, and hospice care in the home. Subjects were terminal cancer patients with a prognosis of less than 6 months of life who were attended by a primary concerned person. There were 1,732 patients 18-99 years old-293 conventional care, 612 hospital-based hospice care, and 827 hospice home care patients. Data sources were the patient, the primary concerned person, the family, and the medical record. Data were obtained at initial interview for the study, 1-week followup, reassessment every 2 weeks, and bereavement interviews. Information was grouped in the following categories: patient functional status, patient psychological outlook, symptomatology, medical condition, and characteristics of the primary concerned person and family. Conclusions were reached by univariate and multivariate analysis. First, a progression of functional disability was found to exist among care sites, from hospice home care for the least disabled to hospital-based hospice care to conventional care for patients with the greatest disabilities. The location of care was best explained by the patient's functional capacity. Second, the location of care was found to be poorly explained by extent of organ involvement or specific symptoms. Third, the primary concerned persons of patients under hospice home care experienced more stress but reacted no differently when compared with primary concerned persons at other care sites. Fourth, patients under hospice home care survived the longest and reported greater family closeness than other care groups.