Nursing home candidates: hospital inpatient trial to identify those appropriately assignable to less intensive care.

Primary care physicians are often confronted with demands that elderly patients be institutionalized. Patients who appear to have no new medical problems present a therapeutic dilemma. Should they be admitted to a general hospital or should they be placed directly in a nursing home? This longitudinal prospective study involved 29 consecutive elderly patients judged to be in need of nursing home placement but who were without medical indications for admission. They were assessed and treated in a 30-bed ward of a general hospital. The mean length of stay was 19 days. Of the 29 patients, 8 (24 percent) benefited from the hospitalization; the outcome was placement and retention at a level of care less intensive than that in a nursing home. On admission to this special ward, the presence of 2 of the following 3 characteristics identified the patients who would benefit from this preplacement hospitalization (with 75 percent sensitivity and 86 percent specificity): 1) a score of A or B (Katz Scale) for activities of daily living; 2) a score of 3 or fewer errors on the mental status scale (Pfeiffer); and 3) the presence of family members willing to care for the patient, although unable to do so at the time of admission. These 3 factors constitute a screening tool to differentiate elderly patients who will benefit most under a regimen of intensive rehabilitation from those who will be inevitable recipients of long-term care.