Psychiatric consultations with depressed medical inpatients: a randomized controlled cost-effectiveness study.

Nonspecific, supportive psychiatric consultations were performed with a random sample of thirty-three general medical inpatients scoring thirteen or more on the Beck Depression Inventory. The control group consisted of thirty-five patients, matched for sex, marital status, somatic history, and seriousness of illness. The number of patients receiving no analgesic and/or psychotropic medication in the consult group (39%) was significantly greater than that in the control group (17%). When compared with their mean BDI score on admission, the BDI score just before discharge had decreased significantly in the consult group (from 20 to 13), but not in the control group (from 19 to 16). Probably because the patient sample was too heterogeneous, with too low a prevalence of mental disorders (45%), a significant reduction in other medical care expenditures and in length of hospital stay could not be demonstrated.