Subsequent research has begun to define the parameters of importance. For example, the offset effect is particularly marked in the case of alcoholism34,43 and is considerably weaker among persons of lower socioeconomic status.43 One striking, well-controlled study (that of Levitan and Kornfeld42) showed that the introduction of a part-time psychiatrist into an orthopedic ward decreased the median duration of stay from 42 to 30 days among elderly patients who underwent surgical repair of the femur. In New York City at this time (1980), hospital costs averaged $200 per day, and the authors estimated that a reduction of this magnitude in length of stay resulted in savings of $55,200 over the six-month period while the cost of the psychiatrist was only $5,000. Given the costs of underdiagnosis of psychiatric disorders and the inadequacy of psychiatric treatment, introducing psychiatric programs into medical care of the elderly should realize major cost savings.
Four studies provide convincing evidence of the value of outpatient psychiatric treatment. McCaffree44 demonstrated that cost reductions for custodial care more than compensated for increased costs of active intervention in the Washington State mental hospital system. In addition, Cassel and colleagues13 and Endicott and coworkers15 found that community-based psychiatric treatment was as effective, and less expensive, than hospital-based treatment. Finally, the Hu research group35 showed that the costs of caring for an elderly demented person at home were no more than half the costs incurred in a nursing home.
One type of cost specific to depression is suicide. Stoudemire59 has estimated a total mortality cost of suicide resulting from depression at $4.2 billion, but the costs in human suffering are more difficult to estimate. Although official statistics show that suicide claims 30,000 lives in the United States each year,60 this figure is almost certainly larger owing to underreporting. Underreporting of suicide probably affects older persons more strongly than younger ones (R. Butler, Mount Sinai Medical Center, personal communication, June 1989) because physicians and the families of the elderly may be likely to ascribe death to causes other than suicide out of sentiment and compassion.
More than half of all suicides occur in persons suffering from depression.2 The increased risk of suicide among depressed persons of all ages is 30 times that of the non-affectively ill population; the lifetime risk of suicide of persons suffering from depression is 15