This is a study of the effect of a crisis respite alternative to traditional psychiatric hospitalization within a system of care based in an urban community setting. Previous research has determined crisis respite alternatives to be an effective and cost saving alternative to traditional psychiatric hospitalization for many psychiatric patients. This project will be directed towards people with prolonged mental illness and/or those who have a psychiatric condition that makes them at risk for psychiatric hospitalization, who are poor, and who are deemed to require psychiatric hospitalization by an experienced staff of crisis/evaluation clinicians. Patients will be randomly assigned to traditional psychiatric hospitalization or a crisis respite/day hospital alternative. Patients with dual diagnoses of substance abuse and psychiatric illness will be included. This study explores hypotheses addressing the differential effect of traditional hospitalization and crisis respite based on patient diagnosis and other clinical and demographic features (race, gender, multiple user of services, etc.), the costs of respite care, and the mechanism of action. Specifically, we hypothesize that those patients who do not have a schizophrenic diagnosis assigned to crisis respite will experience superior outcome in terms of social adjustment, family burden, and self esteem measures than those assigned to traditional impatient care. However, along dimensions of symptomatic improvement, we hypothesize that those patients with a schizophrenic disorder will do better with hospitalization than with crisis respite. We propose that the expected greater social adjustment of those assigned to respite care will be correlated with the capacity for more rapid re-integration within the natural network of family and work. In addition to patient level outcome data, systematic measures of cost to the State and impact on the system of care will be made to determine the cost and benefits of implementing a crisis respite program to the system of care as well as to individual patients. MANOVA procedures with demographics, patient diagnosis, social integration, and treatment condition as independent variables or covariates will be modeled for outcome measures of symptomatic change, utilization costs, social adjustment after treatment, family burden, patient satisfaction, and self-esteem measures.
