Predicting rehospitalization of persons with severe mental illness

Patients with severe and persistent mental illness who were discharged from a state hospital were followed for approximately four years to identify predictors of rehospitalization. Length of time in the community between discharge and readmission was the main outcome variable. Clinical interviews, hospital medical records, and a statewide data communications network which tracks client movement throughout the service delivery system were used to collect data. Race, number of prior psychiatric hospitalizations, type of aftercare residence, and assignment to a residential program were all predictors of who remained in the community the longest. Severity of psychopathology, diagnosis, and degree of family support were not good predictors of community stability.