Psychiatric comorbidity measures as predictors of retention in drug abuse treatment programs.

OBJECTIVE: To examine lifetime and current psychiatric comorbidity measures as predictors of drug abuse treatment retention, and to test the generalizability of results across treatment agencies in diverse settings and with varying practices. DATA SOURCES/STUDY SETTING: The national Drug Abuse Treatment Outcome Studies (DATOS), a longitudinal study of clients from 96 treatment agencies in 11 U.S. cities. STUDY DESIGN: The design is naturalistic and uses longitudinal analysis of treatment retention in long-term residential, outpatient drug-free, and outpatient methadone treatment modalities; client background (including psychiatric comorbidity) and program service provision are predictors. Clinical thresholds for adequate treatment retention were 90 days for long-term residential and outpatient drug-free, and 360 days for outpatient methadone. Psychiatric indicators included lifetime DSM-III-R diagnoses of depression/anxiety and antisocial personality, and dimensional measures of current symptoms for depression and hostility. DATA COLLECTION/EXTRACTION METHODS: Data include structured interviews with clients, a survey of treatment program administrators, and program discharge records. PRINCIPAL FINDINGS: Dimensional measures of current psychiatric symptoms emerged as better predictors than lifetime DSM-III-R diagnoses. In addition, the predictive association of hostility with retention varied significantly across treatment agencies, both in the long-term residential and outpatient drug-free modalities. Other notable findings were that on-site mental health services in long-term residential programs were associated with better retention for clients with symptoms of hostility. CONCLUSIONS: Assessment issues and stability of results across programs are important considerations for treatment research and practice.