Mental Health Indicator Interaction in Predicting Substance Abuse Treatment Outcomes in Nevada

Background: Indicators of co-occurring mental health and substance abuse problems routinely collected at treatment admission in 19 State substance abuse treatment systems include a dual diagnosis and a State mental health (cognitive impairment) agency referral. These indicators have yet to be compared as predictors of treatment outcomes. Objectives: 1. Compare both indices as outcomes predictors individually and interactively. 2. Assess relationship of both indices to other client risk factors, e.g., physical/sexual abuse. Methods: Client admission and discharge records from the Nevada substance abuse treatment program, spanning 1995–2001 were reviewed (n = 17,591). Logistic regression analyses predicted treatment completion with significant improvement (33%) and treatment readmission following discharge (21%). Using Cox regression, the number of days from discharge to treatment readmission was predicted. Examined as predictors were two mental health indicators and their interaction with other admission and treatment variables controlled. Results: Neither mental health indicator alone significantly predicted any of the three outcomes; however, the interaction between the two indicators significantly predicted each outcome (p < .05). Having both indices was highly associated with physical/sexual abuse, domestic violence, homelessness, out of labor force and prior treatment. Conclusions and Scientific Significance: Indicator interactions may help improve substance abuse treatment outcomes prediction.