Depression and substance use: The CES-D's utility in predicting treatment outcomes in a longitudinal multi-site study of residential treatment centers.

This investigation utilized a large sample of individuals undergoing substance use treatment to examine the CES-D's structural validity and establish its predictive utility relative to treatment discharge. The sample included 5750 individuals who were receiving substance abuse treatment at 19 different residential treatment facilities. The CES-D was administered to participants over the course of their time in a residential inpatient substance use treatment program. The present study used a split sample method to conduct both exploratory (EFA) and confirmatory (CFA) factor analyses. Results of both the EFA and CFA indicated that the CES-D contains three first order factors measuring negative mood, positive affect, and interpersonal interactions as well as a second-order high score factor which can be used to guide interpretation and symptom monitoring. Individuals receiving residential treatment fell into one of three distinct groups based on their CES-D scores, and that their scores (and subsequent trajectories of scores over time) were differentially related to discharge status. Implications for practice and the utility of the CES-D within substance use populations are discussed.