Adolescent substance abuse: predictors of treatment outcomes

The purpose of this study was to explore the various drug, demographic, participant, and treatment characteristics predicting treatment completion and other treatment outcomes for 1148 adolescents receiving publicly funded substance abuse treatment. The data for the present study were extracted from a larger study evaluating substance abuse treatment that was funded by Medicaid in a midwestern state. Stepwise regression analyses indicated that treatment characteristics (e.g., multiple entries in treatment and type of treatment) predicted frequency of drug use and days missed at school during treatment. However, treatment characteristics failed to predict participants\u27 arrests during treatment. Participants\u27 arrest history and gender predicted arrests that occurred during treatment. Discharge status was predicted by characteristics of the treatment program and characteristics that participants brought to the treatment. For example, being white and being referred from the criminal justice system predicted completion of treatment. Use of marijuana, being male, having repeated entries in treatment, and having multiple arrests predicted involuntary termination of treatment. And, those participants who were older in age at admission were more likely to drop out of treatment. There were many gender differences found in the initial analyses using ANOVA and Chi-square statistics. For example, compared to males, females tended to miss more days at school, had multiple hospitalizations, and entered the treatment at an earlier age. Males tended to have multiple arrests, were more likely to live with families, and started using drugs at an earlier age than did females. However when combined with other variables in regression analyses gender was suppressed and only predicted number of arrests during treatment