Methods:  A longitudinal, prospective study was conducted with participants as they entered a community-based, residential substance abuse treatment program designed for women offenders.  All participants were interviewed within the first week of admission.  Participants completed the Addiction Severity Index (ASI), the Childhood Trauma Questionnaire (CTQ), and the Trauma Symptom Inventory (TSI).  Length of stay (LOS) for each study participant was tracked for six months (N = 105).  Bivariate analyses were used to determine relationships among pairs of variables.  Variables that were significantly correlated with length of stay from the bivariate tests were analyzed using multiple regression to identify the best predictive model for retention.

Results:  Predictors of length of stay were the woman’s age, current use of prescribed pain medication, sexual concerns, employment problems, and importance of drug treatment were significant predictors.  Substance use characteristics did not predict retention  Specifically, women remained in treatment longer if they were older, were not taking any prescription pain medication, reported concerns about employment, reported concerns about sexual problems, and reported lower importance of drug treatment, yet higher personal commitment to recovery.  Higher scores for childhood emotional trauma, emotional neglect, and physical neglect were correlated with retention, but were not statistically significant.

Conclusions:  Successful treatment programs must address women’s medical, employment, sexual and interpersonal relationships, and emotional trauma needs in this vulnerable population.