Risk factors for dropout from psychological substance use disorder treatment programs in criminal justice settings

Background: Substance use disorders are highly prevalent in people within the criminal justice system. Psychological programs are the most common type of treatment available and have been shown to decrease recidivism, but dropping out of treatment is common. Risk factors associated with treatment dropout remain unclear in this setting, and whether the risk factors differ by treatment form (group -based vs. individual). Methods: Outcome (treatment dropout) was defined as not finishing the program due to client's own wish, misbehavior, no-shows, or because program leader found client to be unsuitable. Predictors of treatment dropout included a comprehensive set of individual -level clinical, socioeconomic, and crime -related pre-treatment characteristics. Multivariable regression models were used to estimate the associations between predictors and dropout. Findings: The study cohort included 5239 criminal justice clients who participated in a psychological treatment program (group -based or individual). Multivariable logistic regression models showed that female sex (OR =1.64, 95% CI 1.20-2.25), age (0.99, [0.97-1.00]), sentence length (0.98, [0.97-0.98]), higher education (0.54, [0.28-1.00]), number of violent offenses (1.03, [1.01-1.05]), and anxiety disorders (1.32, [1.01-1.72]) were associated with dropout from the individual treatment program. For the group -based program, age (OR =0.98, 95% CI 0.96-1.00), sentence length (OR =0.96, 95% CI 0.94-0.98), stimulant use disorder (OR =1.48, 95%, 1.00-2.19), and self -harm (OR 1.52, 95% CI 1.00-2.34) were associated with dropout. Conclusions: We identified certain sociodemographic, crime -related, and clinical characteristics that were particularly important in predicting dropout from psychological treatment. Further, we find that there are similarities and differences in predictors of dropout from group -based and individual treatment forms.