Predictors of retention in a homeless veteran intervention program

Abstract : This study assessed the value of a specific set of client variables in predicting treatment phase retention and overall program completion of an intervention treatment program for homeless veterans. Treatment programs can improve the lives of homeless individuals, but many participants leave prematurely. Certain characteristics, which are evident upon admission, may help to identify those veterans at greater risk of early discharge and program failure. Spanning three years (2009-2011), the records of 680 unique clients from a homeless veterans program were reviewed. Logistic regression models yielded significant association between treatment completion, graduation, and certain observable client characteristics. Ultimately, information regarding a client's mental health, chronic health, and the client's immediate prior residence before admission to the Veteran Rehabilitation Center program proved to be significant independent predictors of premature discharge from treatment. From the significant covariates for the treatment model, a simple, scoring-scheme heuristic was developed to enable treatment providers to expeditiously and accurately assess the relative risk of premature discharge among a cohort of veteran clients. Using demographic information from three simple questions, the scoring scheme has a 98.3% correlation to the theoretical probability of failure from treatment and gives providers a simple and accurate way to identify those at greater risk of early exit. These results can inform targeted intervention strategies to maximize program effectiveness and efficiency. They provide a decision support tool to help high-risk veterans remain engaged in treatment, attain treatment goals, graduate, and fully prepare them to reintegrate into a sober, self-sustaining lifestyle.