Providing intensive addiction/housing case management to homeless veterans enrolled in addictions treatment: A randomized controlled trial.

This study sought to determine whether homeless veterans entering Veterans Affairs (VA) substance use treatment randomized to intensive addiction/housing case management (AHCM) had improved housing, substance use, mental health, and functional outcomes and lower acute health care utilization, compared to a housing support group (HSG) control. Homeless veterans (n = 181) entering outpatient VA substance use treatment were randomized to AHCM and HSG and received treatment for 12 months. AHCM provided individualized housing, substance use and mental health case management, life skills training, and community outreach. The control condition was a weekly drop-in housing support group. Adjusted longitudinal analyses compared groups on baseline to month 12 change in percentage of days housed and functional status, substance use, and mental health outcomes (36-Item Short-Form Health Survey; Addiction Severity Index [ASI]). Negative binomial regression models compared groups on health care utilization. Both conditions significantly increased percentage of days housed, with no differences detected between conditions. In total, 74 (81.3%) AHCM and 64 (71.1%) HSG participants entered long-term housing (odds ratio = 1.9, 95% confidence interval [0.9, 4.0], p = .088). HSG participants experienced a greater decrease in emergency department visits than AHCM (p = .037), whereas AHCM participants remained in substance use treatment 52.7 days longer (p = .005) and had greater study treatment participation (p < .001) than HSG. ASI alcohol composite scores improved more for HSG than AHCM (p = .006), and both conditions improved on ASI drug and psychiatric scores and alcohol/drug abstinence. AHCM did not demonstrate overarching benefits beyond standard VA housing and substance use care. For those veterans not entering or losing long-term housing, different approaches to outreach and ongoing intervention are required. (PsycINFO Database Record