Meta-analysis of day treatment and contingency-management dismantling research: Birmingham Homeless Cocaine Studies (1990-2006).

Four successive randomized clinical trials studying contingency management (CM), involving various treatment arms of drug-abstinent housing and work therapy and day treatment (DT) with a behavioral component, were compared on common drug abstinence outcomes at 2 treatment completion points (2 and 6 months). The clinical trials were conducted from 1990 to 2006 in Birmingham, Alabama, with a total of 644 homeless persons with primary crack cocaine addiction. The meta-analysis utilized the weighted least squares approach to integrate data encompassing 9 different treatment arms to assess the effects of CM and DT (neither, DT only, CM only, and CM = DT) on a common estimate of prevalence of drug abstinence. Taken together, the results show much stronger benefits from CM = DT and from CM only than for DT alone. Throughout all of the Birmingham Homeless Cocaine Studies, the CM = DT consistently produced higher abstinence prevalence than did no CM.