Follow-up contact bias in evaluation of substance abuse treatment programs

This study examined the problem of follow-up contact bias in evaluation of substance abuse treatment programs using administrative data sets. Complete d discharges in Iowa were compared to clients contacted 6 months after trea tment. The percent of clients reporting no substance use doubled between ad mission to treatment and follow-up. Unemployment, arrests, and number of da ys of work or school missed declined. However, clients included in the foll ow-up sample were significantly different in terms of risk factors for adve rse outcomes. The authors conclude that improvements in social functioning 6 months after discharge cannot be generalized to the entire population of clients treated, since those not found have a different combination of risk factors.