Lessons Learned From Project MATCH

Although Project MATCH failed to furnish support for the matching hypothesis in 15 of 16 instances, there are those who would argue that the results of this $27 million study support the empirical efficacy of substance abuse treatment. The problem with this argument is that Project MATCH used a selective group of participants—highly trained therapists—and failed to control for placebo or historical/maturational effects. Accordingly, the internal and external validity of pre–post differences in drinking behavior for MATCH participants is in doubt. While matching client characteristics to specific intervention strategies, as was done in Project MATCH, may not work, other forms of matching may prove more effective. Research on the therapeutic alliance is used to illustrate how matching clients to interventions based on commonalities in belief system may have a better chance of yielding meaningful results than matching client characteristics to specific treatment modalities.