Residential drug abuse treatment research: are conventional control designs appropriate for assessing treatment effectiveness?

There has been renewed interest in residential drug abuse treatment research since the late 1980s. Earlier research employed control group designs. This article argues that for today's treatment research, the applicability of control groups created by either matched groups or by random sampling is limited. Four difficulties with control group designs are (1) client differences make assembling matched treatment and control groups untenable, (2) implementing random designs is practically impossible in field research, (3) there are ethical, political, and legal dilemmas in withholding treatment from a control group, and (4) randomly selected client populations will behave differently from the more realistic treatment circumstance involving client populations who are selected based on need and motivation. A more applicable perspective for drug abuse treatment research than the control group design needs to recognize that the client influences the treatment as well as that the treatment influences the client. Also, it must be understood that most clients view residential treatment as only one aspect of their life. Most importantly, individual differences must be taken into account when predicting outcome. A perspective that incorporates these concepts can use client history or baseline as its no-treatment comparative condition, and make use of multivariate analyses to help account for many other influences on outcome.