Methodological issues in the evaluation of treatment of drug dependence

Studies of treatment outcome for drug dependence are methodologically inferior to those for alcohol dependence, probably because scientific interest in drug use is more recent. Also, the behavior of drug dependence is very complex, since multiple drug use is normative for psychoactive drugs. Compared to clients in alcohol and tobacco studies, those in studies of drug dependence are more heterogeneous in their drug use. It is recommended that research priority be given to the identification of homogeneous subsets of the drug user population, including users of alcohol and tobacco. Several methods of measuring drug use have been advocated, but all have limitations, and little is known about their impact on the interpretation of treatment outcome. Recently the quality of designs in this area of research has improved, but avoidance of randomized trials continues to retard the acquisition of knowledge about the relative effectiveness of treatment procedures. Variability in the duration of follow-ups also impedes interpretation of findings; a minimum of six months post treatment is recommended. It is concluded that measures of drug use should be the principal outcome in evaluations of treatment effectiveness, that the role of therapist variables deserves research attention, and that criteria be developed for the interpretation of nonabstinent outcomes, which are the norm of clients treated for drug dependence.