Therapist effects in substance abuse treatment: a naturalistic study

Objective: The purpose of this naturalistic study was to investigate between-therapist variation in the working alliance, treatment effectiveness and client satisfaction. Methods: The subjects were Finnish outpatients (N = 327) with various substance use disorders and their therapists (N = 33). The clients were randomly assigned to the therapists, and they were followed-up for 6 months. Regression analyses with mixed models were used. Results: The findings indicated that between-therapist variation was wider in the therapists' ratings of the alliance than in the clients' ratings. The therapist effect was present at follow-up in prolonged abstinence period and client satisfaction with the therapist. According to the clients' ratings of the alliance and post-treatment satisfaction, the therapist effect on satisfaction in general increased as treatment progressed. The therapists' and clients' ratings of the alliance differed significantly from each other, the clients' ratings being more positive than the therapists' ratings, although correlations showed that they were significantly interrelated. Conclusions: It can be concluded that the size of the therapist effect varied with regard to the treatment aspects measured.