Therapeutic alliance, cohesion and outcome in a long-term analytic group. A preliminary study.

We studied the ongoing relationship of patients and their therapist in a long-term, analytic group. The therapeutic alliance was rated weekly and group cohesion was rated every month, by patients and therapist. The patients' symptoms (SCL-90-R) and interpersonal problems (IIP-C) were rated every third month during therapy (self-reports). There was a steady increase in the alliance ratings by patients and therapist during the first 2 years of therapy. This differs from findings in short-term therapies, where the alliance quickly reaches a high level and then remains stable throughout therapy. Therapist ratings of early alliance correlated significantly with positive symptomatic outcome, but did not predict interpersonal change. Patients' alliance ratings did not predict change. Early cohesion ratings did not predict change. The concordance between the patients' and the therapist's alliance ratings was highest between 16 and 30 sessions, and was essentially maintained throughout therapy. An early concordance of patient and therapist alliance ratings predicted a better symptomatic outcome. The measures of therapeutic alliance and cohesion used in this study seem to address different elements in the group process.