Therapeutic alliance and cohesion variables as predictors of outcome in short-term group psychotherapy

The current study explored the relative ability of aggregate therapeutic alliance and cohesion variables to predict short-term group therapy outcome. Data were collected from a comparative trial of two forms of time-limited group psychotherapy for complicated grief (Piper, McCallum, Joyce, Rosie, & Ogrodniczuk, 2001). The therapeutic alliance and elements of the cohesion construct were measured from the perspectives of each patient and the group therapist at intervals during the groups; scores were aggregated across assessments. Hierarchical multiple regression analyses, adjusting for the effects of treatment approach (interpretive vs. supportive) and specific group membership, demonstrated that the patient-rated alliance was a consistent predictor of outcome. Two cohesion measures, reflecting other participants' (therapist, other members) views of the patient's "fit" with the group, also accounted for variation in outcome. Implications of the findings for research and clinical practice, and the limitations of the measurement approach taken in this study, are considered.