Quality of early working alliance in psychotherapy: diagnoses, relationship and intrapsychic variables as predictors

The aim of the study was the prediction of the quality of early working alliance, using possible predictors among patient pretreatment variables: diagnoses, current and past relationships and intrapsychic ones. Data are from the ongoing, naturalistic Norwegian Multisite Project on Process and Outcome of Psychotherapy (NMSPOP).

The sample, n = 270, is recruited from 15 outpatient clinics; 61.1% of the patients have personality disorders. Alliance was assessed with the Working Alliance Inventory (WAI), and predictors include independent clinicians' evaluations of diagnostic/interpersonal/intrapsychic characteristics and the patients' self-reports on similar and additional variables.

Four of 6 hypotheses were supported: Quality of working alliance is difficult to predict, early alliance is better predicted than later, diagnostic variables do not predict quality of working alliance, but quality of both current and past relationships is associated with working alliance. In a hierarchical multiple-regression analysis, 7% variance of working alliance in the 3rd session was explained from current relationship variables, whereas alliance in the 12th session was not predicted by the same model. Intrapsychic variables predicted the therapists' ratings of alliance, but not the patients' ratings.

The results are in line with previous research, and also with the theoretical model for working alliance.