Effects of perceived working alliance quality and congruence on naturalistic psychotherapy outcome: A response surface analysis

Psychother Res. 2025 Sep 22:1-14. doi: 10.1080/10503307.2025.2553636. Online ahead of print.

ABSTRACT

OBJECTIVE: Patient- and therapist-reported (i.e., dyadic) working alliance evaluations have been identified as robust predictors of psychotherapy outcome. However, it remains unclear how their congruence relates to outcome and how treatment duration, a variable factor in naturalistic psychotherapy, influences this relationship. This study examined the effects of dyadic alliance judgments and their congruence on multidimensional outcomes, further assessing the role of treatment duration.

METHOD: N = 353 outpatients undergoing cognitive behavioral therapy and their N = 95 therapists provided alliance ratings at treatment onset, with patients reporting distress, depression, and quality of life pre- and post-treatment. Polynomial regressions with response surface analyses were used to examine the relationships between alliance, congruence, and outcome, further controlling for session count.

RESULTS: Dyadic alliance ratings were moderately correlated, with patient-reported ratings exceeding those of therapists. Therapist-reported but not patient-reported alliance predicted greater reductions in depression and distress. Higher levels of jointly high alliance ratings, but not congruence, were associated with improvements in distress and depression. There were no incongruence effects. Controlling for session count yielded identical result patterns.

CONCLUSION: These findings highlight the value of alliance ratings and their alignment in predicting psychotherapy outcome, emphasizing the need to consider their interplay, regardless of treatment duration.

PMID:40981945 | DOI:10.1080/10503307.2025.2553636