Influence of Patient and Therapist Agreement and Disagreement About Their Alliance on Symptom Severity Over the Course of Treatment: A Response Surface Analysis

The alliance is dyadic in its nature with both the patient and the therapist contributing. Relatively little is known about the effects of congruence between patient and therapist perception of alliance on treatment outcome. The current study investigated how patient and therapist agreement and disagreement about the alliance predict symptom severity over the course of long-term psychotherapy. We investigated N = 361 patients nested within N = 102 therapists longitudinally every 5th session across long-term treatment. Multilevel polynomial regression with response surface analysis was used to predict symptom severity five sessions later from congruence of the alliance ratings. Throughout treatment, patient and therapist agreement about stronger alliances significantly predicted lower subsequent patient-reported symptom distress. Patient and therapist disagreement was a marginally significant predictor of subsequent symptom distress. There was no significant difference in the effects of alliance agreement and disagreement on symptoms across time in long-term treatment. Findings support the importance of alliance agreement and disagreement as predictors of subsequent patient symptom severity. Public Significance Statement This study suggests that agreement between therapists and patients about the quality of their alliance may be related to lower patient symptoms later in treatment, while disagreement about the alliance quality could be related to more severe symptoms later in treatment. The study further suggests that alliance agreement and disagreement affect later symptoms steadily over the course of long-term psychotherapy. Thus, therapists need to direct their attention toward patients' potentially diverging perspective on the alliance and take appropriate actions to foster better agreement on their working relationship throughout treatment.