(PS1-57) Patient and therapist session-to-session perceptions of alliance ruptures and their effect on outcome in CBT for generalized anxiety disorder

Abstract Text: Ruptures in the patient-therapist alliance (i.e., breakdowns in the emotional bond and/or disagreements on therapeutic tasks or goals) are associated with poorer outcomes across various forms of psychotherapy for diverse mental health problems (Eubanks et al., 2018). However, when effectively repaired, ruptures may catalyze improvement, including to levels beyond patients who never experience a rupture (Eubanks et al., 2018). Of course, a first step toward repairing a rupture is that both members of a dyad recognize it. Yet, most research on this topic has included static, single-rater assessments of ruptures, thereby neglecting the dynamic and dyadic nature of the rupture-repair process. Addressing these gaps, this study replicates and extends prior research by examining the within-dyad associations among patients and therapists’ perceptions of alliance ruptures (i.e., degree of discrepancy or similarity) and the effect that such dyadic discrepancy/similarity has on next-session symptom change. Data derive from a trial in which 106 patients with generalized anxiety disorder received 14 sessions of cognitive-behavioral therapy (CBT), either alone or integrated with motivational interviewing (MI; Westra et al., 2016). Patients and therapists provided alliance rupture ratings after each session on a scale from 1 (“Not at all”) to 5 (“Constantly”) (Muran, 2009), and patients reported worry at baseline and after each session. Dyadic multilevel modeling revealed there was no average pattern of discrepancy between participants’ ratings (γ000 = 0.10, p = .10). However, the direction and degree of discrepancy varied significantly across dyads, such that some therapists tended to perceive more ruptures than their patient and vice versa. When either the patient (γ100 = 7.37, p < .001) or therapist (γ200 = 4.25, p < .001) independently perceived an alliance rupture, it was associated with poorer next-session outcomes. In contrast, when both the patient and therapist agreed on the presence of a rupture, its negative impact on the next session's outcome was mitigated (γ300 = -7.64, p < .001). Taken together, these findings highlight the importance of monitoring alliance ruptures throughout treatment, including with a brief measure, and suggest that patient-therapist agreement on the presence of a rupture might play a crucial role in mitigating its negative effects. A mutual awareness and acknowledgment of alliance breakdowns can facilitate a constructive repair process, whereas a hypervigilant therapist stance—where only the therapist recognizes the rupture—may be just as detrimental as failing to recognize a patient-perceived rupture.