Leveraging observational data to identify in‐session patient and therapist predictors of cognitive processing therapy response and completion

Cognitive processing therapy (CPT) is an evidence-based treatment for posttraumatic stress disorder (PTSD), but little is known about in-session process variables that predict symptom reduction and treatment completion during CPT. Examining potentially malleable factors that may promote or impede recovery can inform care delivery and enhance outcomes. The current study used observational ratings of CPT session recordings to examine in-session patient and therapist factors in cognitive, affective, and interpersonal domains to identify their relative contributions to predicting symptom outcomes and treatment completion. Participants were 70 adult survivors of interpersonal violence who received CPT. Predictors of better posttreatment PTSD outcomes included less patient fear, β = .32, and less patient avoidance of engaging with the therapist, β = .35. When using the last available PTSD score, less fear, β = .23, and avoidance, β = .28, continued to predict better outcomes, and more patient cognitive flexibility emerged as a stronger predictor of outcome, β = -.33. Predictors of a higher likelihood of treatment completion included more therapist use of Socratic dialogue, OR = 6.75, and less therapist encouragement of patient affect, OR = 0.11. Patient sadness and anger and therapist expression of empathy did not predict symptom outcomes or treatment completion versus dropout. The results highlight the importance of patients' cognitions, emotions, and engagement with their therapist in CPT as well as the role of therapist behaviors in patient completion of treatment.