Abstract
Objective: Little is known about how therapy processes relate to outcome in cognitive-behavioral therapy (CBT) for panic disorder (PD). This study examined whether client resistance predicts CBT for PD outcomes beyond the effects of established pre-treatment predictors. A secondary aim was to assess the consistency of resistance over treatment. Method: Data were from 71 adults participating in up to 24 biweekly sessions of CBT in a randomized controlled trial. Panic severity was assessed before, during (at Weeks 1, 5, and 9), and at termination of treatment (Week 12) using the Panic Disorder Severity Scale. Trained coders reliably rated resistance in videos of Sessions 2 and 10 using the Client Resistance Code. Results: Resistance was found to be moderately consistent (r = .64). Although overall resistance was unrelated to outcomes, hierarchical linear modeling revealed that openly hostile resistance at Session 10 predicted significantly diminished symptom change (r = .28, CI
95%
= [.01, .51]), beyond the effects of pretreatment predictors. Hostile resistance at Session 2 predicted attrition (r
rb
= −.30, p = .001), even after established predictors were controlled. Conclusions: Although some forms of resistance may be benign, openly hostile resistance is an important therapy marker that warrants increased clinical and research attention.