A systematic review of predictors and moderators of improvement in cognitive-behavioral therapy for panic disorder and agoraphobia

Background: Despite the considerable efficacy of cognitive-behavioral therapy (CBT) for panic disorder (PD) and agoraphobia, a substantial minority of patients fail to improve for reasons that are poorly understood. Objective: The aim of this study was to identify consistent predictors and moderators of improvement in CBT for PD and agoraphobia. Data sources: A systematic review and meta-analysis of articles was conducted using PsycInfo and PubMed. Search terms included panic, agoraphobi*, cognitive behavio*, CBT, cognitive therapy, behavio* therapy, CT, BT, exposure, and cognitive restructuring. Study selection: Studies were limited to those employing semi-structured diagnostic interviews and examining change on panic- or agoraphobia-specific measures. Data extraction: The first author extracted data on study characteristics, prediction analyses, effect sizes, and indicators of study quality. Interrater reliability was confirmed. Synthesis: 52 papers met inclusion criteria. Agoraphobic avoidance was the most consistent predictor of decreased improvement, followed by low expectancy for change, high levels of functional impairment, and Cluster C personality pathology. Other variables were consistently unrelated to improvement in CBT, understudied, or inconsistently related to improvement. Limitations: Many studies were underpowered and failed to report effect sizes. Tests of moderation were rare. Conclusions: Apart from agoraphobic avoidance, few variables consistently predict improvement in CBT for PD and/or agoraphobia across studies. •Agoraphobic avoidance predicted less improvement from pre- to post-treatment.•Functional impairment and low expectancy for change predicted less improvement.•Comorbid depression and medication use consistently did not predict improvement.•Few studies examined moderators of improvement in CBT vs. other treatments.