Who improves in trauma-focused treatment: A cluster analysis of treatment response in VA patients undergoing PE and CPT

Background: Although most veterans with posttraumatic stress disorder (PTSD) benefit from evidence-based treatments, questions persist concerning the profiles of those at risk for poor outcomes. To help address these gaps, this study analyzed a large clinical cohort of veterans receiving prolonged exposure (PE) or cognitive processing therapy (CPT). Methods: Cluster analysis using Ward's method with Euclidian distances identified clinically meaningful sub-groups of veterans in a national cohort (n = 20,848) using variables maintained in the electronic medical record. The clusters were then compared via one-way analysis of variance and Tukey's HSD on indicators of treatment progress including PTSD symptom change, clinical recovery, clinically significant change, remission, and treatment completion. Results: Effect size differences on clinical outcome measures for PE and CPT were negligible. Less than half of veterans achieved at least a 15-point reduction in PCL-5 score and half completed treatment. We identified 10 distinct clusters. Higher rates of PTSD service-connected disability were linked to poorer outcomes across multiple clusters, especially when combined with Post-Vietnam service era. Non-White race was also linked with poorer clinical outcomes. Factors associated with better outcomes included a greater proportion of female vet-erans, especially when combined with recent service era, and longer PTSD diagnosis duration. Conclusions: This study suggests the need to improve PTSD treatment outcomes for non-White and male veterans, examine treatment response in Post-Vietnam era veterans, and consider ways in which the service connection process could hinder treatment response. The results from this study also indicate the benefits of integrating elements of clinical complexity into an analytic approach.