Level 1: Likely reliable evidence
Female veterans with post-traumatic stress disorder (PTSD) and clear memory of the trauma that caused it at least three months before (N=284) were randomized to prolonged exposure (a type of cognitive-behavioral therapy) vs. present-centered therapy (a type of supportive therapy) for 10 weeks (JAMA. 2007;297:820-830). Both groups had therapy delivered by standard protocol in weekly 90-minute sessions. The prolonged-exposure group utilized imaginal exposure in sessions 3-10 while present-centered therapy focused on current life problems as manifestations of PTSD. The primary outcome was Clinician-Administered PTSD Scale (CAPS) score; outcome assessors were blinded. All women were included in intention-to-treat analysis, but 38 women (13%) were lost to follow-up.
Prolonged-exposure therapy was associated with a higher rate of treatment dropout compared with present-centered therapy (38% vs. 21%, P = .002) and fewer mean number of therapy sessions attended (8 vs. 9.3, P <.001). Comparing prolonged-exposure vs. present-centered therapy, mean CAPS score was 77.6 vs. 77.9 at baseline (not significant), 52.9 vs. 60.1 at 10 weeks (P <.05), 49.7 vs. 56 at three months (P <.05), and 50.4 vs. 54.5 at six months (not significant). The proportion of women who no longer met criteria for PTSD was 39% vs. 20.3% at 10 weeks (P =.002, NNT 6), 39% vs. 28% at three months (not significant), and 39.7% vs. 32.9% at six months (not significant).