The relative efficacy of bona fide psychotherapies for post-traumatic stress disorder: a meta-analytical evaluation of randomized controlled trials

Ulrich S. Tran, Bettina Gregor

In the treatment of PTSD, meta-analyses suggest comparable efficacy of cognitive behavioural therapies and various trauma focused treatments, but results for other treatments are inconsistent. One meta-analysis found no differences for bona fide therapies, but was critizised for overgeneralization and a biased study sample and relied on an omnibus test of overall effect size heterogeneity that is not widely used. We present an updated meta-analysis on bona fide psychotherapies for PTSD, contrasting an improved application of the omnibus test of overall effect size heterogeneity with conventional random-effects meta-analyses of specified treatment types against all others. Twenty-two studies were eligible, reporting 24 head-to-head comparisons in randomized controlled trials of 1694 patients. Head-to-head comparison between trauma focused and non-trauma focused treatments revealed a small relative advantage for trauma focused treatments at post-treatment (Hedges' g = 0.14) and at two follow-ups (g = 0.17, g = 0.23) regarding PTSD symptom severity. Controlling and adjusting for influential studies and publication bias, prolonged exposure and exposure therapies (g = 0.19) were slightly more efficacious than other therapies regarding PTSD symptom severity at post-treatment; prolonged exposure had also higher recovery rates (RR = 1.26). Present-centered therapies were slightly less efficacious regarding symptom severity at post-treatment (g = -0.20) and at follow-up (g = -0.17), but equally efficacious as available comparison treatments with regards to secondary outcomes. The improved omnibus test confirmed overall effect size heterogeneity. Trauma focused treatments, prolonged exposure and exposure therapies were slightly more efficacious than other therapies in the treatment of PTSD. However, treatment differences were at most small and far below proposed thresholds of clinically meaningful differences. Previous null findings may have stemmed from not clearly differentiating primary and secondary outcomes, but also from a specific use of the omnibus test of overall effect size heterogeneity that appears to be prone to error. However, more high-quality studies using ITT analyses are still needed to draw firm conclusions. Moreover, the PTSD treatment field may need to move beyond a focus primarily on efficacy so as to address other important issues such as public health issues and the requirements of highly vulnerable populations.