To assess efficacy, comparative effectiveness, and harms of psychological and pharmacological treatments for adults with posttraumatic stress disorder (PTSD) and to update the original 2013 review.
MEDLINE, CINAHL, Cochrane Library, Cochrane Clinical Trials Registry, PILOTS (Published International Literature on Traumatic Stress), PsycINFO, and reference lists of published literature (May 2012–September 2017).
Two investigators independently selected, extracted data from, and rated risk of bias of relevant studies. We conducted meta-analyses or network meta-analyses using random-effects models when we had evidence from three or more studies with low heterogeneity. We graded strength of evidence (SOE) following established Agency for Healthcare Research and Quality guidance.
We included 193 randomized controlled trials (207 articles) for this review. Several psychological treatments were associated with the reduction of PTSD symptoms and loss of PTSD diagnosis compared with inactive comparators; high SOE supports efficacy of cognitive behavioral therapy (CBT)-exposure and CBT-mixed treatments, and moderate SOE supports efficacy of cognitive processing therapy (CPT), cognitive therapy (CT), eye movement desensitization and reprocessing (EMDR), and narrative exposure therapy (NET). When directly comparing two treatments of interest, moderate SOE favors CBT-exposure over relaxation therapy. Several pharmacological treatments reduced PTSD symptoms; moderate SOE supports the efficacy of fluoxetine, paroxetine, and venlafaxine compared with placebo. Our network meta-analysis (33 trials; N=4,817) of Clinician-Administered PTSD Scale (CAPS)-measured PTSD symptoms showed no differences in effectiveness between medications with at least moderate SOE of efficacy (fluoxetine, paroxetine, and venlafaxine) (low SOE for no difference). Studies provided insufficient strength of evidence for serious adverse events associated with any treatments of interest. The majority of psychological studies reported no information about adverse events. Among pharmacological treatments with evidence of efficacy (moderate SOE), we found increased risk of nausea with venlafaxine compared with placebo (moderate SOE). Our review found insufficient strength of evidence for the comparative effectiveness of any psychological versus pharmacological treatment and for differences in the efficacy or comparative effectiveness of treatments by patient characteristics (e.g., co-occurring conditions) or type, number, severity, or chronicity of trauma exposure(s). We did not find evidence for many of our outcomes of interest or interventions of interest, including the newer treatments added since our prior review.
Several psychological and pharmacological treatments have moderate to high SOE of efficacy for treating adults with PTSD. Future research is needed on the comparative effectiveness of treatments (including different comparisons of psychological and pharmacological treatments), differences in treatment benefits by trauma type or other patient characteristics, and adverse events associated with treatments.
