Psychotherapy Versus Pharmacotherapy For Posttraumatic Stress Disorder: Systemic Review And Meta-Analyses To Determine First-Line Treatments

Authors

Daniel J. Lee, JBSA Fort Sam Houston
Carla W. Schnitzlein, Uniformed Services University of the Health Sciences
Jonathan P. Wolf, Walter Reed National Military Medical Center
Meena Vythilingam, Headquarters U.S. Marine Corps, Health Services
Ann M. Rasmusson, Boston University School of Medicine
Charles W. Hoge, Walter Reed Army Institute of Research

Date of this Version

2016

Citation

Depression and Anxiety 33:792–806, 2016.

Comments

U.S. Government Work

Abstract

Background: Current clinical practice guidelines (CPGs) for posttraumatic stress disorder (PTSD) offer contradictory recommendations regarding use of medications or psychotherapy as first-line treatment. Direct head-to-head comparisons are lacking. Methods: Systemic review of Medline, EMBASE, PILOTS, Cochrane Central Register of Controlled Trials, PsycINFO, and Global Health Library was conducted without language restrictions. Randomized clinical trials >8 weeks in duration using structured clinical interview-based outcome measures, active-control conditions (e.g. supportive psychotherapy), and intent-to-treat analysis were selected for analyses. Independent review, data abstraction, and bias assessment were performed using standardized processes. Study outcomes were grouped around conventional follow-up time periods (3, 6, and 9 months). Combined effect sizes were computed using meta-analyses for medication versus control, medication pre-/posttreatment, psychotherapy versus control, and psychotherapy pre-/posttreatment. Results: Effect sizes for trauma-focused psychotherapies (TFPs) versus active control conditions were greater than medications versus placebo and other psychotherapies versus active controls. TFPs resulted in greater sustained benefit over time than medications. Sertraline, venlafaxine, and nefazodone outperformed other medications, although potential for methodological biases were high. Improvement following paroxetine and fluoxetine treatment was small. Venlafaxine and stress inoculation training (SIT) demonstrated large initial effects that decreased over time. Bupropion, citalopram, divalproex, mirtazapine, tiagabine, and topiramate failed to differentiate from placebo. Aripiprazole, divalproex, guanfacine, and olanzapine failed to differentiate from placebo when combined with an antidepressant. Conclusions: Study findings support use of TFPs over nontrauma-focused psychotherapy or medication as first-line interventions. Second-line interventions include SIT, and potentially sertraline or venlafaxine, rather than entire classes of medication, such as SSRIs. Future revisions of CPGs should prioritize studies that utilize active controls over waitlist or treatment-as-usual conditions. Direct head-to-head trials of TFPs versus sertraline or venlafaxine are needed.