Examining military population and trauma type as moderators of treatment outcome for first-line psychotherapies for PTSD: A meta-analysis

Abstract

There is conflicting evidence as to whether military populations (i.e., veteran and active-duty military service members) demonstrate a poorer response to psychotherapy for posttraumatic stress disorder (PTSD) compared to civilians. Existing research may be complicated by the fact that treatment outcomes differences could be due to the type of trauma exposure (e.g., combat) or population differences (e.g., military culture). This meta-analysis evaluated PTSD treatment outcomes as a function of trauma type (combat v. assault v. mixed) and population (military v. civilian). Unlike previous meta-analyses, we focused exclusively on manualized, first-line psychotherapies for PTSD as defined by expert treatment guidelines. Treatment outcomes were large across trauma types and population; yet differences were observed between trauma and population subgroups. Military populations demonstrated poorer treatment outcomes compared to civilians. The combat and assault trauma subgroups had worse treatment outcomes compared to the mixed trauma subgroup, but differences were not observed between assault and combat subgroups. Higher attrition rates predicted poorer treatment outcomes, but did not vary between military populations and civilians. Overall, manualized, first-line psychotherapies for PTSD should continue to be used for civilians and military populations with various trauma types. However, greater emphasis should be placed on enhancing PTSD psychotherapies for military populations and on treatment retention across populations based on findings from this meta-analysis.

Keywords: Civilians; First-Line treatments; Meta-Analysis; Military; Posttraumatic stress disorder; Trauma type; Veterans.

Copyright © 2019 Elsevier Ltd. All rights reserved.

Figures

Figure 1.

PRISMA Flow Chart of Included Studies.

Figure 2.

Repeated Measures Treatment Effect Sizes with SAMD. N = 1,228; g = Hedges’ g; SE = Standard Error; SAMD = sample adjusted-meta-analytic deviance; ES = effect size. *The study was identified as an outlier based n the SAMD statistic > 2.25. The total random effect size increased, g = 1.46, SE = .08 , p < .001, with the outlier study (Schnurr et al., 2007) removed from the model.

Figure 3.

Repeated Measures Trauma Type Subgroup Analysis. g = Hedges’ g; SE = Standard Error; ES = effect size. Three studies were separated by trauma type for trauma type analyses. Two studies included some participants with non-combat traumas, but only the participants with combat trauma were included in the trauma type analysis due to small n in other trauma type categories (i.e., Rauch et al., 2009: n = 8; Thorp et al., 2012: n = 5). One study was separated by trauma type in trauma type analysis as follows: Nacash et al. (2011; n = 10) = Combat, Nacash et al. (2011; n = 5) = Mixed.

Figure 4.

Repeated Measures Population Type Subgroup Analysis. g = Hedges’ g; SE = Standard Error; ES = effect size. One study included military and civilian participants and therefore was separated by population for population type analyses. Lee et al. (2002) was separated by population type as follows: Lee et al. (2002-1; n = 4) = Military, Lee et al. (2002-2; n = 8) = Civilian.