In the article, “Support Mechanisms and Vulnerabilities in Relation to PTSD in Veterans of the Gulf War, Iraq War and Afghanistan Deployments: A Systematic Review,” by Breanna K. Wright, Helen L. Kelsall, Malcolm R. Sim, David, M. Clarke, & Mark C. Creamer, (Journal of Traumatic Stress, 2013, Vol. 26, No. 3, pp, 310–318, doi: 10.1002/jts.21809), there were errors in Figures 2 and 3 on page 315 regarding the results from studies by Pietrzak et al. (2009, 2011). SEs were used instead of SDs when computing effect sizes for the relationship of PTSD with unit cohesion and social support for the Pietrzak et al. (2009) study. In addition, a minor typographical data error reported in Pietrzak et al., 2011 was also incorporated. Consequently, the effect sizes computed for the Pietrzak et al. 2009 study in our review were markedly large (ds < −7.00).
When we recomputed these effect sizes using SDs, the revised d values and 95% confidence intervals were d = −0.62 [0.91,−0.33] for unit cohesion in Pietrzak et al. (2009) and d = −1.04 [−1.34,−0.74] and d = −1.04 [−1.37,−0.70] for social support in Pietrzak et al. (2009, 2011, respectively).
As shown in the revised Figures 2 and 3, the standardized mean differences and 95% confidence intervals in the meta-analyses using these revised values were 0.55 [0.73,−0.37] for unit cohesion; and 0.97 [CI = −1.08,−0.87] for social support. These revised results are consistent with the findings from the sensitivity analyses reported in our paper, which omitted the Pietrzak et al. (2009) results. The overall conclusions of our systematic review are not changed by these revised meta-analyses—both higher unit cohesion and social support remained associated with lower severity of PTSD symptoms.