... Help-seeking stigma 2 RCTs (Mulligan et al., 2012;); 4 cohort comparisons (Russell et al., 2014;Twardzicki and Jones, 2017;Watson and Andrews, 2018;Gould, Greenberg, and Hetherton, 2007) Not applicable Meta-regression was not possible Insufficient Return to duty 8 case studies with no comparator (Holsenbeck, 1992;Hassan et al., 2010;Hoyt et al., 2015;Hung, 2008 Biggs et al., 2016;Carlier, Voerman, and Gersons, 2000;Castro et al., 2012;Deahl et al., 1994;Foran et al., 2013;Fornette et al., 2012;Greenberg et al., 2010;Hourani et al., 2018;Ilnicki et al., 2012;Mulligan et al., 2012;Pinna et al., 2017;Roy, Highland, and Costanzo, 2015;Shipherd, Salters-Pedneault, and Fordiani, 2016;Sipos et al., 2012;Sipos et al., 2014;Stetz et al., 2009;Twardzicki and Jones, 2017;Young, 2012;Zimmermann et al., 2015); 18 case studies in 19 articles (Bryan and Morrow, 2011;Bush, Ouellette, and Kinn, 2014;Carr et al., 2013;Gahm et al., 2009;Garber and Zamorski, 2012;Griffith and West, 2013;Hourani et al., 2017;Jarrett, 2013;Jones et al., 2011;Burdett et al., 2011;Judkins and Bradley, 2017; (Wesemann et al., 2016;Tuckey and Scott, 2014;Wilson et al., 2001;Cacioppo et al., 2015); 4 cohort comparisons (Larsson, Michel, and Lundin, 2000;Deahl et al., 1994;Gould, Greenberg, and Hetherton, 2007) Indirect Biggs, 2016;Carlier, Voerman, and Gersons, 2000;Castro et al., 2012;Deahl et al., 1994;Foran et al., 2013;Fornette et al., 2012;Greenberg et al., 2010;Hourani et al., 2018;Ilnicki et al., 2012;Mulligan et al., 2012;Pinna et al., 2017;Roy, Highland, and Costanzo, 2015;Shipherd, Salters-Pedneault, and Fordiani, 2016;Sipos et al., 2012;Sipos et al., 2014;Stetz et al., 2009;Twardzicki and Jones, 2017;Young, 2012;Zimmermann et al., 2015); 18 case studies in 19 articles (Bryan and Morrow, 2011;Bush, Ouellette, and Kinn, 2014;Carr et al., 2013;Gahm et al., 2009;Garber and Zamorski, 2012;Griffith and West, 2013;Hourani et al., 2017;Jarrett, 2013;Jones et al., 2011;Burdett et al., 2011;Judkins and Bradley, 2017; Overall, meta-analyses found that COSC programs had no significant difference in effect on sleep, alcohol misuse, anxiety, depression, PTSD symptoms (usually measured by PCL score), or help-seeking stigma compared with active interventions, such as a single, standard stressmanagement class. Meta-analyses also found no significant effect on alcohol misuse, depression, PTSD symptoms, PTSD case rate, stress level, or help-seeking stigma compared with no intervention. ...