Predictors of treatment completion in group psychotherapy for male veterans with posttraumatic stress disorder

Group therapy is a frequently used therapy format for posttraumatic stress disorder (PTSD). However, factors contributing to treatment completion remain understudied. The current study examined predictors of treatment completion, defined as having completed 10 out of 14 sessions within 16 weeks, in veterans with PTSD who engaged in a hybrid efficacy-effectiveness randomized controlled trial of group psychotherapy for PTSD. Veterans (N = 198) were randomly assigned to 14 sessions of either group cognitive behavioral treatment (GBCT; n = 98) or group present-centered treatment (GPCT; n = 100). Four primary domains of predictors were examined, encompassing sociodemographic factors, the severity of PTSD and comorbid conditions, modifiable predictors, and treatment condition. Multilevel binomial logistic regression models following the Fournier analysis approach were used to examine significant predictors within domains, which were then included in a final model. Overall, 70.7% of participants completed treatment (GCBT = 61.6%, GPCT = 79.8%). Participants in the GPCT condition were 2.389 times, 95% CI [1.394, 4.092], more likely to complete treatment than those in the GCBT condition. Older age, a higher income and level of educational attainment, more lifetime and current mental health diagnoses, and higher use of positive reappraisal ER skills predicted treatment completion. Higher levels of depressive symptoms, cumulative trauma burden, and use of positive refocusing ER skills predicted treatment noncompletion. These findings are discussed in the context of current clinical and research practices for examining treatment noncompletion, with attention to the inclusion of translational predictors.