key patterns and predictors of response to treatment for military veterans with post-traumatic stress disorder: a growth mixture modelling approach

Background to determine the patterns and predictors of treatment response trajectories for veterans with post-traumatic stress disorder (ptsd). methods conditional latent growth mixture modelling was used to identify classes and predictors of class membership. in total, 2686 veterans treated for ptsd between 2002 and 2015 across 14 hospitals in australia completed the ptsd checklist at intake, discharge, and 3 and 9 months follow-up. predictor variables included co-morbid mental health problems, relationship functioning, employment and compensation status. results five distinct classes were found: those with the most severe ptsd at intake separated into a relatively large class (32.5%) with small change, and a small class (3%) with a large change. those with slightly less severe ptsd separated into one class comprising 49.9% of the total sample with large change effects, and a second class comprising 7.9% with extremely large treatment effects. the final class (6.7%) with least severe ptsd at intake also showed a large treatment effect. of the multiple predictor variables, depression and guilt were the only two found to predict differences in response trajectories. conclusions these findings highlight the importance of assessing guilt and depression prior to treatment for ptsd, and for severe cases with co-morbid guilt and depression, considering an approach to trauma-focused therapy that specifically targets guilt and depression-related cognitions.