Using Bayesian methods to clarify the role of trauma-related cognitions on the course of posttraumatic stress in nonspecific trauma-focused therapy

Objective: There is strong empirical evidence that a reduction of trauma cognitions lessens PTSD symptoms, but there are discrepancies, including evidence that baseline negative posttrauma cognitions are associated with more, less, or are not associated with changes in PTSD symptoms. Discrepancies may be a function of power, sample size, analytic method, or measure. Methods: The rate of PTSD symptoms change across 16 trauma-focused treatment sessions in a community clinic (n = 56) was estimated using a Bayesian mixed-effects model with repeated measures nested within participants. Number of treatment sessions was the level-1 predictor variable with baseline levels of trauma-related cognitions (overaccommodation, assimilation, accommodation, and optimism) as time-invariant level-2 predictors. The relations between baseline trauma-related cognitions and PTSD symptoms change across sessions were assessed by cross-level interactions. Results: PTSD symptoms declined over treatment (b = −1.57, 95% CrI [−1.89, −1.25]). Higher levels of overaccommodation and assimilation were associated with attenuated (b = .38, 95% CrI [.03, .73]) and greater (b = −.36, 95% CrI [−.69, −.02]) rates of symptom reduction, respectively. The relations between PTSD symptom reduction and accommodation (b = −.12, 95% CrI [−.43, .20]) and optimism (b = −.13, 95% CrI [−.45, .20]) were uncertain. Conclusions: There may be a nuanced role of trauma-related cognitions on PTSD symptoms during treatment. More research is needed to examine theoretically grounded trauma-related cognitions that align with the different treatments for PTSD, particularly in reference to the current diagnostic criteria for PTSD. (PsycInfo Database Record (c) 2023 APA, all rights reserved)