Mediators of change in a condensed online exposure-based intervention provided soon after trauma: insights from a randomised controlled trial

Abstract

Background: The active mechanisms of change are unclear in early-provided exposure-based interventions for psychological trauma. This study aimed to address this gap by analysing weekly data from a randomised trial involving a 3-week therapist-guided online intervention based on prolonged exposure compared to a waiting-list control group.Method: The objective was to investigate whether changes in each of the four subscales of the Posttraumatic Stress Disorder Checklist, fifth version (PCL-5; i.e. intrusions, avoidance behaviours, negative alternations in cognitions and hyperarousal) during the three-week intervention period mediated subsequent improvements in other post-traumatic stress symptoms at the controlled 1-month follow-up. We included baseline levels of both the mediator and the outcome as well as changes in the outcome from baseline to week 3 as covariates in a mediation model.Results: The results showed that reductions in avoidance during the intervention period mediated reduced symptom levels of intrusions, negative alternations in cognitions, and hyperarousal at week 7 (z-scores of indirect effect estimates = −0.12 to −0.07). No other PCL-5 subscales were found to be mediators of change.Conclusions: The results from this study highlight the importance of addressing avoidance behaviours in online exposure-based interventions provided in the early aftermath of trauma. Sensitivity analysis showed that the mediation effects were sensitive to assumptions related to mediator-outcome confounders, which could be considered a study limitation.Trial registration: ClinicalTrials.gov identifier: NCT03850639.

Keywords

• Post-traumatic stress
• trauma
• early intervention
• Internet
• prolonged exposure
• mediation