Heterogeneity in Treatment Effect in Posttraumatic Stress Syndrome Trials: A Meta-Regression Analysis

Abstract

Objective

To evaluate the heterogeneity in treatment effect in posttraumatic stress disorder (PTSD) trials.

Patients and Methods

We downloaded data from a publicly available repository that captured PTSD trials published from January 1988 through February 2023. We applied restricted maximum-likelihood random-effect meta-analyses and meta-regression to explore potential moderators of treatment effect including methodologic study features (risk of bias domains and control group response rate), characteristics of the population, and intervention features following the theme, intensity, and platform framework.

Results

We included 199 PTSD trials that reported the outcomes of diagnosis resolution (122 trials, 8437 patients) and clinically meaningful improvement (133 trials, 9895 patients). Multiple treatments demonstrated effectiveness but with significant heterogeneity. Statistically significant moderators included risk of bias domains of randomization sequence and outcome measurement, control group response rate reflecting severity of PTSD in the enrolled population, and whether the psychotherapeutic approach was trauma focused (P values <0.05). There was no statistically significant effect for the frequency of treatments per week, format of the intervention (eg, individual vs group), duration of the intervention, or delivery method (in person vs not), (P values <0.05). Characteristics of the population such as sex, age, and military status did not appear to significantly affect the treatment effect (P values <0.05).

Conclusion

Trauma focused psychotherapies should be considered the first-line intervention to induce remission. Several patient characteristics or treatment context did not modify the treatment effect, which allows tailoring care based on patient values, preferences and logistics.