Internet-based interventions for posttraumatic stress: A meta-analysis of randomized controlled trials

• CBT-IBIs reduce PTSD symptoms significantly at post-test compared to passive controls.
• CBT-IBIs with different components are all more efficacious than passive controls.
• CBT- and EW-IBIs are not superior to active control conditions.
• Efficacy of CBT-IBIs is given at post-test, but does not carry through to follow-up.
• However, the number of studies was low for sufficiently powered testing.

Posttraumatic stress disorder (PTSD) is a prevalent and highly distressing affliction, but access to trauma-focused psychotherapy is limited. Internet-based interventions (IBIs) could improve the delivery of and access to specialized mental health care. Currently, no meta-analytical evidence is available on IBIs for PTSD. We conducted a meta-analysis of 20 randomized controlled studies, including 21 comparisons, in order to summarize the current state of efficacy for the treatment of PTSD and to identify moderator variables. Studies tested internet-based cognitive behavioral therapy (CBT) and expressive writing (EW) against active or passive comparison conditions, including subclinical and clinical samples. Results show that at post-assessment CBT-IBIs are significantly more efficacious than passive controls, resulting in medium to large effects on the PTSD sum and all sub-symptom scores (0.66 < g < 0.83), but both EW and CBT are not superior to active controls. EW differed from controls only at follow-up in reducing intrusions and hyperarousal, but based on merely two studies. Subgroup analyses reveal that for CBT none of the program components such as provision of therapeutic support, reminders, or number of sessions serves as a moderator. Overall, results for CBT-IBIs are promising, but the number of includable studies for subgroup analyses was low, limiting statistical power. Future research is necessary to systematically investigate the impact of treatment components and test against active controls with optimal power.