Can virtual reality exposure therapy gains be generalized to real-life? A meta-analysis of studies applying behavioral assessments

In virtual reality exposure therapy (VRET), patients are exposed to virtual environments that resemble feared real-life situations. The aim of the current study was to assess the extent to which VRET gains can be observed in real-life situations. We conducted a meta-analysis of clinical trials applying VRET to specific phobias and measuring treatment outcome by means of behavioral laboratory tests or recordings of behavioral activities in real-life. Data sources were searches of databases (Medline, PsycInfo, and Cochrane). We included in total 14 clinical trials on specific phobias. Results revealed that patients undergoing VRET did significantly better on behavioral assessments following treatment than before treatment, with an aggregated uncontrolled effect size of g = 1.23. Furthermore, patients undergoing VRET performed better on behavioral assessments at post-treatment than patients on wait-list (g = 1.41). Additionally, results of behavioral assessment at post-treatment and at follow-up revealed no significant differences between VRET and exposure in vivo (g = -0.09 and 0.53, respectively). Finally, behavioral measurement effect sizes were similar to those calculated from self-report measures. The findings demonstrate that VRET can produce significant behavior change in real-life situations and support its application in treating specific phobias.