Ample evidence supports the use of Virtual Reality (VR) for anxiety disorders. Nonetheless, currently there is no evidence about moderators or potential negative effects of VR treatment strategies. An Individual Patient Data (IPD) approach was employed with 15 retrieved datasets. The current study sample was composed of 810 patients. Randomized control trials (RCTs) for each primary outcome measure were performed, in addition to moderator analyses of the socio-demographic variables. Deterioration rates were 14 patients (4.0%) in VR, 8 (2.8%) (...)

Ample evidence supports the use of Virtual Reality (VR) for anxiety disorders. Nonetheless, currently there is no evidence about moderators or potential negative effects of VR treatment strategies. An Individual Patient Data (IPD) approach was employed with 15 retrieved datasets. The current study sample was composed of 810 patients. Randomized control trials (RCTs) for each primary outcome measure were performed, in addition to moderator analyses of the socio-demographic variables. Deterioration rates were 14 patients (4.0%) in VR, 8 (2.8%) in active control conditions, and 27 (15%) in the WL condition. With regard to receiving treatment, patients in a waiting list control condition had greater odds of deteriorating than in the two active conditions, odds ratios (ORs) 4.87, 95% confidence interval (CI) [0.05, 0.67]. In the case of the socio-demographic variables, none of them were associated with higher or lower odds of deterioration, with the exception of marital status in the WL condition; married people presented a significantly lower probability of deterioration, OR 0.19, 95% CI [0.05, 0.67]. Finally, when comparing pooled effects of VR versus all control conditions, the OR was 0.61 (95% CI 0.31–1.23) in favor of VR, although this result was not statistically significant. This study provides evidence about the deterioration rates of a therapeutic VR approach, showing that the number of deteriorated patients coincides with other therapeutic approaches, and that deterioration is less likely to occur, compared to patients in WL control groups.