Virtual Reality Exposure for Socially Anxious Adolescents

This study is going to be a single-blinded, randomized controlled trial (RCT) with three arms: VRE, IVE, and a WL condition. It will follow a mixed-subjects design with condition as a between factor and measurement time-points as a within factor. Due to the study's goal to unravel the effects of VRE, the focus of the intervention will be on exposure in both active conditions (without explicitly including other treatment components like cognitive restructuring). The intervention will take place either at participants' schools or at KU Leuven (i.e., PraxisP - the outpatient treatment facility of the Faculty of Psychology and Educational Sciences, PSI - Psychological Institute) depending on the participants' preference and availability. In case this is not possible, neighboring healthcare practices will be explored as an alternative option. In any case, the training will be held in a separate, private room where the privacy of the participant can be guaranteed. Each active condition will comprise of 7 training sessions in total approximately 90 minutes each.

At pre-assessment, all primary and secondary outcome measures and most of the predictors will be assessed approximately one week prior to the first session. The outcome measures will be re-assessed during the post-assessment (one week after the last session) and the follow up assessments (3-months \& 6-months following the post assessment). During the training sessions, measurements of the mechanisms, remaining predictors, and participant feedback will be acquired and at mid-assessment an additional measurement of the primary outcome measures (e.g., social anxiety measures) will be obtained. Moreover, a brief social anxiety measure will also be conducted on a weekly basis to record symptom-related changes from session to session. Following the training, a qualitative interview will be conducted. This semi-structured interview will be conducted in a convenience sample of participants from both active conditions (e.g., 10-16 participants per condition due to pragmatic considerations and based on general recommendations as to when theoretical data saturation can be reached) to investigate VRE's acceptability and to compare it to IVE. To decrease demand effects, the (online) interview will be conducted by an independent interviewer (e.g., not the therapist) and will be audio-taped so that the interviewer can focus solely on the interaction.

Moreover, in line with good clinical practice and individualized treatment, the specific nature and content of the exposure exercises will be determined for each participant individually by presenting them with an extensive list of the possible exercises (e.g., ordering food in a restaurant, asking a question in class, engaging in a group conversation) in combination with a functional analysis. To ensure that the active conditions are comparable, the IVE condition will be matched to the VRE condition in terms of exposure duration and homework assignments will not be actively encouraged in both conditions. The crucial difference between the two active conditions is that in the IVE condition the exercises take place in real life, whereas in VRE the exercises take place in virtual reality. For the WL condition, the (primary and secondary) outcome measures will be administered before, during, and after the given waiting period (i.e., duration of training sessions) at the same instances as with the active conditions. After that time, participants in the WL condition will have the opportunity to be randomized into IVE and VRE.