Efficacy of the Virtual Reality Intervention VR FestLab on Alcohol Refusal Self-Efficacy: A Cluster-Randomized Controlled Trial

Abstract

It is currently unknown whether a virtual social environment can support young people in building their skills to overcome peer pressure when offered alcohol. This study evaluated the efficacy of the newly developed virtual reality simulation game VR FestLab on the refusal self-efficacy regarding social pressures to drink of Danish male and female students aged 15-18. VR FestLab features a party setting where adolescents can "steer" their own party experience. Eleven schools were included in a cluster-randomized controlled trial and allocated to either the intervention (n = 181) or the active control group (n = 191). Students in intervention schools played VR FestLab, while those in the control group played the VR game Oculus Quest-First Steps. The primary outcome measure was the social pressure subscale of the drinking refusal self-efficacy scale (DRSEQ-RA). The intervention effects were measured immediately after the intervention/control session (T1) and after a 6-week follow-up (T2). Data were examined using linear mixed regression models. Our study did not demonstrate a significant effect of drinking refusal self-efficacy at T1. For all secondary outcomes, we observed no substantial differences between the intervention and control groups. This study provides new insights into the feasibility and effectiveness of an innovative virtual reality alcohol prevention tool. VR FestLab can be an innovative and promising contribution to complement existing school-based alcohol prevention, but more research is needed to improve its effectiveness.

Keywords: adolescents; alcohol; alcohol prevention; cluster-RCT; drinking refusal self-efficacy; intervention; peer pressure; school-based prevention; virtual reality.

Figure A1

Theory of change of VR FestLab using the Behavior Change Wheel categorization based on Michi et al. [34].

Figure 1

Participant flow through the trial.

Figure 2

Intervention effects for drinking refusal self-efficacy (social pressure subscale of DRSEQ-RA) at T1 (n = 378), in total sample and by sex, age, family wealth, baseline value, lifetime binge drinking, and sensation seeking * based on linear mixed models ^#. * Variable names in graph: drinking refusal self-efficacy: “resist soc press drink”; baseline value of DRSEQ-RA less than or over the mean of 20: “baseline <20/20+”, lifetime binge drinking: “lifetime binge”; sensation seeking with values less than or over the mean of 27: “sens seek <27/27+”; family affluence: “affluence”. ^# The model for the first five estimates is adjusted for sex, age, and baseline value. Additional separate models were used for subgroup analyses of lifetime binge drinking, sensation seeking, and family affluence. All these models were adjusted for age, sex, and baseline value.

Figure 3

Intervention effects for secondary outcomes (susceptibility to social pressure, drug refusal skills, outcome expectation, knowledge on blood alcohol concentration, communication skills, and social support) * at T1 (n = 378) based on linear mixed models ^#. * Labels used in graph: social pressure: “sp”; blood alcohol concentration: “bac”. ^# Separate regression models were used for each outcome. The models are adjusted for sex, age, and the particular baseline value.

Figure 4

Intervention effects for secondary outcomes (drinking refusal self-efficacy, susceptibility to social pressure, drug refusal skills, outcome expectation, knowledge on blood alcohol concentration, communication skills, and social support) * at T2 (n = 378), based on linear mixed models ^#. * Labels used in graph: social pressure: “sp”; blood alcohol concentration: “bac”; drinking refusal self-efficacy: “resist soc press drink”. ^# Separate regression models were used for each outcome. The models are adjusted for sex, age, and the particular baseline value.