JMIR Publications

ABSTRACT

Background:

Alcohol use disorder (AUD) is associated with substantial individual and societal burden. Cue exposure approaches can support treatment by targeting conditioned cue reactivity and high-risk contextual associations. However, routine implementation is limited by therapist time, scalability constraints, and the difficulty of reproducing ecologically valid drinking situations in standard care settings. Virtual reality (VR) and artificial intelligence (AI) have been proposed as promising tools to support scalable and contextually rich interventions in mental health care.

Objective:

This study aims to develop and technically evaluate a prototype virtual reality therapy assistant (VRTA) for narratively guided alcohol cue exposure in immersive VR.

Methods:

This is a single-site technical feasibility study conducted at the Central Institute of Mental Health Mannheim, Germany. The study started in January 2026. It builds on a participatory development process including multi-stakeholder workshops with developers, researchers, clinicians, and patient and public involvement (PPI) contributors. The study began with workshops to refine implementation priorities. A VRTA prototype will be developed to guide alcohol cue exposure scenarios through speech-based interaction with a computer-generated character in immersive VR, informed by established exposure manuals. An iterative qualitative expert appraisal will be conducted during development. In addition, 20 adult patients with AUD will participate in an observer-based evaluation in which therapeutic staff demonstrate the prototype and patients provide qualitative feedback.

Results:

Results are expected on the technical feasibility of the VRTA prototype, including interaction quality and integration into therapeutic workflows. Qualitative findings from expert appraisal and patient feedback (observed from a third-person perspective) will provide insights into usability, manual conformity, acceptability, and perceived therapeutic relevance.

Conclusions:

This study is intended to generate design specifications and feasibility data to inform subsequent prototype refinement and future clinical evaluation studies.