JMIR Publications

ABSTRACT

Background:

People with a Mild to Borderline Intellectual Disability (MBID, IQ = 50-85) are particularly affected by Alcohol Use Disorders (AUDs). Given their susceptibility to peer pressure, patients with MBID and AUD should learn refusal skills in a tailored manner to avoid relapse. For this, Immersive Virtual Reality (IVR) appears promising to train managing peer pressure in a playful manner, given that conventional role-plays in the clinical setting lack realistic situations and standardizable actors. However, such complex social interactions have not yet been explored in IVR for MBID/AUD.

Objective:

We aimed to develop an IVR peer pressure simulation for AUD therapies in patients with MBID and AUD. For this, we involved experts from an addiction clinic for our initial Persuasive System Design (PSD) and exploration of usability, immersion procedures, and therapeutic goals for IVR therapy.

Methods:

We conducted a comprehensive co-creation method with five experts (i.e. psychologist, nurse specialist, psychomotor therapist, psychiatrist) from a Dutch addiction clinic for people with MBID and substance use disorder. Three focus groups were held to design the IVR environment, persuasive virtual agent, and dialogue. Afterwards, we developed and tested our IVR prototype with the same experts to improve the application/procedures and discuss promising IVR therapy approaches.

Results:

Our experts described visiting a friend at home with multiple friends to be the most relevant social situation for relapse. Together with the experts, we designed an IVR peer pressure simulation, during which patients can select coping responses of various riskiness levels to train proper refusal scripts. Our expert-based evaluation showed the need for natural speech with paralinguistic features and group dynamics (e.g. two against one) to improve the agent’s persuasive power. Further, facilitators (e.g. embodied interactions) and barriers (e.g. text-based procedures) for usability were reported. For clinical applications, IVE difficulty, content, and therapeutic goals should be tailorable to the patient’s needs. Lastly, experts preferred therapist-delivered interventions over stand-alone approaches to avoid a perilous trial and error.

Conclusions:

Our work establishes a first PSD for IVR peer pressure simulations in patients with MBID and AUD. With this, scholars can create comparable IVR simulations using an analogous co-creation approach, replicate findings, and identify active PSD elements. To boost the persuasive power of virtual humans, delivering subtle emotional information (e.g. paralinguistics) and group dynamics are promising. However, previous rapport building may be needed to ensure that agents are experienced as cognitively capable entities with certain (persuasive) power. Future works should validate our PSD with patients and explore therapeutic goals, including treatment protocols, using interdisciplinary teams.