Co-Design and Feasibility Testing of an AI-Based Virtual Reality Application to Prepare People With Intellectual Disability for Healthcare Visits.

All ten participants were recruited through a disability advocacy organisation and were experienced at giving feedback, so they are not typical of people with intellectual disability as a whole. Everyone in the study was already engaged with health services and had a regular doctor, so the paper says nothing about people who avoid healthcare altogether or who do not speak up for themselves. Participants had to speak English, tolerate a headset and handle a controller, and people with epilepsy were excluded, so the findings do not cover people who communicate in other ways or cannot use a head-mounted display. Each person used the system once only, in a university lab, so there is no evidence about whether anything is learned or remembered, or how it would work in a real clinic or at home. Only self-report was collected. There was no objective measure of learning, confidence or anxiety. The researcher held the controller and stayed within arm's reach, prompting people when the AI did not understand them, which the authors say may have hidden usability problems. The two researchers expected positive responses going in, and no member checking or independent verification of the coding was done.

Funded by the UNSW Disability Innovation Institute Research Seed Funding 2024 and the UNSW Health Systems Research Theme 2025 Collaborative Grant. Individual authors report support from a Dementia Australia Research Foundation post-doctoral fellowship and National Health and Medical Research Council grants, and the National Centre of Excellence in Intellectual Disability Health is funded by the Australian Government Department of Health, Disability and Aged Care. The authors declare no conflicts of interest, and no VR or AI company is named as a funder. The authors note their affiliation with the National Centre may have influenced what participants expected of the study.

Do not read the enthusiasm in these interviews as evidence that the programme makes healthcare visits easier. Ten people used it once, in a university lab, with a researcher holding the controller and prompting them, and there was no objective measure of learning or confidence. Nothing here shows the app reduces anxiety or improves care.