Know Me, Support Me: AI reasoning tool for identifying and managing health issues of people with intellectual disability

Background: People with intellectual disability often need caregiver support to access health care. However, signs of deteriorating physical or mental health can be missed due to diagnostic overshadowing, communication barriers, and inadequate caregiver health literacy. Symptoms may be misattributed to the disability itself or behaviour of concern, while comorbid conditions and caregiver underreporting further complicate care. Effective management requires regular screening for symptoms, improved communication, and tailored interventions to initiate and access healthcare through an interdisciplinary approach involving caregivers and health service providers.

 

Approach: We are developing and implementing an easily accessible mobile AI evidenced-based reasoning tool called EarlyCare AI to support people with intellectual disability, their caregivers, and health service providers. The AI tool is being designed to: (1) analyse subtle cues from people with intellectual disability to help caregivers detect health issues early to get appropriate treatment; (2) assist with communicating symptoms to health service providers to guide appropriate care. (3) generate personalised support plans, in consultation with the health service providers, outlining treatments and expected progress.

 

Development includes the following: (1) identifying evolving needs, understanding the capacity to interact with technology, identifying possible input (including wearables and health records), and monitoring data sources; (2) co-designing the AI tools and innovative training programs, including virtual reality, with stakeholders; (3) cycles of testing the tool with people with intellectual disability, caregivers, and health providers; and (4) conducting a pilot trial with 200 people with intellectual disability and their supporters comparing (a) AI-supported care with (b) usual support Trial arms will be compared using service provider data, consumer experiences, and costs through surveys, interviews, and service records.

 

Results: We are using an existing Generative AI-Based Agent initially designed to transform mental health management. We are currently in the first and second stages of the research identifying the needs and planning the co-design of the AI tool. We will present on these findings, demonstrate the prototype and provide information on planning for the cycles of testing and the pilot trial.

 

Implications: The National Roadmap for Improving the Health of People with Intellectual Disabilities (2021) highlights the need for models of care that include the development of digital solutions that support decision making and communication. Within the context of the healthcare system, the provision of healthcare through assistive technologies, particularly in rural areas, provides the opportunity for monitoring and assessment activities to be provided to individuals for proactive healthcare management.