Olera CareNavigator - Increasing Affordability and Accessibility of Senior Care For Dementia Family Caregivers

Abstract Senior care is unaffordable for millions, yet a large percentage of elderly Americans miss out on $30 billion yearly in unclaimed financial aid because they are unaware that they qualify for assistance with food, housing, prescriptions, and healthcare or they do not know how to apply. This presents a high-impact opportunity to help millions with transformative innovation. Recent advances in artificial intelligence (AI) called Large Language Models (LLMs) could allow for mass eligibility-screening of elderly Americans and optimal resource-matching with underutilized local, state, or federal aid programs and other support programs. Novel applications of AI in the eldercare ecosystem have the potential to help caregivers especially for individuals with demanding conditions like Alzheimer's and related dementia (AD/ADRD) facing complex care demands and frequent access and affordability issues. In this proposed study, we aim to develop, evaluate, and deploy an AI-powered care planner technology to enable Americans to reclaim billions in underutilized aid and better access vital support resources. This collaborative project, involving industrial, academic, and community partners, is oriented around three specific aims: (Specific Aim 1) Development of novel specialized AI agents for elder care planning. Three categories of AI “agents'' will be engineered for specialized care planning functionalities including: needs assessment agents, resource & aid matching agents, and personalized care planning agents. (Specific Aim 2): Integration of multi-agent network & preliminary usability testing of a novel intelligent user interface (UI) among AD/ADRD caregivers. The agents will be integrated into a Multi-Agent Network and intelligent UI. The UI will be evaluated among a pilot group of AD/ADRD family caregivers (n=25) using the User Experience in Intelligent Environments (UXIE) framework to test for its acceptance and usability. Following a Build-Measure- Learn approach, we will iteratively refine the UI through survey data, platform usage data, and interview feedback. (Specific Aim 3): Evaluation of the user experience and caregiver perceptions of novel AI-powered care planning technology among AD/ADRD caregivers with diverse backgrounds. The network will be evaluated for usability, acceptance, and caregiving experience among AD/ADRD family caregivers (n=200) with diverse backgrounds regarding demographic characteristics and health needs factors. The modified Technology Acceptance Survey (TAS), the modified Mobile Application Rating Scale (MARS), Caregiver Self-Efficacy Scale (CSES-8), and the positive and negative appraisals of caregiving (PANAC) will be used as key quantitative survey tools to evaluate the technology acceptance and usability, and the caregiving experience in decision making before and after using the technology. At Phase IIB conclusion, the care-planner technology will be ready for service, commercially viable, and prepared for nationwide scaling.