Abstract
BACKGROUND: The "Describe-Investigate-Create-Evaluate" (DICE) approach has been considered a guide for managing behavioral and psychological symptoms of dementia (BPSD). However, limited resources may limit the implementation of the DICE approach. With the development of AI technology, the effectiveness of the AI-aided DICE algorithm for BPSD management has yet to be determined. Therefore, this study aims to examine the effectiveness of the AI-aided DICE algorithm for managing BPSD in low-resource settings. METHODS: The cluster randomized controlled trial will be conducted in 12 medical facilities where geriatric psychiatrists are not fully installed. One hundred eighty-four persons with mild and moderate BPSD will be enrolled and randomized to the AI-aided DICE group (n=92) and usual care group (n=92). In the AI-aided DICE group, all participants will receive a comprehensive assessment on a digital triage platform to identify individualized needs and target symptoms, be prescribed a personalized management plan based on the AI-aided decision process, be monitor the implementation of the management plan, and receive follow-up assessment to evaluate the effectiveness. The neuropsychiatric inventory questionnaire and caregiver burden inventory will measure primary and secondary outcomes. The study duration for each participant will be 12 weeks. DISCUSSION: The study will examine the effectiveness of the AI-aided DICE algorithm for managing BPSD in low-resource settings. The findings will support the implementation of an AI-aided algorithm and leverage the practice of quality care for dementia.