Implementation and Evaluation of an AI-Assisted Telerehabilitation System for Postdischarge Continuation of Rehabilitative Care: Protocol for a Randomized Controlled Trial

作者：Tan CYM, Seah XY, Seah SSY, Loh XH, Chua WHP, Xia OJ, D/O Chandran C, Hosain H, Qiu W, Chong K, Tai BC, Low LL · 发表于：JMIR research protocols · 年份：2026 · DOI：10.2196/78400 · 研究领域：Artificial Intelligence、Humans、Telerehabilitation、Pragmatic Clinical Trials as Topic、Singapore、Patient Discharge、Cost-Benefit Analysis、Female、Male、Hip Fractures、Randomized Controlled Trials as Topic

BACKGROUND: The World Health Organization (WHO) launched a Rehabilitation 2030 initiative to call for global action to scale up rehabilitation efforts. Rehabilitation needs are growing, and efforts should be made to strengthen and integrate rehabilitation into all levels of health care, including building research capacity and expanding evidence for rehabilitation. Postdischarge rehabilitation is essential for reducing hospital readmissions and maintaining patients' health within the community. However, locally, the shift toward community-based rehabilitation is often hampered by long waiting times for admission to day rehabilitation centers (DRCs), cost and logistical barriers, and the lack of a structured program to onboard patients and caregivers to the digital solutions required for rehabilitation in Singapore. Telerehabilitation systems that incorporate wearables within a structured rehabilitation program support inpatient rehabilitation and enable patients to continue with physical rehabilitation after discharge while awaiting admission to a DRC. OBJECTIVE: The aim of this randomized controlled trial (RCT) is to investigate the clinical and cost-effectiveness of ATLAS, an AI-assisted telerehabilitation system, among patients admitted to community hospitals (CHs) for rehabilitation. METHODS: This is a 2-arm pragmatic RCT. Participants admitted to CHs for rehabilitation for hip fracture, musculoskeletal conditions, or deconditioning will be enrolled and randomized to ei...