Digital Technology and Clinical Evidence (DTCE) in Future Healthcare for Elderly
The care of elderly patients presenting with acute, often undifferentiated symptoms poses an increasing challenge for healthcare systems. In Emergency Departments (EDs), disposition decisions are typically diagnosis-driven and based on snapshot assessments of vital signs, which may inadequately reflect risk in older adults with multimorbidity and atypical presentations. Consequently, some patients are unnecessarily hospitalized—exposing them to risks such as delirium and functional decline—while others discharged may deteriorate and re-present. Existing Hospital-at-Home (H@H) and MIC@Home models remain constrained by restricted inclusion criteria, intermittent monitoring, and manpower-intensive workflows. This project introduces the Senior AI-guided Gateway for Emergency Care (SAGE) framework to address these gaps through two integrated aims: (1) developing and validating a symptom-based, data-driven AI triage algorithm to support safer ED and clinic disposition decisions, and (2) designing PRIME-NEWS, an AI-enabled remote monitoring system for continuous, irregular-time risk prediction with uncertainty estimation using home physiological data. Together, these innovations aim to reduce avoidable admissions, enable safe home-based care, improve early detection of deterioration, and alleviate healthcare system burden.
SAGE enables safer, data-driven triage and continuous home monitoring for elderly patients with acute symptoms. By reducing unnecessary admissions, detecting deterioration earlier, and supporting scalable Hospital-at-Home care, the project improves patient safety, experience, and outcomes while alleviating hospital bed pressures and manpower constraints across the healthcare system.

C-AIM Lead:
![]() | Joseph SUNG |
Recruitment contact: Asst. Prof Fan Xiuyi xyfan@ntu.edu.sg
Related Research Domain: Ageing/Frailty, Hospital @ Home
Keywords: Hospital at Home, Emergency Medicine Triage, Home Monitoring

.tmb-listing.jpg?Culture=en&sfvrsn=2a62f216_1)

