TY - JOUR
T1 - Artificial Intelligence-Enabled Hospital Command Center for Predicting Patient Transfer Bottlenecks Using Admission Requests, Unit Occupancy, Bed Cleaning Duration, Isolation Requirements, and Staffing Constraints
AU - James Walker
AU - Olivia Harris
JF - Journal of Health Informatics and Digital Systems
JO - J. Health Inform. Digit. Syst.
SN - 3149-8973
Y1 - 2023
VL - 3
IS - 2
DO - 10.68159/h134661506
SP - 83
N2 - Patient transfer delays from the emergency department, post-anaesthesia care unit, and outside facilities are a major source of hospital congestion. These delays can convert local unit constraints into system-wide capacity failure when demand and transfer readiness are not coordinated in real time. Hospital command centres increasingly monitor bed status, queue length, and operational pressure, but many remain reactive. They often identify congestion only after transfer queues have formed, particularly when demand, cleaning delays, isolation needs, and staffing shortages converge. This article proposes an AI-enabled hospital command centre framework for predicting patient transfer bottlenecks. The system would fuse real-time admission requests, unit occupancy, bed cleaning duration, isolation requirements, and staffing constraints into a dynamic bottleneck risk score. The framework includes an admission request projection module, a unit occupancy forecasting engine, a bed-cleaning-time estimation model, an isolation-delay calculator, and a staffing-aware transfer-capacity reasoner. Together, these components would estimate whether each receiving unit can absorb expected transfer demand. The system would provide command centre staff with a rolling risk map of potential transfer bottlenecks up to four hours ahead. This would support earlier load-balancing, cleaning prioritisation, staffing escalation, and transfer-routing decisions. A predictive command centre could shift hospital flow management from reactive queue monitoring to proactive bottleneck prevention. Such a framework should be evaluated prospectively before operational deployment.
UR - https://cirpublications.com/h134661506
ER -