Agentic AI for Clinical Urgency Mapping and Queue Optimization in High-Volume Outpatient Departments: A Simulation-Based Evaluation

A team of researchers has published a paper detailing a novel agentic AI framework designed to tackle severe overcrowding in hospital outpatient departments (OPDs). The system, developed by Ravish Gupta, Saket Kumar, and Maulik Dang, moves beyond the standard First-Come-First-Served model by integrating six components: voice-based multilingual symptom capture, an LLM for severity scoring, load-aware physician assignment, adaptive queue optimization, a multi-objective orchestrator, and a crucial 'Patient Memory System' that provides longitudinal, context-aware triage by remembering a patient's history.

Evaluated through a discrete-event simulation modeling a District Hospital in Jabalpur with 368 synthetic patients, the results are striking. The AI framework achieved a 94.2% rate of critical patients being seen within 10 minutes, a massive improvement over the 30.8% rate under traditional FCFS. It successfully detected approximately 236 simulated 'urgency drift' events—where a patient's condition deteriorates while waiting—and identified about 12 additional hidden-critical cases using its patient memory. The system dynamically recomposed the queue, significantly increasing the number of high-urgency patients seen first, all while maintaining a comparable throughput of roughly 40.4 patients per hour, proving efficiency doesn't have to come at the cost of fairness or clinical safety.