Abstract
OBJECTIVE: The surge in the number of emergency department (ED) visits due to a growing population, aging society, and easier access to healthcare highlights the need for an effective triage process. Our goal in this study was to compare the clinical and operational performance of a triage system supported by artificial intelligence (AI) with two traditional methods-the Emergency Severity Index and the Manchester Triage System-in a high-volume ED. METHODS: In this prospective study, 18,000 adult patients were randomized equally to one of the three triage systems. Primary and secondary outcomes included patient wait time, complication and mortality rates, resource utilization, medical errors, legal issues, and patient satisfaction. RESULTS: Compared with the Manchester Triage System, the AI-supported system was associated with significantly lower in-ED mortality (OR 0.39, 95% CI, 0.32-0.47; P < .001) and lower complication rates (4.42% vs 10.25%), as well as higher patient satisfaction scores (9.0 vs 7.0; P < .001). Resource utilization was also more balanced in the AI-supported triage cohort. CONCLUSION: The AI-assisted triage system showed favorable clinical and operational patterns relative to traditional methods. However, the single-center design and short observation period limit generalizability, and causal inferences could not be firmly established.