Chang Ying

Zibo Polytechnic University

Abstract：

The overcrowding of Emergency Departments (ED) and the inherent subjectivity of traditional triage protocols frequently lead to resource misallocation and compromised patient prognosis. This study investigates the quantitative relationship between the implementation of a data-driven optimized emergency triage system and the subsequent clinical outcomes of patients. A retrospective cohort study was conducted utilizing electronic health records from a Grade-A Tertiary Hospital, encompassing 2,450 emergency admissions between January 2023 and December 2023. The cohort was stratified into a control group (n=1,200) managed via the traditional Emergency Severity Index (ESI) and an observation group (n=1,250) managed via an optimized, algorithmically assisted electronic triage system. The optimized architecture integrates dynamic vital sign weighting and automated risk stratification. Statistical evaluation revealed that the optimized system significantly reduced triage wait times from an average of 14.5 ± 4.2 minutes to 5.2 ± 1.8 minutes (p<0.001). Furthermore, prognostic indicators demonstrated substantial improvement: the mistriage rate decreased from 9.8% to 2.4%, while the unexpected 24-hour Intensive Care Unit (ICU) transfer rate dropped from 7.5% to 3.8% (p=0.012). The findings substantiate that optimizing the triage system through data-driven decision support not only enhances operational throughput but also significantly correlates with improved short-term patient survival and resource utilization.

Key Words：

emergency triage; patient prognosis; system optimization; data-driven modeling; emergency severity index; decision support systems