Combining a Modified Emergency Severity Index Score with Chief Complaints - Validation of a New Allocation Algorithm for Integrated Emergency Centers

Objectives Current health policy reforms in Europe aim to synchronize emergency departments and acute outpatient clinics (AOC)/primary care services through the establishment of Integrated Emergency Centers (IECs). A central challenge of this model is the upstream triage at a central point of assessment (CPA). To date, there is a lack of standardized concepts for a single-stage assessment conducted by emergency-qualified staff that also facilitates cross-sectoral allocation. To address this gap, this study introduces the optimized leading-symptom-weighted modified Emergency Severity Index (L-mESI opt.) triage instrument. This tool aims to safely distribute the clinical workload between emergency departments (EDs) and AOCs by linking a modified emergency severity index (mESI)—omitting vital signs—with the patient's chief complaint (CEDIS). Methods A retrospective analysis of walk-in emergency patients was conducted at two university hospitals in 2024 (training dataset: n=26,845). Outpatient treatment rates were determined for combinations of chief complaints and mESI categories. These probabilities informed an optimized allocation algorithm (L-mESI opt.) via receiver operating characteristic (ROC) analysis and the maximum Youden Index. The model was verified using a separate validation dataset (n = 11,512). Results The pretest probability for AOC allocation was identified at 74%. The optimized L-mESI-opt. algorithm correctly allocated 54.2% of patients to the AOC, the ED workload was reduced by 46.9%. Compared with standard ESI-based steering, the L-mESI-opt. significantly improved sensitivity for inpatient care from 54.5% to 69.5% and achieved a negative predictive value of 85.4%. Although 14.6% of AOC-allocated patients required subsequent ED admission ("Upstream") in the validation set, critical outcomes remained minimal (observation unit transfer: 0.3%; mortality: 0.1%). Validation confirmed consistent quality and safety. Conclusion L-mESI opt. model demonstrates the potential to effectively reduce ED workload while maintaining high patient safety in a simulated environment. These results show that its adaptability could allow for broad application in IECs. However, the successful implementation of such an algorithm will likely depend on close spatial and organizational links between EDs and AOCs to manage bilateral patient flow efficiently. Prospective clinical trials are warranted to confirm these findings in real-world settings.