Journal Article

External validation of the International Early Warning Score in non-traumatic emergency department patients: a prospective cohort study

Bayram, Fatma, Deniz, Turgut, İlhan, Buğra, Kan, Zeynep, Eroğlu, Oğuz

2025

Overview

Background Emergency department (ED) overcrowding has become a global public health concern, underscoring the importance of rapid and reliable risk stratification tools. Early warning scores are widely used to identify patients at risk of deterioration and mortality. The recently developed International Early Warning Score (IEWS), which incorporates age and sex adjustments into the National Early Warning Score (NEWS) model, has shown promising results and has undergone initial external validation in a Danish cohort; however, no prospective external validation has yet been conducted, and broader international validation remains limited. This study aimed to evaluate the performance of IEWS compared with NEWS in predicting in-hospital mortality, 30-day mortality, and ICU admission among adult ED patients. Methods This prospective observational cohort study was conducted between July and August 2024 in a tertiary university hospital ED with an annual census of ~ 70,000 visits. Adult patients presenting to the ED were included, while trauma cases, patients without vital signs on arrival, interhospital transfers, and cases with incomplete data were excluded. IEWS and NEWS were calculated at presentation. The primary outcome was all-cause in-hospital mortality; secondary outcomes included 30-day mortality and ICU admission. Results A total of 8,666 patients were analyzed. The median age was 40 years (IQR: 26–58), and 51.5% were female. In-hospital mortality was 1.5% ( n  = 134), and 30-day mortality was 1.9% ( n  = 163). IEWS demonstrated excellent discriminative ability for in-hospital and 30-day mortality (AUC: 0.944 and 0.930, respectively), and good performance for ICU admission (AUC: 0.876). In contrast, NEWS showed good performance for in-hospital and 30-day mortality (AUC: 0.884 and 0.848, respectively) and moderate performance for ICU admission (AUC: 0.781). IEWS consistently outperformed NEWS across all outcomes ( p  < 0.05, DeLong’s test). Conclusion IEWS outperformed NEWS in predicting in-hospital mortality, 30-day mortality, and ICU admission among non-traumatic ED patients. Given its high sensitivity, specificity, and overall discriminative performance, IEWS may serve as a reliable bedside tool for patient risk stratification in the ED. Large-scale multicenter studies are needed to confirm its generalizability across diverse populations. Clinical trial number Not applicable.

• Publisher: BioMed Central,BioMed Central Ltd,Springer Nature B.V,BMC
• Subject: Adult / Calibration / Cardiac arrest / Clinical deterioration / Clinical outcomes / Cohort analysis / Comparative analysis / Consciousness / Consent / Decision making / Early Warning Score / Emergency department / Emergency medical care / Emergency Medicine / Emergency service / Emergency Service, Hospital - statistics & numerical data / Female / Health aspects / Hospital Mortality / Hospitals / Humans / Inclusion / Intensive care admission / Intensive Care Units - statistics & numerical data / International Early Warning Score / Male / Medical research / Medicine / Medicine & Public Health / Medicine, Experimental / Middle Aged / Mortality / Mortality prediction / National Early Warning Score / Patient satisfaction / Patients / Prospective Studies / Public health / Risk Assessment / Validation studies / Vital signs