Agreement and classification performance of ChatGPT-5-generated and nurse-led emergency department triage using an expert reference decision: A retrospective study
International Emergency Nursing, cilt.89, 2026 (SCI-Expanded, SSCI, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 89
- Basım Tarihi: 2026
- Doi Numarası: 10.1016/j.ienj.2026.101918
- Dergi Adı: International Emergency Nursing
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Social Sciences Citation Index (SSCI), Scopus, CINAHL, EMBASE, MEDLINE, Health Research Premium Collection (ProQuest)
- Anahtar Kelimeler: ChatGPT-5, Decision support system, Emergency department, Nurse, Triage, Under-triage
- Atatürk Üniversitesi Adresli: Evet
Özet
This study evaluated the agreement and classification performance of ChatGPT-5-generated triage compared with nurse triage using an expert reference decision in adult emergency department patients. In this retrospective cross-sectional study, 1564 adult patients with documented yellow or green triage classifications were included. Standardized triage-level patient information was entered into ChatGPT-5 using an identical prompt for all patients. Agreement was assessed using Cohen's κ, and classification performance was evaluated using sensitivity, specificity, and overall accuracy. The mean age was 39.8 ± 18.5 years, and 55.9% of patients were female. The expert reference decision classified 50.0% of patients as yellow, compared with 48.5% by ChatGPT-5-generated triage and 25.6% by nurse triage. ChatGPT-5-generated triage showed higher agreement with the expert reference decision than nurse triage (κ = 0.670 vs. 0.467), higher overall accuracy (83.5% vs. 73.3%), and higher sensitivity (82.0% vs. 49.0%). Compared with nurse triage, ChatGPT-5-generated triage produced fewer green classifications among expert-reference yellow patients (18.0% vs. 51.0%), whereas nurse triage produced fewer yellow classifications among expert-reference green patients (2.3% vs. 15.0%). These findings indicate that ChatGPT-5-generated triage more closely approximated the expert reference decision than nurse triage under the conditions evaluated in this study.