Predictive value of adding end-tidal CO₂ and perfusion index to NEWS2 for 30-day in-hospital mortality: a prospective...

This study aimed to evaluate whether integrating end-tidal carbon dioxide (ETCO₂) and perfusion index (PI) into the National Early Warning Score 2 (NEWS2) enhances its ability to predict 30-day in-hospital mortality in patients visitted to the emergency department (ED).

This prospective, single-center cohort study was conducted between March 15 and May 15, 2025, in a tertiary ED. Adult patients requiring continuous monitoring were included. ETCO₂ and PI were measured upon admission and incorporated into the NEWS2 score using predefined thresholds. The primary outcome was 30-day all-cause in-hospital mortality. Model performance was assessed via ROC analysis, Brier score, Hosmer-Lemeshow test, decision curve analysis (DCA), and Net Reclassification Index (NRI).

A total of 375 patients were analyzed (mean age: 76.5 ± 7.6 years; 56.8% female), with a 30-day mortality rate of 5.9%. The base NEWS2 had an AUC of 0.804 (95%CI: 0.734-0.868), while the extended NEWS2 + ETCO₂ + PI model demonstrated a significantly higher AUC of 0.893 (95%CI: 0.824-0.944; p < 0.001). The optimal cut-off for the combined model was > 12, yielding 77.3% sensitivity and 90.9% specificity. It also showed superior calibration (Brier score: 0.039; HL p = 0.265) and the highest net clinical benefit between threshold probabilities of 0.22 and 0.45. NRI analysis showed a net reclassification improvement of 46.3%.

The integration of ETCO₂ and PI into the NEWS2 score significantly improves its predictive accuracy for short-term mortality. The combined model may contribute to earlier and more accurate identification of high-risk patients in emergency care settings.