Comparison of qSOFA Score and NEWS for Predicting Hospital Admission and Mortality in the Emergency Department: A Prospective Observational Study

Background/Objectives: This study compared the performance of the National Early Warning Score (NEWS) and the quick Sequential Organ Failure Assessment (qSOFA) score for predicting hospital admission and in-hospital mortality among adult patients triaged to the high-acuity area of the emergency department (ED). qSOFA was derived and validated in patients with suspected infection, and only 8.1% of this cohort carried a preliminary diagnosis of sepsis; its evaluation here is therefore exploratory. Methods: A total of 270 patients were included in this prospective observational study. Vital signs, Glasgow Coma Scale scores, and both scores were recorded at ED presentation. Discrimination was assessed by receiver operating characteristic (ROC) analysis, with the two scores compared using the DeLong test. Sensitivity and specificity were then examined separately at both conventional and data-derived thresholds. Bootstrap internal validation, calibration plots, Brier scores and decision-curve analysis were also performed, all computed using the development dataset. Results: Of 268 patients evaluable for the admission outcome (two ED deaths were excluded as a competing outcome), 35.8% (n = 96) were admitted, and 7.0% (n = 19 of 270) died in the hospital. No statistically significant difference was found between the AUCs of the two scores for admission (0.709 vs. 0.695; p = 0.627) or mortality (0.900 vs. 0.862; p = 0.150). At the conventional qSOFA score threshold of ≥2, the expected high-specificity, low-sensitivity profile was reproduced for mortality (96.8% and 52.6%), whereas the cohort-derived qSOFA score threshold of ≥1 favored sensitivity (84.2%). Bootstrap correction showed modest optimism. Calibration assessment was limited by the small number of events, and both score-specific models showed apparent net benefit over treat-all and treat-none strategies across most thresholds examined. Each score was independently associated with mortality in exploratory score-specific models. Conclusions: NEWS and qSOFA showed no statistically significant difference in discrimination in this high-acuity ED cohort. With only 19 deaths, the mortality analyses and the cohort-specific thresholds are exploratory and require confirmation in larger, multicenter cohorts before any clinical use.