ROX index and NEWS2 for 28-day mortality prediction in emergency department patients with acute dyspnea: a prospective observational study

Abstract

Acute dyspnea is a frequent and high-risk presentation in emergency departments (ED), requiring rapid identification of patients at increased risk of deterioration and mortality. This study aimed to compare the accuracy of the Respiratory Rate–Oxygenation (ROX) index and National Early Warning Score 2 (NEWS2) score for predicting 28-day mortality in ED patients presenting with acute dyspnea. This prospective, single-center observational study was conducted in the ED of Ankara Bilkent City Hospital. Adult patients presenting with acute dyspnea were consecutively enrolled. We recorded demographic characteristics, comorbidities, physiological parameters, NEWS2 scores, and ROX index values at ED admission. The primary outcome was 28-day all-cause mortality. We performed receiver operating characteristic (ROC) analysis and multivariable logistic regression to evaluate prognostic performance. A total of 521 patients presenting to the ED with acute dyspnea were included, of whom 78 (15.0%) died within 28 days. Non-survivors were older and had higher NEWS2 scores and lower ROX index values than survivors. NEWS2 demonstrated modest discrimination for 28-day mortality (AUC 0.639, 95% CI 0.596–0.680), as did the ROX index (AUC 0.616, 95% CI 0.572–0.658). Pairwise comparison using DeLong’s test showed no statistically significant difference between the two AUCs (ΔAUC 0.023, 95% CI − 0.041 to 0.087; p = 0.483). At a threshold of NEWS2 ≥ 4, sensitivity was 73.0%, and specificity was 41.0%, while a ROX threshold of ≤ 22.2 yielded a sensitivity of 66.7% and specificity of 50.3%. In multivariable logistic regression, age (OR 1.03, 95% CI 1.01–1.05), Pre-existing lung cancer (OR 2.11, 95% CI 1.02–4.38), COPD (OR 0.33, 95% CI 0.17–0.65), and NEWS2 score (OR 1.25, 95% CI 1.13–1.39) were independently associated with 28-day mortality. Both NEWS2 and the ROX index demonstrated only modest discrimination for predicting 28-day mortality in ED patients with acute dyspnea. Although combining NEWS2 with selected clinical variables modestly improved discrimination, the observed model miscalibration and lack of external validation indicate that these findings should be interpreted cautiously and require further validation before clinical implementation.