The Role of REMS, MEWS, and NEWS scoring systems in predicting in-hospital mortality in non-trauma emergency cases requiring hospitalization

Abstract

This study aimed to evaluate the prognostic performance of the Rapid Emergency Medicine Score (REMS), Modified Early Warning Score (MEWS), and National Early Warning Score (NEWS) for predicting in-hospital mortality and subsequent intensive care unit (ICU) transfer in non-trauma adult emergency department (ED) patients requiring hospitalization. This prospective, single-center observational study included 311 adult non-trauma patients requiring hospitalization from the ED. REMS, MEWS, and NEWS were calculated using the first physiological measurements recorded during the initial ED assessment and were not used for clinical decision-making. The primary outcome was in-hospital mortality. The secondary outcome was subsequent ICU transfer among patients initially admitted to a general ward. Score distributions were compared between outcome groups, and discriminatory performance was evaluated using receiver operating characteristic analysis with area under the curve (AUC) estimates and 95% confidence intervals (CIs). Among 311 patients, 11 (3.5%) died during hospitalization. MEWS was significantly higher among patients who died than among survivors [2 (1, 2, 3–4) vs. 1 (0–2), p = 0.035], whereas REMS (p = 0.212) and NEWS (p = 0.211) did not differ significantly. MEWS demonstrated limited discrimination for in-hospital mortality, with an AUC of 0.681 (95% CI, 0.533–0.829). At a cutoff of ≥ 4, sensitivity was 36.4% (95% CI, 15.2–64.6) and specificity was 92.3% (95% CI, 88.8–94.8). Among 292 patients initially admitted to a general ward, 47 (16.1%) subsequently required ICU transfer. REMS, NEWS, and MEWS were significantly higher among patients requiring subsequent ICU transfer (p = 0.009, p = 0.002, and p = 0.005, respectively), but discrimination was modest, with AUCs of 0.619 (95% CI, 0.528–0.709), 0.639 (95% CI, 0.554–0.724), and 0.624 (95% CI, 0.537–0.711), respectively. MEWS was associated with in-hospital mortality in this cohort; however, its discriminatory performance was limited. REMS, NEWS, and MEWS demonstrated only modest discrimination for subsequent ICU transfer. None of the three scores demonstrated sufficiently strong performance to support their use as standalone prognostic tools in this population. Given the small number of mortality events, the mortality findings should be interpreted cautiously and require confirmation in larger cohorts.