20 The Use of Delta-Modified Early Warning Scores Predicts Additional Mortality Risk Within Quick Sequential Organ Failure Assessment-Negative Emergency Department Patients

In the emergency department (ED), Quick Sequential Organ Failure Assessment (qSOFA) and Modified Early Warning Scores (MEWS) offer quick references that have been validated as predictive for in-hospital morbidity and mortality. When considering the importance of changes of these scores within the emergency department, there remains a paucity of data. We examined the correlations of the calculated differences in serial MEWS scores (deltaMEWS or dMEWS) and patient outcomes, particularly among patients who did not present with abnormal vital signs. We conducted a retrospective analysis on ED patients with MEWS that were admitted to a single, academic Level 1 Trauma Center over a 5-year period (December 2013-2018) We calculated a Triage-to-Last deltaMEWS by subtracting the patient’s MEWS at triage by their last MEWS before leaving the ED. Each dMEWS was then correlated to in-hospital mortality, ICU admission, length of stay (LOS), diagnosis of sepsis, 48-hour escalation of care from the floor to ICU. We analyzed the relationships using regression models incorporating the covariates: sex, age, Charlson Comorbidity Index, home beta-blocker use, hemoglobin level, time to antibiotics, confirmed influenza, and intravenous fluids, vasopressors, blood transfusions received in the ED. Sub-group analysis was then conducted among patients who had a qSOFA score at triage of 0. Our analysis included 33,339 ED patients with an ICU admission rate of 22.4% and a mortality rate of 2.4%, 1.1% among patients with a negative triage qSOFA (score = 0) and 5.6% among patients with a positive triage qSOFA (score ≥1). In our sample, for every point increase from Triage-to-Last deltaMEWS was predictive of mortality (OR 1.42, 95% CI 1.35-1.50), ICU admission (OR 1.23, 95% CI 1.32-1.38), developing Sepsis (OR 1.24, 95% CI 1.20-1.29), LOS (B 0.46, 95% CI 0.41-0.51), escalation from floor to ICU (OR 1.21, 95% CI 1.15-1.29). Among patients who had a positive or negative triage qSOFA, every point increase in their dMEWS was associated with increased mortality (OR 1.37, 95% CI 1.29-1.46 and OR 1.53, 95% CI 1.40-1.70, respectively), see Figure 1. Moreover, increasing dMEWS while in the emergency department was as predictive of mortality as qSOFA scores at triage (AUROC 0.8299 vs. 0.8242, p=0.333). Failure to normalize deltaMEWS in the emergency department, and every point of worsening deltaMEWS after triage, are both strongly associated with mortality, ICU admission, LOS, development of sepsis, and unplanned escalation of care. Further, among patients considered lower-risk by triage qSOFA, subsequent hemodynamic decompensation in the emergency department, as modeled by deltaMEWS, was strongly associated with outcomes, and was as predictive as initial qSOFA for all outcomes as rising qSOFA scores at triage.