Mortality prediction by quick Sequential Organ Failure Assessment, National Early Warning Score 2, Shock Index, and Modified Shock Index in suspected sepsis: A retrospective study

Objective To compare the predictive performance of quick Sequential Organ Failure Assessment, National Early Warning Score 2, Shock Index, and Modified Shock Index for in-hospital mortality in adult emergency department patients with sepsis. Methods This retrospective observational study included 187 adult patients who presented to the emergency department and met the Sepsis-3 definition of sepsis between January 2022 and December 2023. The predictive performance of quick Sequential Organ Failure Assessment, National Early Warning Score 2, Shock Index, and Modified Shock Index was evaluated using receiver operating characteristic curve analysis, DeLong’s test, and diagnostic performance metrics, including sensitivity, specificity, positive predictive value, and negative predictive value. Multivariable logistic regression was performed as an exploratory secondary analysis. Results Among 187 patients, 37 (19.8%) died during hospitalization. Non-survivors showed greater physiological derangement at emergency department presentation, including higher respiratory rate, heart rate, and lactate levels and lower oxygen saturation. National Early Warning Score 2 demonstrated the highest discriminatory performance for in-hospital mortality (area under the curve, 0.94), followed by quick Sequential Organ Failure Assessment (area under the curve, 0.82), Modified Shock Index (area under the curve, 0.79), and Shock Index (area under the curve, 0.77). Pairwise comparisons using DeLong’s test showed that National Early Warning Score 2 performed significantly better than the other evaluated tools. In exploratory multivariable analysis, quick Sequential Organ Failure Assessment ≥2, Modified Shock Index ≥1.3, lactate >2.0 mmol/L, and chronic kidney disease were independently associated with in-hospital mortality. Conclusions National Early Warning Score 2 showed the best predictive performance for in-hospital mortality in adult emergency department patients with sepsis and may be a useful tool for early risk stratification in this setting.