Monocyte Distribution Width, Neutrophil-to-Lymphocyte Ratio, and Platelet-to-Lymphocyte Ratio Improves Early Prediction for Sepsis at the Emergency

(1) Background: Sepsis is a life-threatening condition, and most patients with sepsis first present to the emergency department (ED) where early identification of sepsis is challenging due to the unavailability of an effective diagnostic model. (2) Methods: In this retrospective study, patients aged ≥20 years who presented to the ED of an academic hospital with systemic inflammatory response syndrome (SIRS) were included. The SIRS, sequential organ failure assessment (SOFA), and quick SOFA (qSOFA) scores were obtained for all patients. Routine complete blood cell testing in conjugation with the examination of new inflammatory biomarkers, namely monocyte distribution width (MDW), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR), was performed at the ED. Propensity score matching was performed between patients with and without sepsis. Logistic regression was used for constructing models for early sepsis prediction. (3) Results: We included 296 patients with sepsis and 1184 without sepsis. A SIRS score of >2, a SOFA score of >2, and a qSOFA score of >1 showed low sensitivity, moderate specificity, and limited diagnostic accuracy for predicting early sepsis infection (c-statistics of 0.660, 0.576, and 0.536, respectively). MDW > 20, PLR > 9, and PLR > 210 showed higher sensitivity and moderate specificity. When we combined these biomarkers and scoring systems, we observed a significant improvement in diagnostic performance (c-statistics of 0.796 for a SIRS score of >2, 0.761 for a SOFA score of >2, and 0.757 for a qSOFA score of >1); (4) Conclusions: The new biomarkers MDW, NLR, and PLR can be used for the early detection of sepsis in the current sepsis scoring systems.