This study aimed to re-evaluate whether the scoring systems, neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) were effective in predicting prognosis and severity of COVID-19 patients in the emergency department (ED). COVID-19 patients enrolled in this retrospective study divided into the death (DEA) and survival (SUR) groups, the severe/critical (SC) and non-severe/critical (non-SC) groups. The Acute Physiology and Chronic Health Evaluation (APACHE) II, Sequential Organ Failure Assessment (SOFA), National Early Warning Score (NEWS) and CCEDRRN COVID-19 Mortality Score were calculated. The neutrophil, lymphocyte and platelet counts were extracted from the first routine blood examination, and NLR and PLR were calculated accordingly. Receiver Operating Characteristic (ROC) curve and logistic regression were performed. All the scoring systems, as well as NLR and PLR, significantly increased in both the DEA and SC groups. The ROC curve showed that the CCEDRRN COVID-19 Mortality Score had the highest predictive value for mortality and severity (AUC 0.779, 0.850, respectively), which outperformed the APACHE II, SOFA and NEWS. NLR presented better predictive ability for severity (AUC 0.741) than death (AUC 0.702). The APACHE II, NEWS and CCEDRRN COVID-19 Mortality Score were positively correlated with both prognosis and severity, whereas NLR only with severity. The NEWS and CCEDRRN COVID-19 Mortality Score were reconfirmed for early and rapid predicting the poor prognosis and severity of COVID-19 patients in ED, especially the CCEDRRN COVID-19 Mortality Score with the highest discrimination capacity, and NLR was more appropriate for predicting the severity.
Read full abstract