Prediction of the severity of Corona Virus Disease 2019 and its adverse clinical outcomes

Abstract

This study aims to investigate blood and biochemical laboratory findings in patients with severe Corona Virus Disease 2019 (COVID-19) and to develop a joint predictor for predicting the likelihood of severe COVID-19 and its adverse clinical outcomes, to provide more information for treatment. We collected the data of 88 patients with laboratory-confirmed COVID-19. Then patients were divided into a non-severe group and a critical group (including critically ill cases). Univariate analysis showed that the absolute lymphocyte count, albumin level, albumin/globulin (A/G) ratio, lactate dehydrogenase (LDH) level, interleukin-6 (IL-6) level, erythrocyte count, globulin level, blood glucose level, and age were significantly correlated with the severity of COVID-19. The multivariate binary logistic regression model revealed that Age, absolute lymphocyte count, and IL-6 level were independent risk factors in patients with COVID-19. The receiver operating characteristic (ROC) curve revealed that the combination of IL-6 level, absolute lymphocyte count and age is superior to a single factor as predictors for predicting severe COVID-19, regardless of whether it is the area under curve (AUC) or the prediction sensitivity and specificity. Early application is beneficial to early identification of critically ill patients and timing individual treatments to reduce mortality.