Preliminary study to identify severe from moderate cases of COVID-19 using NLR&RDW-SD combination parameter

changzheng wang; Chengbin Li

According to the changing trend of NLR and RDW-SD, we use LDA to carry out linearly fitting on the different hematological parameters because when the development or diagnosis of a disease is affected by two mutually masked parameters, the diagnostic performance will be lower if only one parameter is used, for example one parameter with high sensitivity while the other with high specificity. The fitting of the two different parameters by ax + by + c = w into one combined parameter has the advantages of the two parameters in specificity and sensitivity at the same time and can be more effectively used for the diagnosis and prediction of diseases. In this study, after fitting analysis of different parameters, NLR&RDW-SD combined parameter is selected as the best indicator to distinguish moderate COVID-19 patients from severe cases. The AUC is up to 0.938 and the diagnostic accuracy rate up to 85.7%. The combined parameter can help clinicians prejudge the staging of patients and take effective treatment measures in advance.