ROC curve for significant different parameters between nucleic acid positive and negative pneumonia group
Factors that were significantly different in above analysis (Table 1 ) were used to analyze the correlation with SARS-CoV-2 nucleic acid positive/negative diagnosis by calculating the AUC (area under curves) of ROC (receiver operating characteristic) (Figure 1 ).
It was apparent that the top three candidates were LYM%, Hb and WBC, whose AUC reached 0.844, 0.757 and 0.712 (Figure 1 ), indicating that they are good predictors to distinguish the SARS-CoV-2 nucleic acid positive from negative pneumonia patients.