Fig. 3

Biomarker analysis to discriminate SARS-CoV-2 positives developing COVID-19 pneumonia from asymptomatic. A Receiver operating characteristic (ROC) curves comparing the No Pneumonia group (WHO 1) to Pneumonia (WHO 2–7) group for serum IgM, IgA and IgG concentrations and B the different combinations of IgA with the others Igs. C ROC curve of serum IgA levels together with other previously related factors to COVID-19 severity by binary logistic regression. All ROC curves were performed using IBM SPSS Statistics 21.0. D Information of logistic regression analysis of combined selected factors, including circulating IgA concentrations, to differentiate between subjects developing COVID-19 pneumonia and those without developing COVID-19 pneumonia in the COVID-19 study cohort at the time of admission. Predictor variables were coded as follows: fever no = 0, fever yes = 1; cough no = 0, cough yes = 1; dyspnea no = 0, dyspnea yes = 1. P-value < 0.05 was considered significant and marked in bold