Univariate analysis
A comparison of patient characteristics between severe and mild patients is shown in Table 1 . A total of 12 variables were significant in the univariate analyses (P < 0.1). Considering that laboratory variables may change during patient’s hospitalization, and length of hospital stay is the result of the degree of disease. We did not include the laboratory variables and length of hospital stay in the multivariate logistic regression analysis. Finally, a total of seven variables were entered into multivariable logistic regression analysis, including age, number of comorbidities, hypertension, diabetes, cardiovascular disease, respiratory system disease, and past surgical history.