Fig. 3

Relevance to clinical prognosis and percentage of complications in each group during in-patient treatment of COVID-19. a Univariate analysis of the proportion of poor prognosis cases in each group. b Comparison of odds ratios for poor clinical outcomes in each group. c Multivariate logistic regression analysis of the relationship between respiratory symptoms and critical outcomes in the whole cohort (adjusted for age, sex, BMI, smoking history and comorbidities (hypertension, cardiovascular disease, chronic kidney disease and chronic liver disease). d Univariate analysis of the proportion of patients with bacterial infection in each group. e Univariate analysis of the proportion of patients with heart failure in each group. f Univariate analysis of the proportion of patients with acute kidney injury in each group. g Univariate analysis of the proportion of patients with thrombosis in each group