Multivariable logistic regression
Multiple logistic regression analysis found that advanced age (OR = 1.098, 95% CI = 1.020-1.183) and number of comorbidities (OR = 6.067, 95% CI = 1.078-34.143) were significant risk factors for severe patients with COVID-19 (Table 2 ). The multiple logistic regression model is expressed as follow:
\(\text{Logit}\left(P\right)=-5.185+0.093\times age+1.803\times\text{comorbidity\ }\)Equation 1
where age means years of age, and comorbidity represents the number of comorbidities, including hypertension, diabetes, cardiovascular disease, respiratory system disease, and previous surgical history. This risk model had acceptable discriminative power, with the area under curve (AUC) of 0.930 (95% CI = 0.862-0.999) (Figure 1 ), and was well-calibrated with the Hosmer-Lemeshow χ2 statistic of 7.565 (P = 0.477).