Patient clinical characteristics
Of 2541 patients included in this study, 271 patients had severe COVID-19 and 2270 were non-severe cases. Supplemental Table 1 indicated the basic characteristics of the study cohorts. Compared with non-severe cases, patients were older (p < 0.001) and more males (p < 0.001) were found in the severe disease cohort. In addition, there were more comorbidities and smokers in severe cases (p < 0.001). The proportion of patients with panting was higher significantly in the severe cases (p < 0.001). Significantly higher levels of white blood cell (WBC) count, neutrophil count, CRP, PCT, IL-6, total bilirubin (TBIL), alanine aminotransferase (ALT), aspartate aminotransferase (AST), lactate dehydrogenase (LDH), γ-glutamyl transpeptidase (γ-GT), and creatinine (Cre) were identified in severe cases (p < 0.001), while the levels of lymphocyte count, platelet (PLT), and oxygen saturation (SaO2) of patients with severe diseases were significantly lower than those of non-severe ones (p < 0.001). There were 186 patients with no manifestations of chest CT in non-severe COVID-19 cohort, and no significant differences of the findings of chest CT between severe and non-severe cases were found (p > 0.05). By the end of April 4, 2020, there were 58 and 39 patients died in the severe and non-severe disease group, respectively. (Supplemental Table 1).
The basic characteristics of the severe patients are listed in Supplemental Table 2. The median age of the patients was 68 (range 20 – 100) years. One hundred and fifty-two (56.09%) patients were males and there were one hundred nineteen (43.91%) female patients. There were 73 (26.94%) patients who had a high white blood cell (WBC) count of >10*10^9/L, and 148 patients (54.61%) with lymphopenia defined as lymphocyte count of ≤1.0*10^9/L. Ninety-five (35.06%) patients had a high neutrophil count of >6.3*10^9/L, while 26 (9.59%) patients had a low platelet (PLT) count of <100*10^9/L. Ground-glass opacity and consolidation was found in 122 (45.02%) and 114 (42.07%) patients, respectively. In addition, Twenty-eight (10.33%) and seven (2.58%) patients had thickened interlobular septa and nodular lesions, respectively in chest CT. Of 271 severe patients, there were 58 patients died during the study period (Supplemental Table 2).
Comparison of baseline characteristics between patients in training and validation cohorts can be seen in Table 1. There were significant differences of the age, proportion of smokers, incidence of panting, white blood cell count (WBC), neutrophil count, CRP and SaO2 at admission between the three cohorts (p < 0.05). There was no significant difference in the other variables between the three cohorts (p > 0.05). By the end of April 4, 2020, 22 severe COVID-19 patients died in the training group, and 22 and 14 patients died in the validation group 1 and validation group 2, respectively (Table 1).
The baseline characteristics of patients in the training cohort were shown in Table 2. There were no significant differences in gender, TB, ALT, AST, LDH, γ-GT, Cre, PLT, and the proportion of smoker between survivors and non-survivors (p > 0.05). Survivors were significantly younger than the non-survivors in the training cohort (p < 0.05), however, the proportion of patients with multiple comorbidity and panting (breathing rate ≥30/min) was significantly higher in non-survivors (p < 0.05). In addition, WBC and neutrophil count, CRP, D-dimer, PCT, and IL-6 was also significantly higher in non-survivors (p < 0.05). The lymphocyte count was significantly lower in non-survivors (p < 0.05).
Independent High-risk Factors Associated with the Fatal Outcome
All variables listed in Table 1were analyzed by univariate and multivariate Cox regression analysis. Multivariate Cox analysis indicated that age ≥ 70 years (HR 1.184, 95% CI 1.061-1.321), Panting(breathing rate ≥ 30/min) (HR 3.300, 95% CI 2.509-6.286), lymphocyte count < 1.0 × 109/L (HR 2.283, 95% CI 1.779-3.267), and IL-6 >10pg/mL (HR 3.029, 95% CI 1.567-7.116) were independent risk factors associated with fatal outcomes (Table 3).
Survival Analysis in the Patients with High Level of IL-6
Due to high level of IL-6 correlating with poor outcomes in severe COVID-19 patients, the therapeutic effect of tocilizumab in the patients with high IL-6 was further analyzed. In the training cohort, it was demonstrated that the prognosis of patients receiving tocilizumab was better than the that of patients not receiving tocilizumab, but without significance (p = 0.105) (Supplemental Figure.2A). Similar results were also observed in the validation cohort 1 and validation cohort 2, respectively (p = 0.133, p = 0.210) (Supplemental Figure.2B-C).
Construction and validation of the nomogram
Four independent risk factors found to be associated with the risk of mortality of patients in the multivariate analyses were incorporated into the nomogram (Figure.1). The ROC curve was employed to assess the predictive ability of the established nomogram, and the result demonstrated that the AUC was 0.900 (95% CI: 0.841-0.960) in the taring cohort, with a sensitivity of 95.5% and specificity of 77.5% (Figure. 2A). Moreover, the calibration curves for nomogram predicted mortality indicated that a good consistency between observed actual outcomes and predicted ones in the training cohort (Figure.3A).
In the validation cohort 1, the AUC was 0.811 (95%CI: 0.763-0.961) for patients with a sensitivity of 77.3% and specificity of 73.5% (Figure. 2B). In the validation cohort 2, the AUC was 0.862 (95%CI: 0.698-0.924) for patients with a sensitivity of 92.9% and specificity of 64.5% (Figure. 2C). The calibration curves also showed good agreement between prediction and observation in the risk of mortality in the two validation cohorts (Figure.3B-C).
Clinical application of the nomogram
DCA based on the net benefit and threshold probabilities was performed to assess the clinical applicability of the risk prediction nomogram. The DCA showed that our risk prediction nomogram had a superior net benefit with a wide range of threshold probabilities in the training cohort and validation cohorts (Figure. 4A-C).