Fig 1.
Selection of prediction models for validation.
Table 1.
Predictors and outcomes in the four prediction models.
Table 2.
Validation of the four prediction models.
Table 3.
Univariate analysis of predictors in mild/moderate and severe disease.
Fig 2.
Discrimination of the models at the different sites.
AUROCs from validation of the four models at the KCH and OUH cohorts, and the original AUROC from development cohorts [21–23]. Also shown are the results from the external validation of the Xie and Zhang models by Gupta et al. [6]. Lines represent the 95% CIs of the AUROCs. For the development cohorts only Xie reported confidence intervals.
Fig 3.
Calibration plots for OUH and KCH after recalibration.