Skip to main content

Table 4 Comparison of performance with worldwide developed scoring systems

From: Machine learning-based prediction of COVID-19 mortality using immunological and metabolic biomarkers

Model

AUC

RF model(This study)

0.98(95% CI:0.89 - 0.98)

CURB-65 [19]

0.81 (95% CI: 0.68 - 0.94)

ISARIC-4C [20]

0.79 (95% CI:0.78 - 0.79)

NEWS2 [26]

0.86(95% CI:0.84 - 0.88)

qCSI [12]

0.81 (95% CI:0.73 - 0.89)