Validation of nomogram performance
ROC curve was conducted to evaluate the prediction ability (Figure 2).
The AUC of this nomogram was 74.5% and the cut-off value for risk
probability in this model was 1.3, with a specificity and sensitivity of
75.3% and 63.2%, respectively. The calibration curve of this nomogram
for the risk between the actual and predicted probability was consistent
(Figure 3A). The DCA analysis evaluated the clinical value of the
nomogram (Figure 3B), which indicated that recipients with the threshold
risk of about 20 to 50% were recommended to using this nomogram.