A Practical Model for Predicting Esophageal Variceal Rebleeding in Patients with Hepatitis B-Associated Cirrhosis

Figure 5

Decision curve analysis (DCA) of LR and XGBoost prediction models. x axis indicates the threshold probability of critical care outcome and y axis indicates the net benefit. Red solid line = XGboost model; blue solid line = LR model. Its net benefit is larger in XGBoost than in the range of the LR model: (a) DCA shows a comparison of models predicting rebleeding within 6 weeks. (b) DCA shows a comparison of models predicting rebleeding within 1 year.

(a)

(b)