Early Prediction of Hepatic Decompensation in Cirrhosis Using Optimised XGBoost Models at the Initial Outpatient Hepatology Visit.

Key Points

• To determine the effectiveness of optimised XGBoost models in predicting hepatic decompensation among patients with compensated cirrhosis.
• Utilised optimised XGBoost models for prediction
• Assessed predictive accuracy during initial outpatient visits
• Focused on patients with compensated cirrhosis
• Models showed high predictive accuracy for medium- and long-term outcomes
• Facilitated early risk stratification
• Enabled personalised management strategies to avert clinical deterioration