Delisting From Clinical Improvement in Liver Cirrhosis: A Machine Learning Decision Tree Analysis - PubMed
5 hours ago
- #Liver Cirrhosis
- #Machine Learning
- #Clinical Decision-Making
- The study examines predictors of removal from the liver transplant waitlist due to clinical improvement using machine learning on UNOS data from 2000 to 2025.
- A cohort of 127,978 LT candidates included 6.6% (8,493) delisted because of improvement. The random survival forest model showed strong performance at 1, 5, and 15 years.
- Key predictors of recovery identified were diagnosis, age, and serum albumin. These variables could aid in developing predictive models for individualized LT decision-making.