Predictive Analysis of Cardiometabolic Risks in Liver Transplantation - A Case Study in Uruguay

Area of Science:

• Cardiology
• Hepatology
• Medical Informatics

Background:

• Cardiovascular diseases are a leading global cause of mortality.
• In Latin America, 48 million people lived with heart disease in 2021.
• Cardiometabolic risk factors significantly impact liver transplant patient survival and recovery.

Purpose of the Study:

• To analyze a cohort from Uruguay's National Liver Transplantation Program.
• To employ machine learning for predicting post-transplant cardiometabolic diseases.
• To utilize pre-transplant health indicators for risk assessment.

Main Methods:

• Evaluation of multiple machine learning models over five years.
• Utilizing the Extra Trees algorithm for predictive analysis.
• Analysis of pre-transplant clinical data for risk stratification.

Main Results:

• The Extra Trees algorithm achieved the highest predictive accuracy of 88% (AUC: 0.94).
• Demonstrated the potential of predictive analytics in risk assessment.
• Highlighted the enhancement of patient outcome prediction in liver transplantation.

Conclusions:

• This is the first national study validating machine learning for cardiometabolic risk in Uruguayan liver transplant patients.
• The model offers a data-driven approach for early risk stratification.
• Supports clinicians in mitigating post-transplant cardiometabolic complications.