[Classification tree obtained by artificial intelligence for the prediction of heart failure after acute coronary...

Area of Science:

• Cardiology
• Artificial Intelligence
• Predictive Analytics

Background:

• Coronary heart disease is the primary cause of heart failure (HF).
• Identifying patients at high risk for HF post-acute coronary syndrome (ACS) is crucial.
• Artificial intelligence (AI) shows promise in predicting cardiovascular complications.

Purpose of the Study:

• To develop a tool for identifying patients at higher risk of developing heart failure (HF) after an acute coronary syndrome (ACS).
• To leverage artificial intelligence (AI) for predicting HF development in ACS patients.

Main Methods:

• Utilized decision tree models based on recursive partitioning algorithms.
• Included 7,097 consecutive patients discharged after ACS from two Spanish centers (2006-2017).
• Collected clinical data and followed patients for a median of 53 months.

Main Results:

• Identified eight key variables predicting HF hospitalization: HF at index hospitalization, diabetes, atrial fibrillation, glomerular filtration rate, age, Charlson index, hemoglobin, and left ventricular ejection fraction.
• The AI-driven decision tree model generated 15 distinct clinical risk patterns.
• Observed a 13.6% HF readmission rate among the cohort.

Conclusions:

• AI effectively identified eight variables for predicting HF post-ACS.
• The developed decision tree model provides 15 differentiated risk patterns for HF hospitalization probability.
• An electronic application based on the model is freely available.