Predicting decline in left ventricular function after new-onset left bundle branch block

Area of Science:

• Cardiology
• Artificial Intelligence in Medicine
• Medical Diagnostics

Background:

• Left bundle branch block (LBBB) is linked to increased risk of left ventricular systolic dysfunction.
• Early identification of high-risk patients with LBBB is crucial for timely management.

Purpose of the Study:

• To identify predictors of significant left ventricular ejection fraction (LVEF) decline in LBBB patients.
• Utilize clinical, echocardiographic, and electrocardiographic (ECG) data for prediction.

Main Methods:

• Retrospective analysis of 769 patients with new LBBB and preserved LVEF (≥50%).
• Cox proportional hazards regression to identify predictors of ≥20% LVEF decline.
• Construction of a risk prediction model.

Main Results:

• 31.5% of patients experienced ≥20% LVEF decline over a median of 4.9 years.
• AI-derived ECG scores for LV systolic and diastolic dysfunction independently predicted LVEF decline.
• The predictive model showed modest discriminative ability (C-index 0.615-0.70).

Conclusions:

• AI-derived ECG markers for LV systolic and diastolic dysfunction predict future LVEF decline in new-onset LBBB.
• Echocardiographic parameters can further enhance risk stratification.
• This framework supports monitoring and early intervention in LBBB patients.