Predicting 30-day mortality of aortic valve replacement by the AVR score

Area of Science:

• Cardiovascular Surgery
• Medical Statistics
• Risk Prediction Modeling

Background:

• Aortic valve replacement (AVR) is a critical procedure with associated mortality risks.
• Accurate prediction of 30-day mortality is essential for patient management and surgical decision-making.
• Existing risk stratification tools may require refinement for contemporary AVR patient populations.

Purpose of the Study:

• To develop and validate a straightforward risk score for predicting 30-day mortality following AVR.
• To identify key independent predictors of short-term mortality after AVR.
• To create a practical tool for clinicians to assess AVR patient risk.

Main Methods:

• A risk score was developed using a dataset of 673 patients who underwent AVR between 1990-1993.
• Four independent predictors of 30-day mortality were identified: BMI (≥30 or <20), prior coronary artery bypass grafting (CABG), and recent myocardial infarction.
• The developed AVR score was validated on a separate cohort of 673 patients who underwent AVR approximately two decades later.

Main Results:

• The AVR score categorized patients into four risk classes (I-IV) with distinct 30-day mortality rates.
• Low-risk patients (Class I) had ≤2% mortality, while high-risk patients (Class IV) had >15% mortality.
• The score demonstrated strong predictive accuracy in the validation set, with observed mortalities aligning closely with predicted ranges across all risk classes.

Conclusions:

• The AVR score is a simple and validated tool for predicting 30-day mortality after aortic valve replacement.
• The identified predictors (BMI, prior CABG, recent MI) are robust indicators of short-term risk.
• This score can aid in risk stratification and clinical decision-making for AVR procedures.