Using surgical risk scores in nonsurgically treated infective endocarditis patients

Area of Science:

• Cardiology
• Infectious Diseases
• Medical Informatics

Background:

• Infective endocarditis (IE) poses significant mortality risks.
• Predictive accuracy of surgical risk scores for medically treated IE patients remains unexplored.
• Accurate mortality prediction is crucial for guiding treatment decisions in IE.

Purpose of the Study:

• To evaluate the accuracy of established surgical risk scores in predicting in-hospital mortality for patients with infective endocarditis (IE) who did not undergo surgery.
• To determine if existing cardiac surgery risk scores can be repurposed for medically managed IE patients.
• To identify reliable scores for risk stratification in non-surgical IE management.

Main Methods:

• Retrospective analysis of 192 non-surgically treated IE patients from the AEPEI Registry (2008).
• Application of ten different risk scores: six IE-specific surgical scores, three general cardiac surgery scores, and one 6-month IE mortality score.
• Assessment of score calibration using the Hosmer-Lemeshow test and discriminatory power via Receiver Operating Characteristic (ROC) analysis.

Main Results:

• In-hospital mortality was 19.8% (38/192 patients).
• Independent predictors of mortality included age >70, Staphylococcus aureus infection, and severe sepsis.
• While most scores showed good calibration, only five scores (three IE-specific, two cardiac surgery) demonstrated satisfactory discriminatory power (AUC >0.70).

Conclusions:

• Five of the ten evaluated surgical risk scores demonstrated adequate accuracy for predicting in-hospital mortality in medically treated IE patients.
• These validated scores can aid in risk stratification and clinical decision-making for non-surgical IE management.
• The study highlights the potential utility of repurposed surgical scores in a distinct IE patient population.