The EuroSCORE in western Denmark: a population-based study.

Area of Science:

• Cardiology
• Cardiac Surgery
• Risk Stratification

Background:

• The logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE) is a widely used tool for predicting mortality in cardiac surgery.
• Previous studies have questioned its predictive performance due to methodologic limitations.
• Evaluating its accuracy over time is crucial for clinical practice.

Purpose of the Study:

• To assess the predictive performance of the logistic EuroSCORE in a large cohort of cardiac surgery patients.
• To analyze the EuroSCORE's accuracy from 1999 to 2010, a period of significant changes in surgical practice and patient care.
• To identify potential biases in risk prediction over time.

Main Methods:

• Utilized prospectively registered, population-based data from the Western Denmark Heart Registry.
• Included 21,664 patients undergoing on-pump cardiac surgery between 1999 and 2010.
• Assessed predictive ability using Area Under the Curve (AUC) for discrimination, Hosmer-Lemeshow (HL) calibration test, and estimated-to-observed (E/O) mortality ratio.

Main Results:

• The logistic EuroSCORE demonstrated moderate-to-good discrimination with an overall AUC of 0.79.
• Calibration was poor, with a significant HL test (p < 0.01) and an overall E/O ratio of 1.9.
• Mortality was consistently overestimated across all risk levels and procedures, with increasing overestimation noted over time, particularly in low-to-medium-risk patients.

Conclusions:

• The logistic EuroSCORE offers moderate-to-good discrimination but exhibits poor calibration in predicting 30-day mortality after cardiac surgery.
• Consistent overestimation of mortality suggests improvements in surgical quality and patient care during the study period.
• The EuroSCORE's predictive accuracy needs careful consideration, especially in specific patient subgroups and over extended timeframes.