Search
Close
Search
Advanced Search
Search Menu
AI Discovery Assistant

Purpose: The logistic Euroscore (Euroscore I) has been shown to lack accuracy when applied to contemporary patients. The Euroscore II has therefore been recently proposed to improve the prediction of operative mortality, but external validations are scarce. We compared the predictive performances of the Euroscore I and II in our institution.

Methods: The Euroscore I and II were was computed in 5118 consecutive patients who underwent cardiac surgery over a 5-year period. Discrimination was assessed using the c-index and calibration by comparing predicted and observed mortality. Besides the overall population we also specifically studied elderly patients (age ≥80) and patients with a body mass index (BMI) <25.

Results: Mean age was 63±14 years. Mean Euroscore I was 6.8±8.7%, mean Euroscore II 4.6±6.7% and 30-day mortality 4.8% (245 patients). C-index and comparisons between predicted and observed mortality are detailed in Table 1 for the overall population and for subgroups according to age and BMI.

Open in new tab

Table 1. Predictive properties of the Euroscore I and II

Overall<80 yrs≥80 yrsBMI <25BMI ≥25
c-Index Euroscore I0.750.770.600.750.80
c-Index Euroscore II0.800.810.660.780.83
p Calibration Euroscore I<0.00010.001<0.00010.007<0.0001
p Calibration Euroscore II0.380.390.120.040.73
Overall<80 yrs≥80 yrsBMI <25BMI ≥25
c-Index Euroscore I0.750.770.600.750.80
c-Index Euroscore II0.800.810.660.780.83
p Calibration Euroscore I<0.00010.001<0.00010.007<0.0001
p Calibration Euroscore II0.380.390.120.040.73

Conclusion: The Euroscore II has indeed a better predictive performance than the Euroscore I. Its discrimination and calibration are however less satisfying in patients with BMI<25 or aged ≥80. BMI should thus be taken into account, and in the elderly, other variables such as frailty may help to estimate more accurately surgical risk.

This content is only available as a PDF.
The European Society of Cardiology
Topic:
Issue Section:
Abstracts
You do not currently have access to this article.
Download all slides