ABSTRACT
BACKGROUND: We recommend the use of models to estimate mortality hospital risk in cardiac surgery (CC). The aim of this study was to validate a risk stratification system, widely used as the EuroSCORE (ES), in our institution. MATERIAL AND METHODS: We retrospectively analyzed the records of patients undergoing CC consecutively between January 2006 and November 2008. The ES was calculated retrospectively based on data from medical records. Mortality was considered until the time of discharge. In all patients, the ES was calculated by logistic and additive. To study the validity of the model, we analyzed their ability to calibration and discrimination through the goodness of fit test of Hosmer and Lemeshow and area under the ROC curve, respectively. RESULTS: 153 patients were included, mean age 64.1 ± 9.5 years, 77.8% men. The CRM was 60.8%, 24.8% and 14.4% valvular surgery combined (bypass + valve). Overall mortality was 3.9%. The area under the curve of the logistic model was 0.87 and the Hosmer Lemeshow test was 0.84. The area under the curve of the additive model was 0.86 and the Hosmer Lemeshow test was p = 0.89. A score greater than 7 points is additive had a sensitivity of 66% and 91,5% specificity for predicting hospital mortality. CONCLUSION: The ES is a useful model and can be used to estimate the risk of hospital mortality in CC in the city of Cordoba.
