Introduction: Cardiovascular surgery risk prediction models are widely applied in medical practice. However, they have been criticized for their low methodological quality and scarce external validation. An additional limitation added in Latin America is that most of these models have been developed in the United States or Europe, which present marked geographical differences. The objective of this study is to characterize the postoperative clinical events of cardiovascular surgeries with the use of cardiopulmonary bypass pump in a local setting and to evaluate the prediction of postoperative mortality using the EuroSCORE II predictive model.
Methods: Cross-sectional study in an urban university hospital in Buenos Aires. Patients ≥21 years of age were included, with a clinical indication for on-pump cardiovascular surgery. Patients with incomplete clinical data regarding EuroSCORE II variables or in-hospital survival, ≥95 years of age, or undergoing heart transplantation were excluded.
Results: 195 patients were enrolled. Postoperative mortality estimated by EuroSCORE II presented a clear underestimation of risk (3.0% vs 7.7%). Discrimination (AUC = 0.82; 95% CI 0.74-0.92) and goodness of fit of the model were adequate (χ2 = 7.91; p = 0.4418). The most frequent postoperative complications were postoperative heart failure (35.9%), vasoplegic shock (13.3%), and cardiogenic shock (10.26%).
Conclusion: The EuroSCORE II is an appropriate tool to discriminate between different risk categories in patients undergoing on-pump cardiovascular surgery, although it underestimates the risk.