Introduction:

EuroSCORE is a probabilistic model with good performance in the prognosis of mortality in heart surgery in many latitudes. It is recommendable to validate it in hospitals where it is employed.

Objective:

To validate the EuroSCORE model in adult patients at the Hospital Regional de Alta Especialidad del Bajío (HRAEB) of León, Guanajuato, Mexico.

Material and methods:

We conducted an observational transversal, and retrospective study, accomplishing this through the review of the clinical files of patients submitted to heart surgery with and without extracorporeal circulation pump from 01/01/2008 to 12/31/2013 at the HRAEB. This included mortality up to hospital discharge, utilizing the on-line calculator of the EuroSCORE program to estimate risk of death. In order to validate the EuroSCORE model, we assessed discrimination and calibration through the Area Under the receiver operating characteristics (ROC) curve and χ2 test with Hosmer-Lemeshow (H-L) goodness-of-fit, respectively.

Results:

A total of 342 patients, aged 50.02 ± 16.66 years, 181 males (52.9%) and 161 women (47.1%). The area under the ROC curve of the additive model was 0.763, and of the Hosmer-Lemeshow test was 5.30, with p = 0.62. The area under the ROC curve of the logistic model was 0.761 and of the Hosmer-Lemeshow test, 8.78, with p = 0.36.

Conclusion:

The EuroSCORE model is a reliable score for estimating the probabilities of death in adult patients submitted to heart surgery with or without the pump at the HRAEB.