Mortality prediction following CABG surgery: comparison of EuroSCORE vs general risk scoring systems

Abstract

Background. The aim of the study was to compare the accuracy of general scoring systems on ICU arrival with preoperative evaluation using the EuroSCORE risk model in patients undergoing cardiac surgery. Methods. Ethical committee approved a prospective observational study. Data for 514 consecutive CABG patients operated during one year period were collected. EuroSCORE risk score before surgery and APACHE II, SAPS II and MODS on ICU admission following surgery were calculated for all patients. Calibration of the risk stratification models was performed using the Hosmer-Lemeshow Goodness-of-Fit test, discrimination was made using Receiver Operative Characterictic (ROC) Curves. Predicted values of the risk scoring systems were compared with the actual mortality rate. Results. The observed crude hospital mortality rate of patients during the study period was 2.7%. All risk scoring systems overestimated mortality (Apache  II  –  12.2%, SAPS  II  –  5.5%, MODS  –  6.8%, EuroSCORE  –  4.5%). Hovewer, the preoperative EuroSCORE Risk Stratification System performed better than the postoperative General Severity scoring systems on ICU admission, while the APACHE II system showed the worst discriminatory capability. Conclusions. General Severity scoring systems on ICU arrival following cardiac surgery have worse discriminatory ability in predicting hospital mortality compared to the EuroSCORE Risk Evaluation System performed preoperatively.