Risk stratification analysis of operative mortality in isolated coronary artery bypass graft patients in Pakistan: comparison between additive and logistic EuroSCORE models

Abstract

We compared the performances of the additive and logistic EuroSCORE in predicting mortality in patients undergoing isolated coronary artery bypass grafting at a single institution in Pakistan. Both models were applied to 2004 patients, operated upon at the Aga Khan University Hospital from January 2006 to July 2010. The actual mortality (3.8%) was significantly different from the additive (4.35%) and the logistic (6.41%) estimates. On the basis of degree of risk, actual mortality was 0.6% in the low-risk (additive EuroSCORE 0-4), 4.2% in the medium-risk (EuroSCORE 5-9) and 19.1% in the high-risk (EuroSCORE 10-19) group. With the low risks, both systems slightly overestimated mortality, with the logistic EuroSCORE being more accurate. At a EuroSCORE of between 10 and 19, the additive EuroSCORE underestimated and logistic EuroSCORE overestimated mortality. Both models satisfactorily discriminated outcomes (receiver operating characteristics areas of 0.866 and 0.859 for the additive and the logistic model, respectively). The Hosmer-Lemeshow test showed that calibration was good for the additive model (P=0.424) but turned out to be inadequate for the logistic model (P<0.001). We conclude that the additive EuroSCORE is a more accurate model for risk assessment compared to the logistics model in the Pakistani population.