Prediction of cardiac risk in non-cardiac surgery.

This prospective study was carried out to develop a model for the prediction of cardiac risk in non-cardiac surgery. Detailed data were collected concerning the preoperative status of 2609 consecutive patients, who were followed closely during the postoperative course. Fatal or life-threatening cardiac complications occurred in 68 patients (2.6%). By utilizing logistic regression, a model for prediction of cardiac risk was developed. The model contained six significant preoperative predictor variables: Congestive heart failure (with 3 degrees of severity); ischaemic heart disease (with 2 degrees of severity); diabetes mellitus; serum creatinine above 0.13 mmol l-1; emergency operation; and the type of operation (two categories). With this model it seems possible to discriminate between patients with very different levels of cardiac risk.