Risk of operative mortality in surgery for coronary heart disease (a multiple regression analysis of perioperative hemodynamic and electrocardiographic data).

Using multiple regression analysis, we examined perioperative hemodynamic and electrocardiographic variables as predictors of operative mortality in surgery for coronary heart disease. Data were first analyzed as univariates and, if significantly related to mortality, they were subjected to stepwise logistic multivariate regression analysis. The preoperative predictor variables were: heart rate, ventricular arrhythmia and ST-T shift. The intraoperative predictor variables were: ventricular arrhythmia and ST-T shifts; and the postoperative predictor variables were: left ventricular stroke work index, blood pressure, mixed venous oxygen content and intrapulmonary shunt. When only electrocardiographic data were analyzed, the preoperative variables were: ventricular arrhythmia, ST-T shift and anterior wall infarction. The intraoperative variables were: ventricular arrhythmia and ST-T shift. The postoperative electrocardiogram did not give additional information. The common denominator of the relevant hemodynamic and electrocardiographic variables appears to be an accumulation of pre- and perioperative myocardial damage, which leads to operative mortality.