Multicenter preoperative stroke risk index for patients undergoing coronary artery bypass graft surgery

Background The paradox of present cardiac surgery is that the more elderly and debilitated patients benefit most from cardiac surgery compared with medical therapy, yet they sustain greater overall risk for morbidity and mortality after cardiac surgery. The goal of the present study was to develop a preoperative index predicting major perioperative neurological events in patients undergoing coronary artery bypass graft surgery. Methods and Results As part of a prospective, multicenter, observational study (McSPI Research Group), we enrolled 2417 patients at 24 academic medical centers in the United States. Patients who died intraoperatively or had concomitant open-heart procedures were excluded from analysis, resulting in a total of 2107 for analysis. Sixty-eight patients (3.2%) developed adverse neurological events, defined as cerebrovascular accident, transient ischemic attack (TIA), or persistent coma. Bivariate analysis was applied to determine associations between preoperative variables and neurological events. Significant bivariate predictors were identified then logically grouped, and for each cluster, a score was calculated based on principal components. Key predictor variables were age, history of previous neurological disease, diabetes, history of vascular disease, previous coronary artery surgery, unstable angina, and history of pulmonary disease, the coefficients for which were used to develop a preoperative stroke risk index that was validated by bootstrap (c-index=0.778). Stroke risk could then be determined for each patient, calculating a patient's risk for stroke within 95% confidence intervals. Conclusions With the McSPI stroke risk index developed in this study, neurological risk can be estimated, and the most appropriate group for perioperative therapy can be identified. Further refinement and validation of this index, however, are necessary and are under way in current studies.