Screening carotid ultrasonography and risk factors for stroke in coronary artery surgery patients

Background. The role of noninvasive carotid artery screening in relation to other clinical variables in identifying patients at increased risk of stroke after coronary artery bypass grafting was examined. Methods. Preoperative, intraoperative, and postoperative clinical data were prospectively collected for 1,835 consecutive patients undergoing first-time isolated coronary artery bypass grafting between March 1990 and July 1995, 1,279 of whom had screening carotid ultrasonography. All patients with postoperative neurologic events were identified and reviewed in detail. Average patient age was 65.3 years (range, 33 to 92 years), and 9.3% (171 patients) had a prior permanent stroke or transient ischemic attack. Hospital and 30-day mortality was 2.2% (41 patients). Forty-five patients (2.5%) had a transient or permanent postoperative neurologic event. The data were analyzed by stepwise logistic regression to determine the independent predictors of both significant carotid stenosis and stroke. Results. On multivariate analysis, the clinical predictors of significant carotid stenosis were age (p < 0.0001), diabetes ((p = 0.0123), female sex (p = 0.0026), left main coronary stenosis greater than 60% (p < 0.0001), prior stroke or transient ischemic attack (p = 0.0008), peripheral vascular disease (p = 0.0001), prior vascular operation (p = 0.0068), and smoking (p < 0.0001). When all variables were evaluated for those patients who underwent noninvasive carotid artery screening, the independent predictors of postoperative neurologic event were prior stroke or transient ischemic attack (p < 0.0001), peripheral vascular disease (p = 0.0037), postinfarction angina pectoris (p = 0.0319), postoperative atrial fibrillation (p = 0.0014), carotid stenosis greater than 50% (p = 0.0029), cardiopulmonary bypass time (p = 0.0006), significant aortic atherosclerosis (p = 0.0054), postoperative amrinone or epinephrine use (p = 0.0054), and left ventricular ejection fraction less than 0.30 (p = 0.0744). Conclusions. The etiology of postoperative stroke is multifactorial. Selective use of carotid ultrasonography is of value in identifying patients who are at greater risk of postoperative stroke independent of other variables and should be considered before coronary artery bypass grafting, particularly in patients with a history of neurologic event or peripheral vascular disease.