Perioperative risk and late outcome of nonelective carotid endarterectomy

AbstractPurpose: In an earlier report of our database for 1924 isolated carotid endarterectomies (CEAs) from 1989 to 1995, multivariable analysis results indicated that the urgency of operation unfavorably influenced the combined stroke and mortality rate (CSM). This study was conducted in an attempt to document the features that contribute to perioperative complications and late outcome in 314 patients for whom CEA was considered to be nonelective because of the severity of previous symptoms, carotid stenosis, or medical comorbidities. Methods: All the hospital charts and outpatient records were reviewed retrospectively for the 209 men and 105 women who had undergone nonelective CEAs (median age, 69 years). Information regarding the clinical risk factors, the operative indications (CHAT classification), the severity and distribution of carotid disease, and the surgical management were analyzed to assess the impact on the 30-day CSM and on the long-term survival rate and neurologic events during a median follow-up period of 34 months. Results: Previous symptoms had occurred in 285 patients (91%) and included cortical transient ischemic attacks in 47%, amaurosis fugax in 20%, completed strokes in 14%, unstable strokes in 2%, and nonspecific or miscellaneous symptoms in 8%. Preoperative angiography was performed in 308 patients (98%), which confirmed the presence of 80% to 99% ipsilateral carotid stenosis in 79% of the patients and >90% stenosis in 43%. The median interval between presentation and surgical treatment was 2 days, but 48% of the 314 CEAs were performed within 24 hours of presentation. The 30-day CSM was 6.7% and ranged from 3.4% for 29 patients with severe asymptomatic carotid stenosis to 14% for those patients with unstable strokes. The cardiac and pulmonary risk factors were the only variables that were related statistically to the CSM. During the follow-up period, the risk for ipsilateral stroke was significantly higher in women (risk ratio [RR], 2.38; 95% confidence interval [CI], 1.02 to 5.56; P = .04) and in patients with higher gradients of cardiac and pulmonary risk factors (RR, 2.8; 95% CI, 1.6 to 4.8 per gradient increase; P < .001). The risk was significantly lower in patients who had undergone vein patch angioplasty (RR, 0.29; 95% CI, 0.12 to 0.71; P = .006) in comparison with synthetic patching. However, 38 of the 55 patients (69%) who underwent synthetic patching also had widespread atherosclerosis for which the saphenous veins already had been harvested for coronary bypass grafting surgery or infrainguinal revascularization. Conclusion: In our experience, the perioperative risk of nonelective CEA primarily is determined by incidental cardiopulmonary disease. Vein patch angioplasty appears to enhance late results, but the late stroke rate associated with synthetic patching also may have been influenced by the extent of vascular disease in our study group. (J Vasc Surg 1999;30:618-31.