Predictors of cervical bleeding after carotid endarterectomy.

BACKGROUND

The aim of this study is to determine the incidence of severe cervical bleeding requiring reintervention after carotid endarterectomy (CEA), to identify its predictive parameters, and to find out the influence of these on major complications.

METHODS

This was a retrospective review of 502 CEAs carried out in 455 consecutive patients between 1995-2011 in our institution. The end points were: postoperative cervical bleeding that required reoperation and major postoperative complications (i.e., stroke, myocardial infarction, and death). Patients' demographics, antiplatelet and anticoagulant treatment, anaesthetic technique, surgical details, and perioperative management were registered. The end point predictors were univariate and multivariate analyzed.

RESULTS

Neck bleeding after CEA occurred in 42 cases (8.4%), requiring reoperation in 28 cases (5.6%). In the univariate analysis, chronic anticoagulation and anticoagulation 24 hours before surgery were associated with reoperation for bleeding (16.6% vs. 4.8% [P = 0.02] and 17.8% vs. 4.7% [P = 0.014], respectively). The agent used for antiplatelet treatment before surgery was related to reoperation in the univariate analysis and was the only factor with statistical significance in the multivariate analysis: acetylsalicylic acid (ASA) 100 mg (2.4%), ASA 300 mg (1.5%), clopidogrel 75 mg (7.8%), ASA 100 mg associated with clopidogrel (3.3%), triflusal (5.5%), and ticlopidine (2.2%); there was a higher incidence of reoperation only in the group of patients who had taken clopidogrel 24 hours before CEA (4.7% vs. 1.05% [P = 0.06], respectively) but without statistical significance (odds ratio: 2; 95% confidence interval: 0.95-4.84). No reoperations were registered using vein patch compared to prosthetic patch (0% vs. 6.1% [P = 0.028]). Conversion to general anesthesia (22.2% vs. 4.9% [P = 0.014]) and noncontrollable postoperative hypertension (6.9% vs. 2.5% [P = 0.028]) were associated with a higher rate of reoperation. There were no statistically significant differences in the reoperation rates related to bleeding for anesthetic technique (local versus general), surgical procedure (classic endarterectomy versus eversion technique), type of prosthetic patch (Dacron/politetrafluoroethylene), use of shunt, intraoperative dose of heparin, protamine reversal, activated clotting time monitoring, or surgeon qualification level. The combined rate of stroke mortality was 2.6%. Reoperation for bleeding was not associated with an increased rate of thrombosis, stroke, death, or injury of cranial nerves.

CONCLUSIONS

Postoperative severe bleeding after carotid surgery in our institution is not an uncommon complication. Its incidence is within the range reported in the literature, but it is not associated with major complications or mortality. Antiplatelet treatment with clopidogrel is the main risk factor associated with reintervention. Other factors, such as coagulation control, postoperative hypertension management, and the use of an autologous patch, could help reduce its incidence.