The efficacy of routine completion operative angiography in reducing the incidence of perioperative stroke associated with carotid endarterectomy.

Routine operative angiography was performed during a 14 1/2-year period ending June 1982, during which 603 consecutive carotid endarterectomies were performed. For the purpose of standardization, a functional classification of stroke in terms of severity was established: class I--minimal, class II--moderate, class III--marked, class IV--severe, and class V--coma or death. Patients were analyzed to determine the cause of the stroke, the severity of the stroke, and the efficacy of routine operative angiography in reducing the incidence of perioperative stroke caused by technical error. Perioperative stroke occurred in 18 patients (2.9%), with only one having been caused by technical error. Fifteen patients underwent revision of the endarterectomy before wound closure because of unsatisfactory operative angiography results. None of the patients developed permanent neurologic deficits. Permanent perioperative neurologic deficits resulted from embolization (six patients), hypertensive episodes with cerebral hemorrhage (three patients), conversion of ischemic to hemorrhagic infarcts (two patients), spontaneous thrombosis (one patient), clamp ischemia (two patients), and other factors (three patients). Nine patients died, two of myocardial infarction and seven of stroke, for a combined mortality and morbidity rate of 3.3%. There were three class II, five class III, two class IV, and eight class V strokes that were determined to be related to the procedure. In this series stroke may have been prevented in 15 patients who underwent revision of the endarterectomy because of an unacceptable technical error demonstrated on operative angiography.