Predicting cerebral ischemia during carotid endarterectomy.

We reviewed 86 consecutive patients undergoing elective carotid endarterectomy to determine whether preoperative clinical and angiographic data could be used to predict the risk of intraoperative cerebral ischemia during carotid occlusion. Electroencephalographic (EEG) monitoring with on-line Berg-Fourier transformation was carried out in all patients. A total of 32 patients (37.2%) underwent intraoperative shunting. Of these, 13 had no EEG changes but underwent shunting because of the surgeon's preference, while 19 patients underwent shunting because of EEG changes consistent with cerebral ischemia. There was one permanent (1.2%) and one transient (1.2%) neurologic deficit. Angiographic findings, clinical histories, and intraoperative EEGs were retrospectively reviewed to determine which risk factors best predicted the occurrence of intraoperative cerebral ischemia. Stroke within six weeks increased the risk of intraoperative cerebral ischemia 20-fold. Intracranial disease and contralateral carotid stenosis increased the risk of ischemia 17-fold and 16-fold, respectively. Statistical summation of all risk factors yielded a probability equation for EEG change that accurately quantitated pre-operative risk. Prospective application of this probability equation may simplify operative decision making if EEG monitoring is not available.