MONITORING ELECTROPHYSIOLOGIC FUNCTION DURING CAROTID ENDARTERECTOMY - A COMPARISON OF SOMATOSENSORY EVOKED-POTENTIALS AND CONVENTIONAL ELECTROENCEPHALOGRAM

There is no consensus as to the most appropriate monitor for detecting ischemia during carotid endarterectomy. Accordingly, simultaneous 16-channel continuous electroencephalogram (EEG) and somatosensory evoked potential (SSEP) monitoring were performed in 64 normocapnic patients undergoing carotid endarterectomy and anesthetized with isoflurane or halothane-nitrous oxide (supplemented with fentanyl). Recordings were obtained before, during, and for 15 min after cross-clamping of the internal carotid artery. internal shunt was not used in any patient, regardless of EEG and SSEP changes. Significant amplitude reduction in the cortical component of the primary negative peak (> 50%) in SSEP occurred in 6 patients, and an increase in central conduction time (CCT) (> 1 ms) occurred in 5 patients. Major EEG changes occurred in 6 patients, 4 of whom also had SSEP changes. Two patients had transient neurologic deficits postoperatively, with both having SSEP changes (amplitude reduction > 50%), whereas one had EEG changes. Based on these observations, the relative sensitivity and specificity for EEG and SSEP (amplitude reduction > 50%) in detecting postoperative neurologic deficits were 50% and 92% for EEG and 100% and 94% for SSEP, respectively, differences that were not statistically significant. Regarding SSEP, the use of latency change (CCT) as a criterion was associated with a sensitivity of 0% (P = 0.046 from sensitivity of amplitude) and a specificity of 87% (P = 0.17 from specificity of amplitude). The authors concluded the following, regarding the use of EEG and SSEP as diagnostic tests of postoperative neurologic deficit: 1) both forms of electrophysiologic monitoring are associated with a considerable false-positive rate; 2) compared with conventional EEG, SSEP monitoring during carotid endarterectomy has a similar sensitivity and specificity; and 3) amplitude reduction greater than 50% is a better indicator than latency increase during SSEP monitoring.