Intraoperative somatosensory evoked potentials as a prognostic factor of neurologic state after carotid endarterectomy.

Somatosensory evoked cervical and cortical potentials (SEP) were analyzed under general anesthesia in 106 patients undergoing carotid endarterectomy. Cortical electrical silence occurred in 5 patients without an inlying shunt; all developed a new neurologic deficit postoperatively. Analysis of the SEP in these patients revealed progredient cerebral ischemia as indicated by an increase in central conduction time (CCT) and a decrease in amplitude of the primary cortical response N20P25 resulting in a complete loss of cortical SEP later on during the clamping period. In 6 patients the insertion of a shunt restored the deteriorated SEP, these patients and those with unchanged SEP after carotid clamping showed an uneventful postoperative recovery. Taking the presence or absence of N20P25 as the sole parameter, the sensitivity of this technique was 83%, specificity 99% and predictability 83%. A normal range for CCT and amplitude of N20P25 during anesthesia and criteria for shunt insertion were developed. The presented monitoring regimen appears to be rational and is based on current concepts of cerebrovascular physiology and pathophysiology.