Comparison of monitoring techniques for intraoperative cerebral ischemia

To prospectively compare somatosensory evoked potentials, electroencephalography (EEG) and transcranial Doppler ultrasound (TCD) for detection of cerebral ischemia during carotid endarterectomy (CEA).

Somatosensory evoked potentials and EEG recordings were attempted in 156 consecutive CEAs and TCD was also attempted in 91 of them. Recordings from all three modalities were obtained for at least 10 minutes before CEA, during CEA and for at least 15 minutes after CEA. Somatosensory evoked potentials peak-to-peak amplitude decrease of >50%, EEG amplitude decrease of >75%, and ipsilateral middle cerebral artery mean blood flow velocity (mean VMCAi) decrease >75% persisting for the entire period of internal carotid artery occlusion were individually considered to be diagnostic of cerebral ischemia. Clinical neurological examination was performed immediately prior to surgery and following recovery from general anaesthesia.

Somatosensory evoked potentials, EEG, and TCD were successfully obtained throughout the entire period of internal carotid artery occlusion in 99%, 95%, and 63% of patients respectively. Two patients (1.3%) suffered intraoperative cerebral infarction detected by clinical neurological examination and subsequent magnetic resonance imaging. Somatosensory evoked potentials accurately predicted intraoperative cerebral infarction in both instances without false negatives or false positives, EEG yielded one false negative result and no false positive results and VMCAi one true positive, four false positive and no false negative results. Transcranial Doppler ultrasound detection of emboli did not correlate with postoperative neurological deficits. Nevertheless the sensitivity and specificity of each test was not significantly different than the others because of the small number of disagreements between tests.

A >50% decrease in the cortically generated P25 amplitude of the median somatosensory evoked potentials, which persisted during the entire period of internal carotid artery occlusion, appears to be the most reliable method of monitoring for intraoperative ischemia in our hands because it accurately detected both intraoperative strokes with no false positive or false negative results.