ABSTRACT
BACKGROUND: A controversy exists regarding which monitoring technique is superior in cases in which general anesthesia (GA) is necessary for carotid endarterectomy (CEA). Multimodal evoked potential (mEP) monitoring was investigated under GA during CEA and compared with a historical control group undergoing neurological evaluations awake under loco-regional anesthesia (LA). METHODS: We retrospectively studied 651 patients undergoing elective CEA. In groupHISTORY (N.=349; 1997-1999) LA was provided using superficial or deep/superficial cervical plexus blocks. In groupmEP, (N.=302; 2009-2013) GA was performed by administering remifentanil/propofol infusion. The multimodal EPs included the median-nerve-somatosensory and motor evoked potentials. The primary outcome was the rate of technical failure. The arterio-arterial shunt rate and immediate postoperative motor outcomes were also compared. RESULTS: GroupmEP showed a significantly lower rate of technical failure (OR 0.17; CI 0.03-0.6; P=0.002). Because the groups differed systematically, logistic regression analysis was used to compare shunt rates and motor outcomes. Since shunt rates were 8.3% (groupmEP) versus 8.2% (groupHISTORY), but logistic regression model showed significant differences (OR 3.77; CI 1.67-8.95; P=0.001) correct comparison was impossible. Immediate postoperative deficits were 4.3% (groupmEP) and 4.9% (groupHISTORY); logistic regression analysis: transient OR 0.77, CI 0.28 to 0.22, P=0.61 and permanent OR 0.37, CI 0.02-7.74, P=0.49. CONCLUSION: Monitoring mEPs was associated with less technical failure than awake evaluation and showed similar motor outcomes. Because the groups differed systematically, the interpretation of shunt rates was impossible. Monitoring mEP should be considered to detect intraoperative ischemia in cases in which patients undergo CEA under GA.
