LAUSR

Supplemental Digital Content is available in the text. Background: The aim of this study was to assess the diagnostic ability of near-infrared spectroscopy-monitored regional cerebral oxygen saturation (rSO2) to detect cerebral hypoperfusion during internal carotid artery (ICA) clamping compared with motor and somatosensory evoked potential (EP) monitoring. Methods: This prospective study recruited consecutive patients undergoing carotid endarterectomy under general anesthesia. Significant EP changes (defined as >50% decrease in ipsilateral somatosensory EP amplitude or disappearance of contralateral motor EP on >2 consecutive stimulations) during ICA clamping were considered a warning sign for cerebral hypoperfusion. If significant EP changes occurred, the amplitude of the EPs and simultaneous rSO2 values were recorded before therapeutic intervention. The relationship between reductions in rSO2 and EP amplitudes was analyzed using Spearman rank-correlation analysis. Receiver operating characteristic curve analysis was used to calculate the optimal cutoff value for the relative reduction in rSO2. False-positive rates were evaluated according to immediate postoperative motor outcomes. Results: A total of 203 patients were included for analysis, of whom 23 developed significant EP changes during ICA clamping. There was a positive relationship between decreases in EP amplitude and rSO2 (R 2=0.15, P=0.02). A rSO2 reduction ≥16% from baseline had the optimal diagnostic performance for the detection of cerebral hypoperfusion (area under the receiver operating characteristic curve=0.82; 95% confidence interval: 0.76-0.87). The false-positive rate was 8.9%. Conclusions: Decreases in rSO2 correlated with decreases in EP amplitude during ICA clamping. A relative reduction in rSO2 ≥16% could serve as a warning for clamping-associated cerebral hypoperfusion. The 8.9% false-positive rate is a potential clinical limitation of the use of rSO2 to predict postoperative neurological deficits.