New cutoffs for intraoperative somatosensory and motor evoked potentials during brain surgery

Abstract

We aimed to identify the most reliable cutoff points for changes in somatosensory (SEPs) and motor (MEPs) evoked potentials to predict postoperative motor deterioration after brain surgery. The data of 104 patients who had undergone brain surgery between December 2015 and December 2016 were reviewed. The muscle strength evaluated on the day before, within 48 h and 4 weeks after surgery were reviewed. The sensitivity and specificity of the existing alarm criteria of intraoperative SEP and MEP for postoperative motor deterioration were evaluated, and receiver operating characteristic (ROC) curve analysis was performed to identify the optimal cutoffs. The sensitivities of pre-existing SEPs and MEPs alarm criteria were 8.3% and 16.7%, with specificities of 96.7% and 95.7%, respectively. ROC curve analysis estimated discriminating points for latency change in SEPs and amplitude change in MEPs of 7.1% and 21.0%, respectively. With these cutoffs, the sensitivities of SEPs and MEPs were 66.7% and 91.7%, with specificities of 79.4% and 55.4%, respectively. For brain surgery, we suggest new alarm criteria with cutoff values of a 7.1% prolongation in SEP latency or a 21.0% reduction in MEP amplitude.