Specificity of transcranial motor evoked potential monitoring was significantly improved by compound muscle action potential normalization after peripheral nerve stimulation

BACKGROUND: While intraoperative transcranial motor evoked potential (tcMEP) monitoring is widely used in spinal and brain surgery, it is sometimes not used because it can exhibit low specificity and give false-positive results. METHODS: We studied 1833 muscles in 477 patients and 482 muscles in 206 patients without preoperative paralysis of manual muscle test of 2/5 or less in whom spinal and brain surgery, respectively, was performed under tcMEP monitoring. A receiver operating characteristic (ROC) analysis was used to calculate the cutoff point of amplitude relative values that cause postoperative paralysis, and the sensitivity and specificity of tcMEP monitoring with or without compound muscle action potential (CMAP) normalization after peripheral nerve stimulation. RESULTS: In spinal surgery, the rate of tcMEP amplitude reduction resulting in postoperative paralysis was 54.0% without CMAP normalization and 73.9% under CMAP normalization, with a sensitivity of 92.3% with or without CMAP normalization, a specificity of 95.8% without CMAP normalization vs. 97.4% under CMAP normalization, and this difference was significant by Fisher's Exact Probability Test (two-tailed P=0.0133, one-tailed P=0.0067). In brain surgery, the percentage of tcMEP amplitude reduction resulting in postoperative paralysis was 55.0% without CMAP normalization vs. 68.1% under CMAP normalization, with a sensitivity of 87.0% without vs. 91.3% with CMAP normalization, and a specificity of 91.3% without vs. 95.6% with CMAP normalization, and these values were significantly different by Fisher's Exact Probability Test (two-tailed P=0.0225, one-tailed P=0.0112). CONCLUSIONS: CMAP normalization after peripheral nerve stimulation may significantly increase specificity in intraoperative tcMEP monitoring during spinal and brain surgery.