The efficacy of somatosensory evoked potentials in evaluating new neurological deficits after spinal thoracic fusion and decompression.

Abstract

Posterior thoracic fusion (PTF) is used as a surgical treatment for a wide range of pathologies. The monitoring of somatosensory evoked potentials (SSEPs) is used to detect and prevent injury during many neurological surgeries. The authors conducted a study to evaluate the efficacy of SSEPs in predicting perioperative lower-extremity (LE) neurological deficits during spinal thoracic fusion surgery. The authors included patients who underwent PTF with SSEP monitoring performed throughout the entire surgery from 2010 to 2015 at the University of Pittsburgh Medical Center (UPMC). The sensitivity, specificity, odds ratio, and receiver operating characteristic curve were calculated to evaluate the diagnostic accuracy of SSEP changes in predicting postoperative deficits. Univariate analysis was completed to determine the impact of age exceeding 65 years, sex, obesity, abnormal baseline testing, surgery type, and neurological deficits on the development of intraoperative changes. From 2010 to 2015, 771 eligible patients underwent SSEP monitoring during PTF at UPMC. Univariate and linear regression analyses showed that LE SSEP changes significantly predicted LE neurological deficits. Significant changes in LE SSEPs had a sensitivity and specificity of 19% and 96%, respectively, in predicting LE neurological deficits. The diagnostic odds ratio for patients with new LE neurological deficits who had significant changes in LE SSEPs was 5.86 (95% CI 2.74-12.5). However, the results showed that a loss of LE waveforms had a poor predictive value for perioperative LE deficits (diagnostic OR 1.58 [95% CI 0.19-12.83]). Patients with new postoperative LE neurological deficits are 5.9 times more likely to have significant changes in LE SSEPs during PTF. Surgeon awareness of an LE SSEP loss may alter surgical strategy and positively impact rates of postoperative LE neurological deficit status. The relatively poor sensitivity of LE SSEP monitoring may indicate a need for multimodal neurophysiological monitoring, including motor evoked potentials, in thoracic fusion surgery.