Somatosensory-evoked potential monitoring during instrumented scoliosis corrective procedures: validity revisited.

Background: Intraoperative monitoring (IOM) using somatosensory-evoked potentials (SSEPs) plays an important role in reducing iatrogenic neurologic deficits during corrective pediatric idiopathic procedures for scoliosis. However, for unknown reasons, recent reports have cited that the sensitivity of SSEPs to detect neurologic deficits has decreased, in some to be less than 50%. This current trend, which is coincident with the addition of transcranial motor-evoked potentials, is surprising given that SSEPs are robust, reproducible responses that were previously shown to have sensitivity and specificity of >90%.

Purpose: Our primary aim was to assess whether SSEPs alone can detect impending neurologic deficits with similar sensitivity and specificity as originally reported. Our secondary aim was to estimate the potential predictive value of adding transcranial motor-evoked potentials to SSEP monitoring in idiopathic scoliosis procedures.

Design: This was a retrospective review to analyze the efficacy of SSEP monitoring in the group of pediatric instrumented scoliosis fusion cases.

Patient Sample: We retrospectively reviewed all consecutive cases of patients who underwent idiopathic scoliosis surgery between 1999 and 2009 at Children's Hospital of Pittsburgh. We identified 477 patients who had the surgery with SSEP monitoring alone. Exclusion criteria included any patients with neuromuscular disorders or unreliable SSEP monitoring. Patients who had incomplete neurophysiology data or incomplete postoperative records were also excluded.

Outcome Measures: Major outcomes measured were clinically significant postoperative sensory or motor deficits, as well as significant intraoperative SSEP changes.

Methods: Continuous interleaved upper- and lower-extremity SSEPs were obtained throughout the duration of all procedures. We considered a persistent 50% reduction in primary somatosensory cortical amplitude or a prolongation of response latency by >10% from baseline to be significant. Persistent changes represent significant deviation in SSEP amplitude or latency in more than two consecutive averaged trials. Patients were classified into one of four categories with respect to SSEP monitoring: true positive, false positive, true negative, and false negative. The sensitivity, specificity, positive predictive value, and negative predictive value were then calculated accordingly.

Results: Our review of 477 idiopathic scoliosis surgeries monitored using SSEPs alone revealed a new deficit rate of 0.63% with no cases of permanent injury. Sensitivity = 95.0%, specificity = 99.8%, positive predictive value = 95%, negative predictive value = 99.8%. Using evidence-based epidemiologic measures, we calculated that the number needed to treat was 1,587 patients for one intervention to be performed that would have been missed by SSEP monitoring alone. In addition, the number needed to harm, which represents the increase in false positives with the addition of transcranial electrical motor-evoked potentials, was 200.

Conclusion: SSEP monitoring alone during idiopathic scoliosis continues to be a highly reliable method for the detection and prevention of iatrogenic injury. Our results confirm the high sensitivity and specificity of SSEP monitoring alone published in earlier literature. As such, we suggest the continued use of SSEP alone in idiopathic scoliosis surgeries. At this time we do not believe there are sufficient data to support the addition of MEP monitoring, although more studies and revised criteria for the use of MEP may provide added value for its use in the future.