Evolution of multimodal intraoperative neuromonitoring in pediatric spinal deformity surgery over 39 years: are we better?

Abstract

To describe the incidence and characteristics of multimodal intraoperative neuromonitoring (IONM) in pediatric spinal deformity surgery between two eras (1985-2008 vs. 2008-2024). A review was conducted of 3316 consecutive pediatric spinal deformity surgeries (2008-2024) at a single institution. Neuromonitoring modalities included somatosensory-evoked potentials (SSEP), transcranial electrical motor-evoked potentials (TCeMEP), and descending neurogenic-evoked potentials (DNEP). Due to inherent cohort heterogeneity and evolving surgical standards over four decades, a descriptive analysis was prioritized to characterize shifts in monitoring performance and case complexity.  IONM alerts were more frequent in 2008-2024 at 4.1% (1 in 24) versus 2.2% (1 in 46). Highest alert rates were in kyphosis/kyphoscoliosis (14.8% vs. 9.3%) and revision procedures (9.5% vs. 3.0%). Most alerts occurred during correction/instrumentation (62.2% vs. 47%). Permanent neurologic deficits were more common in the modern cohort: 16 patients (0.48%, 1 in 207) versus 6 (0.17%, 1 in 573). False negatives occurred at similar frequencies (0.15% vs. 0.2%) but were more consequential, with 80% resulting in permanent deficits versus 14% previously. While lower in comparison to the historic era, IONM was associated with a disparity between alert frequent and observed permanent neurologic deficit rates with alerts occurring in 4.1% of cases and permanent deficits in 0.48% (historic era: 2.2 and 0.17%, respectively). However, this coincided with an increase in case complexity as evidenced by higher prevalence of three-column osteotomies and use of traction. Multimodal IONM remains a critical adjunct in pediatric spinal deformity surgery, with alert events occurring in 4.1% of cases and an observed permanent neurologic deficit rate of 0.48%. False negatives, particularly nerve root related, remain a key limitation. Observed differences between eras likely reflect evolving case complexity, surgical techniques, and monitoring practices rather than changes in intrinsic monitoring performance.