Variability of somatosensory cortical evoked potential monitoring during spinal surgery.

The intraoperative variability of somatosensory cortical evoked potentials (SCEPs) has been measured for 320 consecutive spinal surgeries and found to be a function of patient diagnosis, neuromuscular status, age, and procedural factors. In many cases, it is likely that this variability severely limits the reliability and usefulness of spinal cord monitoring in detecting early cord compromise. Patients with idiopathic scoliosis, spondylolisthesis, and pseudarthrosis have the smallest spontaneous variability and strongest amplitudes, while those with congenital, paralytic scoliosis, stenosis, or tumor have very variable, weak SCEPs. Patients with neurologic disorders, particularly cerebral palsy, myelomeningocele, Friedreich's ataxia, and peripheral deficits, also have high variability and weak amplitudes. A monitoring quality scoring system is proposed that may be useful during surgery in judging how well the SCEPs can discern surgically related changes in cord function from background variations.