Surface electrode somatosensory-evoked potentials in spinal surgery: implications for indications and practice.

Abstract

A retrospective study of 442 major spinal operations with spinal cord monitoring performed in a University Hospital between 1982 and 1992 was performed. To assess the reliability of the authors' method for monitoring by somatosensory-evoked potential recording, to determine criteria for intraoperative corrective action, and to redefine the need for the wake-up test. The routine use of somatosensory-evoked potential monitoring in spinal surgery remains controversial. In Nottingham, the authors have used a method of recording from either scalp or high cervical electrodes. The recordings and outcomes of all monitored spinal operations between the years 1982 and 1992 were reviewed. In 442 procedures, 23 technical failures (no reliable monitoring) occurred. Most technical failures were in patients with severe preoperative neurology, identifiable by somatosensory-evoked potential recording before operation. In the remaining 419 procedures, a significant intraoperative change in response occurred in 70 procedures (16.7%). Using the definitions of the American EEG Society, the authors identified 10 true-positives and 60 false-positives. There were no false-negatives. A wake-up test was performed if an amplitude drop greater than 50% from baseline value persisted after attempts to correct any possible identifiable intraoperative cause. This occurred in only 21 patients (5%). In the true-positive group, somatosensory-evoked potential recordings remained persistently abnormal despite an apparently normal subsequent wake-up test. The sensitivity of the method according to current definitions was 100% and the specificity 85.33%. Modified guidelines are needed for routine intraoperative use of somatosensory-evoked potential monitoring in spinal surgery. Such guidelines should avoid the term "false-positive" as currently defined and concentrate on the causative analysis of abnormal responses that warn of critical spinal cord dysfunction before that becomes irreversible and allow for appropriate action. Information from this monitoring method alerted the surgeon to the possible need for corrective action in an additional 9.78% of the reported patients, who traditionally would have been regarded as false-positives. A wake-up test still is indicated in patients with persistent suppression of their somatosensory-evoked potential despite correction of any identifiable cause and in cases of technical failure. The current method proved flexible, versatile, and reliable for future use.