Factors related to false- versus true-positive neuromonitoring changes in adolescent idiopathic scoliosis surgery

A retrospective study of 134 adolescent patients who underwent surgical correction of idiopathic scoliosis between June 1992 and August 1998 was conducted.

To examine factors related to changes in somatosensory-evoked potentials with or without neurogenic motor-evoked potentials.

Studies document and demonstrate threshold criteria for changes in neuromonitoring that predict changes in spinal cord function. Rates of false-negative occurrences are low, yet higher rates of false-positive findings may result.

All the patients had somatosensory monitoring, and 71 patients had both somatosensory-evoked potential and neurogenic motor-evoked potential monitoring. Gender, age, curve types, duration of surgery, type and amount of instrumentation, and amount of correction were examined for their effects on monitoring. Estimated blood volume loss as well as high and low mean arterial pressure and its variance were assessed at the start, middle, and conclusion of the procedure.

According to the findings, 122 patients (91%) had no monitoring changes and no postoperative neurologic deficit. Six patients (4.5%) had false-positive readings. Six patients had a postoperative motor or sensory deficit, all of which resolved within 18 months. False-positive readings were associated with greater variability in mean arterial pressure. No consistent predictions could be made about the incidence of cord injury if neuromonitoring changes returned to baseline before the end of surgery.

Questions remain about the predictive accuracy of somatosensory-evoked and neurogenic motor-evoked potentials. According to the findings in this study, in which there were no false-negative readings and a modest false-positive rate, continued use of these methods is recommended. Higher false-positive rates were seen in patients with greater lability in mean arterial pressure. A wake-up test is recommended for all cases in which threshold monitoring changes occur because cases of spinal cord injury may exist even when monitored variables return to baseline.