Risk factors for spinal cord injury during surgery for spinal deformity.

BACKGROUND

Spinal cord monitoring is now considered standard care during surgery for spinal deformity. Combined somatosensory and motor evoked potential monitoring allows the detection of early spinal cord dysfunction in most patients. The purpose of the current study was to identify clinical factors that increase the risk of intraoperative electrophysical changes and to provide management recommendations.

METHODS

The records of 162 consecutive patients who underwent surgery for the treatment of spinal deformity at a tertiary referral center were reviewed. Electrophysical monitoring of these patients was considered to have been successful if reproducible signals had been obtained. Relevant electrophysical changes included a reduction, as compared with baseline, of >50% in the amplitude of the somatosensory evoked potentials; an increase, as compared with baseline, of >10% in the latency of the somatosensory evoked potentials; a loss of motor evoked potentials; and an abrupt decrease of >75% in the motor evoked potentials.

RESULTS

One hundred and fifty-one (93%) of the 162 patients were monitored successfully. Four of the eleven patients with unsuccessful monitoring had neuromuscular scoliosis. Twelve of the 151 successfully monitored patients had a true electrophysical event, and two of them were found to have new postoperative neurologic deficits that represented a change from the findings of their preoperative neurologic examination. The determined causes of these electrophysical events included curve correction in eight patients, hypotension in two, direct cord trauma in one, and malposition of a pedicle screw in one. The patients with a true electrophysical event had a significantly higher rate of neurologic events than did the patients who did not have a true electrophysical event (p < 0.001). The rate of true electrophysical events was significantly higher in the patients with cardiopulmonary comorbidities than it was in the patients with no comorbidities (p = 0.011).

CONCLUSIONS

Combined somatosensory and motor evoked potential monitoring effectively prevents neurologic injury in most children undergoing surgery for spinal deformity. Despite the potential for false-positive results, we recommend setting a low threshold for defining relevant electrophysical changes. Rapid intervention can reverse these changes and avoid potentially serious neurologic complications. Patients with cardiopulmonary comorbidities may be at a higher risk for having relevant electrophysical events.