Intraoperative neuromonitoring: lessons learned from 32 case events in 2095 spine cases

Study type: Restrospective chart review Introduction: Intraoperative neuromonitoring is becoming the standard of care for many more spinal surgeries, especially with deformity correction and instrumentation. We reviewed our institution's neuromonitored spine cases over the past 4 years to see the immediate intraoperative and postoperative clinical findings when an intraoperative neuromonitoring event was noted. Objective: The main question addressed in this review is how multimodality intraoperative neuromonitoring has affected our ability to avoid potential neurological injury during spine surgery. Methods: We retrospectively reviewed 2,095 neuromonitored spine cases at one institution performed over a period of 4 years. Data from the single neuromonitoring provider (Impulse Monitoring, Inc.) at our institution was collected and any cases with possible intraoperative events were isolated. The intraoperative and immediate postoperative clinical documentation of these 32 cases were reviewed (Table 1). Results: There were 17 cases where changes noted on EMG, SSEP, and/or MEPs affected the course of the surgery, and prevented possible postoperative neurological deficits. Of these 17, five were related to hypotension, seven due to deformity correction, one screw had a low triggered EMG threshold and was repositioned, and four cases had changes related to patient positioning and external pressure (ie, brachial plexus stretch). None of the 17 cases had postoperative motor or sensory deficits (Figure 1). Four cases consisted of intradural cord biopsies or tumor resections that had various positive neuromonitoring findings that essentially serve as controls. These cases confirm that the expected changes were seen on neuromonitoring. Four cases had false-positive neuromonitoring findings due to one technical issue requiring needle repositioning, one low threshold with triggered EMG without a pedicle breach, one case had decreased MEP responses with stable SSEPs, and one case had decreased SSEPs after positioning the patient prone. None of these four cases had any postoperative deficits. Four cases showed improved SSEPs after decompression; three cervical corpectomies, and one thoracic discectomy. Three cases of lumbar instrumentation with spontaneous EMGs each had a medial screw breach without intraoperative findings (Figure 2). They all had a postoperative motor deficit (foot drop). None of these three cases had triggered EMGs performed with the index procedure. Conclusions: Overall, this review reinforces the importance of multimodality neuromonitoring for spinal surgery. The incidence of possible events in our series was 1.5%. It is difficult to determine the true incidence, since it is impossible to know of any missed events due to lack of complete documentation. In a majority of the cases with events, possible postoperative neurologic deficits were avoided by intraoperative intervention, but the possible outcomes without intervention are not known. Clearly, in the three cases with lumbar pedicle screw malposition, triggered EMGs would have likely shown low thresholds. This would allow for screw reposition, and thus avoid a postoperative lumbar radiculopathy and revision surgery. The incidence of false-positive findings was very low in this review, and unfortunately the true incidence of false-negative findings is not able to be elucidated with this database.