A retrospective study of 1162 consecutive patients who underwent spinal deformity surgical procedures at our spine center from January 2010 to December 2013.Objective.
To develop and evaluate a protocol of intraoperative motor evoked potential (MEP) monitoring with the warning criteria we had established on the basis of our clinical experiences and the review of previous literature.Summary of Background Data.
Though MEPs monitoring have become widely used in spinal deformity surgery, different alarm criteria and response protocol used in different studies compromised their comparability; Furthermore, high false-positive rate of MEP reported by previous studies has become an increasingly prominent problem that will limit its clinical use and development.Methods.
The intraoperative monitoring data of 1162 consecutive patients who underwent spinal deformity surgical procedures at our spine center were retrospectively analyzed. Age, sex, diagnosis, preoperative neurological status, intraspinal anomalies, baseline MEP, and MEP change were collected. The protocol with the warning criteria we had established was used. The false-positive rate, false-negative rate, and positive predictive value were calculated.Results.
Significant intraoperative changes were seen in the MEP data in 52 (4.4%) of all the cases. In 25 cases among which, significant MEP changes were synchronously and logically associated with high-risk surgical maneuver (pedicle screw insertion, osteotomy, correction, etc.). The false-positive rate of MEP monitoring was 0.26% (3/1140), whereas the sensitivity and specificity of MEP for detection of clinically significant intraoperative cord injury were 100% and 99.7%, respectively. The positive predictive value of a MEP alert in terms of a new postoperative neurological deficit was 83.3%.Conclusion.
Our study indicates that the appropriate use of MEP monitoring based on our protocol is able to obtain satisfying sensitivity and specificity and thus provide important information for intraoperative decision making.Conclusion.
Level of Evidence: 4