Podium Presentation Abstracts
Summary: This is a prospective multi‐centre study conducted in 3 University teaching hospitals. The aim is to evaluate a new set of warning criteria for intraoperative somatosensory evoked potential(SEP) monitoring considering both gradual and rapid SEP change. The study enrolled 703 scoliosis patients who underwent surgical correction with SEP monitoring. The proposed new warning criterion was found to provide higher specificity, sensitivity and accuracy of monitoring outcomes when compared with the conventional criteria.
Introduction: The conventional criteria for intraoperative somatosensory evoked potential (SEP) monitoring is a decrease of >50% in the amplitude and/or increase of >10% in the latency. However, it did not consider the time course of SEP change, i.e. whether the drop of amplitude is gradual or rapid. We proposed a set of new criteria which takes the time course into account would provide a more reliable method than conventional one. The criteria includes either a drop in SEP as > 10% increase in latency, or >60% amplitude decrease gradually, or a reduction in amplitude by >30% within 30 minutes. This study was designed to evaluate the reliability of intraoperative SEP monitoring based on these new criteria.
Methods: This prospective multi‐centre study was carried out in three University teaching hospitals with a consecutive series of 703 scoliosis surgeries from Jan 2003 to Dec 2007. All patients were monitored by posterior tibial nerve SEP. The monitoring outcomes were compared between the conventional criteria and the proposed criteria. Differences were analyzed using the χ2 test for proportions.
Results: Nine patients showed postoperative neurological complications. Eight of them were detected by the proposed new criteria, while only 5 were detected by the conventional criteria. Within these 9 positive cases, one patient presented permanent neurological complication, which was detected by the new criteria showing a rapid reduction of 39%. The other 8 patients presented transient motor and sensory deficits. There were only 3 false positive cases, much less than 13 false positive cases by conventional criteria. The monitoring outcomes (Table 1) showed that the proposed criteria have less false cases and higher specificity, sensitivity and monitoring accuracy than the conventional criteria.
Conclusion: The proposed warning criteria taking into consideration the rapidity of onset of SEP changes was found to decrease false‐positive warning, as well as avoiding false negative outcomes. This modified criteria were recommended as the reliable warning criteria for intraoperative spinal cord monitoring.
Significance: The proposed monitoring criteria can improve the reliability of intraoperative SEP monitoring for scoliosis surgery.