P498: Reliability of different amplitude-related warning criteria for facial motor evoked potential monitoring during vestibular schwannoma surgery
2014
Introduction Facial motor evoked potentials (facial MEPs) represent an alternative to free-running electromyography to monitor the functional integrity of the facial nerve during surgery for vestibular schwannomas. Yet, warning criteria for facial MEP monitoring remain undetermined. Our goal was to define reliable amplitude criteria to interpret facial MEP changes in the light of clinical outcome. Materials and methods Since 2009 we monitored 100 patients with transcranial facial MEPs during surgery for vestibular schwannomas. Differences in facial MEP amplitude at the end of the surgery, as compared to baselines, where analysed. Three different cut-offs were selected: 30%, 50% and 80% amplitude drop. The clinical outcome was assessed looking at changes in the House–Brackmann grade at discharge from hospital, 6 months and 1 year after surgery, in comparison to the pre-operative grade. We considered a good outcome a HB worsening of no more than one grade. Results Data were available for 89 patients at discharge, 74 patients at 6 months, and 64 patients at one year. The 30% amplitude drop criteria yielded a 51% sensitivity and 82% specificity at discharge, 58% and 73%at 6 months and 63% and 71% at 1 year. The 50% criteria yielded a 39% sensitivity and 94% specificity at discharge, 58% and 87% at 6 months, and 63% and 83% at 1 year. The 80% drop criteria yielded 21% sensitivity and 96% specificity at discharge, 29% and 94% at 6 months, and 36% and 90% at 1 year. Conclusion While a 80% drop criteria appears too little sensitive, we observed no changes in sensitivity using a 50% and 30% drop criteria at the long-term follow-up. Vice versa, specificity was significantly higher in the long term using a 50% drop than a 30% drop criteria. Therefore, we privilege a 50% drop amplitude as the best warning criteria.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
0
References
0
Citations
NaN
KQI