WARNING CRITERIA FOR INTRAOPERATIVE SOMATOSENSORY EVOKED POTENTIAL MONITORING DATA: EXPERIENCE WITH 3783 CASES

AM Padberg, MS,
JD Parkinson, MA,
KH Bridwell, MD,
TJ Wilson–Holden, MA,
BL Raynor,
LG Lenke, MD,
KD Riew, MD
St. Louis, Missouri, USA

INTRODUCTION:
A 50% reduction in response amplitude is the generally accepted indicator of significant change in intraoperative SSEP data. A 60% amplitude decrease has been the standard at this institution for the past 10 years.

PURPOSE:
To demonstrate the sensitivity and increased specificity of a 60% as opposed to 50% decrease in SSEP response amplitude. Our hypothesis is that a 60% criterion will result in fewer wake–up tests without failing to identify a neurologic deficit.

METHODS:
The true positive (data meets warning criteria; correlates with neurologic deficit) and false negative (no significant change in data; postoperative neurologic deficit) results for 3783 monitored surgical procedures from 8/91 through 12/99 were analyzed. Procedures included surgeries for spinal deformity (idiopathic or acquired), degenerative disease, trauma, neoplasm and vascular disease. SSEPs were used alone or in conjunction with other monitoring techniques in all cases. Sciatic, tibial, median and ulnar nerves were stimulated. Identical elicitation, recording and response criteria were utilized in this series. Additionally, a 1–year subset of this population (512 cases) was individually examined to compare & contrast false positive (data meets warning criteria; no neurologic deficit) rates with a 60% versus 50% criteria.

RESULTS:
1.48% (56/3783) of the cases had true positive changes in intraoperative SSEP data using the 60% criteria. These findings correlated with various surgical events (instrumentation, vessel clamping, tumor removal, changes in blood pressure). Intraoperative intervention resulted in improvement of intraoperative data & no post–op neurologic deficit in 66% (37/56) of these cases. 34% (19/56) had partial or no improvement in data with deficit in the immediate post–op period. 0.19% (7/3783) had false negative findings. 2 patients had motor only deficits and the remaining 5 demonstrated nerve root pathology. There was no incidence of SSEP data meeting the 50% but not 60% warning criteria in these 7 cases. A 0.98% (n=5) false positive rate was obtained with the 60% criteria in the 512 patient series. Application of a 50% criterion increased this rate to 3.5% (n=18). Use of a 60% versus 50% reduction resulted in 72% fewer wake–up tests in the 512 patient subset.

DISCUSSION:
Minimizing false negative & false positive outcomes maximizes specificity & surgeons' confidence in intraoperative monitoring. The use of a 50% instead of a 60% criterion would have resulted in 12 additional wake–up tests in the 512 patient series. Extrapolation of that figure across all cases in this study would result in 136 unnecessary wake–up tests. Most importantly, no neurologic deficits were specifically attributed to the use of this standard. Motor and nerve root specific deficits comprised all of the false negative outcomes, which correlates well with the known limitations of the SSEP.

CONCLUSION:
Use of 50% versus 60% criterion does not alter the sensitivity of using SSEP's alone to detect neurologic injury (88.8%). However, specificity of the 50% criterion in predicting normal neurologic outcome is 96.5% as opposed to 99.02% with a 60% standard. This standard should be used in conjunction with motor tract and nerve root monitoring to provide optimal protection of the spinal cord and peripheral nervous system. In this 3783 case series no false negative outcomes could be attributed to the use of a 60% SSEP warning criterion. By reducing the false positive findings, significantly fewer wake–up tests were performed intraoperatively.