Variability of somatosensory evoked potential monitoring during scoliosis surgery

Objective: Somatosensory evoked potentials (SEPs) of 65 patients undergoing scoliosis surgery were monitored by stimulation of posterior tibial nerve to observe variations in latencies and amplitudes. Methods: Monitoring was divided into five stages: pre incision, spine exposure, instrumentation loading, deformity correction, and wound closure (stages 1-5, respectively). Results: We found the latency showed significant increase and the amplitude significant reduction from stages 1 to 2. There was no significant variability from stages 2, 3, and 4, but both latency and amplitude recovered significantly from stage 4 to 5. This variability correlated with the changes in mean arterial pressure and end-tidal concentrations of isoflurane and was not dependent on the type of surgical procedure. If either 50% amplitude reduction or 10% latency prolongation of SEP compared with baseline recordings at stage 1 (pre incision) was used as warning criterion, the overall false-positive rate was 23.1%. It was significantly reduced to 7.7% if stage 2 (spine exposure) recordings were used as the baseline (P < 0.05). The false-positive rate decreased to 0% if a combined 50% amplitude reduction and 10% latency prolongation of SEP compared with the stage 2 baseline were used (P < 0.001). Conclusion: Based on these findings, we concluded that the time to obtain SEP baseline data should be adjusted to be post incision instead of pre incision.