Validation Study: Spectral EEG Entropy as a Marker of Hypnosis in Children

Introduction: EEG derived indices, like Bispectral Index (BIS) and Spectral Entropy, have been shown to be robust indicators for the depth of hypnosis during anesthesia. BIS has already been validated in both adults and children (1). Spectral Entropy has been validated in adults (2), but not in children. There-fore, the aim of this study was to validate Spectral Entropy (State Entropy, SE and Response Entropy, RE) (3) as a marker of the level of hypnosis during general anesthesia in infants and children. Methods: 20 infants (1 month - 1 year) and 40 children (1-16 years) in good general health and under-going surgery were included. Premedication was oral midazolam 0.3 mg/kg. The EEG signal already for the baseline values was collected with two sensors placed on the forehead. Anesthesia was induced with 5% sevoflurane via face mask with 70% nitrous oxide in oxygen until the loss of eyelash reflex, fol-lowed by sevoflurane 2% during the venous cannulation. Thereafter, fentanyl 2 ug/kg and rocuronium 0.5 mg/kg were administered to facilitate tracheal intubation. Maintenance of anesthesia consisted of sevoflurane in 70% N2O and 30% O2, together with fentanyl and rocuronium as needed. All the data of SE, RE, and BIS, end-tidal concentration of sevoflurane and the vital signs were collected. The level of consciousness was also measured with Modified Observer’s Assessment of Alertness/Sedation Scale (OAAS) at various phases. Relationship between SE, RE and BIS values and OAAS scale was analysed with prediction probability (pk). The pk values for different indexes were compared using the Wilcoxon rank-sum test, comparisons of pk values between different age groups were done using the Mann-Whitney test. Relationship between SE, RE, BIS and end-tidal sevoflurane was analysed with correlation coefficients. Results: When all cases were analyzed as a single group, the prediction probabilities of both SE (p-value 0.04, table, ¤) and RE (p-value 0.04, table, *) were higher than BIS. The prediction probability in infants was higher with RE than BIS (p-value 0.02, table, #). The prediction probability of BIS was better in children than infants (p-value 0.01, table, §). All these findings in the operation phase were significant. Also there was a significant inverse correlation between SE, RE, BIS values and the end-tidal sevoflu-rane. Conclusion: Our study showed that both spectral entropies (SE & RE) and BIS reflect reliably the vari-ous phases of anesthesia in infants and children. Entropy seems to perform more consistently with in-fants than BIS