Sevoflurane-induced changes in infants' quantifiable electroencephalogram parameters

Background Electroencephalogram (EEG) based depth of anesthesia algorithms developed in the adult population have not demonstrated the same reliability when applied to infants. This may be due to frequency changes occurring in the EEG during development. Amplitude-integrated EEG (aEEG) is based primarily in the time domain and hence may have greater utility in infants. Objective To investigate the relationship between age adjusted Minimal Alveolar Concentration (MAC) multiples and aEEG in children under 2years of age. Methods The aEEG, Spectral Edge Frequency 90% (SEF90) and Bispectral Index (BIS) were investigated in a prospective study of children <2years of age. After anesthetic induction, and caudal block administration, EEG data were collected simultaneously with BrainZ BRM2 and BIS monitors. Using a randomized crossover design, children received up to three age adjusted concentrations of sevoflurane: 0.75, 1 and 1.25 MAC. After 15min of stable anesthetic delivery EEG readings were obtained. Prediction Probability (Pk) and correlation coefficients were calculated for each EEG parameter. Results From 51 children 102 stable anesthetics concentrations were obtained. For all age groups Pk of aEEG to multiple of age adjusted MAC was <0.72 indicating a poor predictive power for aEEG. In contrast for the SEF90 and BIS there was evidence for better predictive properties in children aged between 6months and 2years, with a Pk >0.81. Conclusion The aEEG is unlikely to be a useful measure of anesthesia depth in young children.