Comparative Analysis of the Performance of Electroencephalogram Parameters for Monitoring the Depth of Sedation During Remimazolam Target-Controlled Infusion

Abstract

BACKGROUND: The changes in hypnotic indicators in remimazolam sedation remain unclear. We investigated the correlation of the electroencephalogram (EEG) parameters with the effect-site remimazolam concentration and the depth of sedation in patients receiving a target-controlled infusion of remimazolam. METHODS: This prospective observational study enrolled 35 patients (32 analyzed) who underwent lower extremity varicose vein surgery or lower extremity orthopedic surgery under spinal anesthesia. We administered remimazolam by target-controlled infusion using the pharmacokinetic model introduced by Schüttler et al. The EEG data were continuously recorded, including the bispectral index (BIS), patient state index (PSI), spectral edge frequency (SEF), and raw EEG signals. The relative beta ratio (RBR), defined as log (spectral power [30–47 Hz]/spectral power [11–20 Hz]), was obtained by analyzing raw EEG. The level of sedation corresponding to each effect-site remimazolam concentration was assessed using the Modified Observer’s Assessment of Alertness/Sedation (MOAA/S). The prediction probability (Pk) and Spearman’s correlation coefficients (R) were calculated between effect-site remimazolam concentration, MOAA/S, and EEG parameters. RESULTS: BIS and PSI showed significantly higher Pk for effect-site remimazolam concentration (Pk = 0.76 [0.72–0.79], P < .001 for BIS; Pk = 0.76 [0.73–0.79], P < .001 for PSI) compared to RBR (Pk = 0.71 [0.68–0.74], P < .001) and SEF (Pk = 0.58 [0.53–0.63], P = .002). BIS, PSI, and RBR showed significantly higher correlation coefficients for effect-site remimazolam concentration (R = −0.70 [−0.78 to −0.63], P < .001 for BIS; R = −0.72 [−0.79 to −0.66], P < .001 for PSI; R = −0.61 [−0.69 to −0.54], P < .001 for RBR) compared to SEF (R = −0.22 [−0.36 to −0.08], P = .002). BIS and PSI also had significantly higher Pk and correlation coefficients for MOAA/S (Pk = 0.81 [0.79–0.83], P < .001; R = 0.84 [0.81–0.88], P < .001 for BIS) (Pk = 0.80 [0.78–0.83], P < .001; R = 0.82 [0.78–0.87], P < .001 for PSI) compared to RBR (Pk = 0.74 [0.72–0.77], P < .001; R = 0.72 [0.65–0.78], P < .001) and SEF (Pk = 0.55 [0.50–0.59], P = .041; R = 0.13 [−0.01 to 0.27], P = .067). CONCLUSIONS: BIS, PSI, and RBR showed an acceptable correlation with the effect-site remimazolam concentration and depth of sedation in this study, suggesting that these EEG-derived parameters are potentially reliable hypnotic indicators during remimazolam sedation. BIS and PSI showed superior performance as hypnotic indicators to RBR and SEF in patients receiving target-controlled infusion of remimazolam.