The Narcotrend monitor (MonitorTechnik, Bad Bramstedt, Germany) performs an automatic analysis of the electroencephalogram (EEG) during anesthesia based on a visual assessment of the raw EEG. Its newest software version 4.0 includes a dimensionless index that, similar to the bispectral index (BIS), ranges from 100 (awake) to 0. We compared the performance of Narcotrend index and BIS as EEG measures of anesthetic drug effect during propofol anesthesia. Eighteen adult patients scheduled for radical prostatectomy were investigated. An epidural catheter was placed in the lumbar space and electrodes for BIS (version XP; Aspect Medical Systems, Natick, MA) and Narcotrend were positioned as recommended by the manufacturers. Narcotrend index, BIS values, and propofol plasma and effect site concentrations as parallelly simulated by Rugloop software (Department of Anesthesia, Ghent University, Belgium) were automatically recorded in intervals of 5 s. Induction of anesthesia consisted of a fentanyl bolus and a propofol infusion. After endotracheal intubation, patients received 15 mL bupivacaine 0.5% epidurally, and 45 min later propofol dosages were subsequently increased and decreased twice. Simulated propofol effect site concentrations ranged from 2.0 +/- 0.4 microg/mL (smallest) to 6.3 +/- 1.3 microg/mL (largest) during these subsequent increases and decreases of propofol. In terms of prediction probability (P(K)) the performance of the Narcotrend index (P(K) = 0.88 +/- 0.03) to predict propofol effect site concentrations was comparable to the BIS (P(K) = 0.85 +/- 0.04). Using the respective EEG index as a measure of drug effect the mean k(e0) was calculated as 0.20 +/- 0.05 min(-1) for Narcotrend index and 0.16 +/- 0.07 min(-1) for BIS. In the observed propofol concentration range Narcotrend index detected differences in EEG dynamics as well as BIS.
Implications: This study in 18 adult patients undergoing radical prostatectomy describes the relationship between Narcotrend index and bispectral index versus predicted propofol effect compartment concentrations. In terms of prediction probability, the performance of the Narcotrend index and the bispectral index to predict propofol effect site concentrations was comparable.