Propofol-induced Electroencephalogram Dynamics: A Missing Piece

We want to congratulate Akeju et al.¹  for their interesting work on the electroencephalographic dynamics of propofol- and dexmedetomidine-induced loss of consciousness (LOC). Nonetheless, we feel that some details should be added in order to apply the provided information to the clinical practice.

The authors used an effect-site (ES) target-controlled infusion (TCI) of propofol starting with a target concentration of 1 μg/ml up to 5 μg/ml and staying 14 min in each target. However, they missed referring which pharmacokinetic model was used to calculate the ES concentrations and to drive the propofol infusion. Some authors used a similar approach in another study²  to induce LOC with propofol, where probably the Schnider model^3,4  was used and presumably LOC occurred at 2 μg/ml, which seems to be a very low ES concentration to induce LOC.^5–7 

From a pharmacokinetic/pharmacodynamic point of view, it would be interesting to correlate the electroencephalographic changes with the predicted ES concentration and the total administered propofol (e.g., at what ES concentration should we expect the occurrence of alpha band?), especially because previous evidence showed a poor correlation between predicted ES concentration calculated by the Schnider model and processed electroencephalogram–derived indices of consciousness.⁸ 

For a possible average patient (male, age 36, weight 70 kg, and height 170 cm), we simulated with Tivatrainer^® software (Gutta BV, The Netherlands, software available for download at http://www.eurosiva.eu, accessed April 22, 2015) two possible ES TCIs of propofol according to the scheme reported by the authors, using the two more common pharmacokinetic models for ES control^3,4,9  (figs. 1 and 2). The concentrations calculated by these two models have different time courses with different total administered doses of propofol: during the experimental period of 14 × 5 min, the total dose administered by the Schnider model is 659 mg of propofol while the Modified Marsh Model administers a total dose of 742 mg of propofol, as a result of different infusion rates, which seem to us to be low to induce the characteristic spectrogram for propofol.

Thus, we consider that a full spectrogram as the one resulting from dexmedetomidine infusion would be valuable information for a better comprehension of the electroencephalographic changes resulting from a stepwise approach of propofol-induced LOC: especially for those who use TCI of propofol, it would be extremely useful to know at which calculated ES concentration by a particular pharmacokinetic model is expected to occur the through-max and peak-max changes.

The authors declare no competing interests.