To the Editor:—
Schüttler and Ihmsen 1have performed a massive task, the evaluation of propofol concentration–time data from a data set that is, in many ways, heterogeneous. The authors start and end their article with the suggestion that the field of target-controlled infusion (TCI) may be broadened by using their results for application of TCI in children and elderly patients. In contrast to this suggestion, and although the results may be well-applicable to children and adults, our evaluation leads us to believe that the described data set should not be used for TCI of propofol in the elderly and even may be harmful to this patient population for various reasons.
The pharmacokinetics of propofol during continuous infusion in the elderly have been described by Dyck and Shafer, 2Schnider et al. , 3and Oostwouder et al. 4A computer simulation of a simple infusion scheme (1.5 mg/kg bolus in 1 min followed by a continuous infusion of 7 mg · kg−1· h−1) based on the pharmacokinetics described by the authors shows that the concentration–time data differ significantly from those based on the three other parameter sets (fig. 1).
This discrepancy may be the result of the following.
First, the central compartment (V1) of Schüttler and Ihmsen 1is much larger compared with the previously described data sets. As a result, the initial bolus of the TCI system to reach the desired target concentration is equivalently larger. The “front end kinetics” are missed or misjudged in the Schüttler and Ihmsen 1parameter set. The larger initial bolus is especially harmful in the elderly in respect to their level of hemodynamic stability during induction. Second, during continuous infusion, the predicted propofol concentration after 360 min of administration is approximately 60% higher based on the data of Schüttler and Ihmsen, 1compared with the average propofol concentration as predicted on the basis of the other three parameter sets (fig. 1). This may be caused predominantly by the small metabolic clearance of less than 1 l/min in a typical elderly patient according to the data set by Schüttler and Ihmsen, 1compared with the 1.5 l/min described by the others. 2–4
How does this translate to the application in TCI? Obviously, the infusion rate needed to maintain a target propofol concentration of, for example, 2.5 μg/ml, is much less when the set of Schüttler and Ihmsen 1is used compared with any of the other three sets (fig. 2).
Implementing the Schüttler and Ihmsen–based infusion scheme needed to maintain a target concentration of 2.5 μg/ml in a computer simulation program provided with the Schnider and Ihmsen 3parameter set (the results are similar when the simulation program is provided with the Oostwouder et al. 4set or the Dyck and Shafer 2set) shows how low the concentration of propofol in the blood may become (1.5 μg/ml) when the population pharmacokinetic set is used for TCI in the elderly (fig. 3).
Therefore, we conclude that use of the population pharmacokinetic parameter set described by Schüttler and Ihmsen 1in a TCI setting in the elderly may cause unwanted low blood and effect-site propofol concentrations, increasing the risk of intraoperative awareness.
What may be the cause of this poor description of the propofol pharmacokinetics in the elderly? From table 1, it is clear that, of the 270 patients studied, only a small minority was aged 65 yr or older (approximately 10%), in contrast to, for example, a large group of patients aged 11 yr or younger (approximately 35%). From the 3 groups of patients that contain elderly patients (groups 3, 5, and 7), the patients from group 5 only were administered a bolus dose of propofol. Clearly, from these patients, the evaluation of the concentration–time data is less useful for the application in a continuous infusion setting, such as TCI. From the remaining elderly patients (groups 3 and 7), concentration–time data were gathered only for a mean period of 55 min. From these data, measured over such a short period, it is difficult, if not impossible, to estimate accurately the clearance or slow distribution of propofol.
Last, the article lacks a retrospective or prospective validation of the parameter set. As a result, nobody knows, also for the adults and children, whether this parameter set predicts the measured propofol concentrations better than previous parameter sets.
The lack of concentration–time data from a significant number of elderly patients who were administered propofol by continuous infusion and from whom data were gathered for an appropriate period of time (3 times the elimination half-life) resulted in a data set far different from those previously described. As a result, this population pharmacokinetic parameter set is unsuitable for application in TCI in elderly patients.