Application of population-based PKPD to improve anaesthetic drug titration in the individual

Over the past decades possibilities to individualize drug titration have expanded rapidly. This development follows an era of increasing knowledge of anesthetic pharmacology. Advanced computer technology has helped to improvement drug administration techniques by providing the computing power necessary for controlling target-controlled infusion systems. However, as mentioned in chapter 1, these systems are commonly based on population-based models, which subsequently have a certain degree of error when applied to the individual. A major goal for an anaesthesist is to titrate drugs accurately. As over- and underdosing are highly undesirable, as they result in increased side effects or insufficient anaesthesia, respectively, reduction of this error seems favorable. This thesis expands our current knowledge by surveying the possibilities to optimize population-based drug titration in the individual and prospectively apply knowledge into clinical practice. The main contribution of this thesis is that it confirms that well performing population-based PK and PD models perform accurately in clinical practice and in pharmacological simulations. As such, this thesis can be used as a plea for the use of these models. Bayesian optimization provides only limited improvement on that matter. It must be emphasized that in an already well performing model, clinically significant improvements are not likely to be expected in the individual patient. The residual error might be caused by biological variability, which is hard to include in a mathematical model, even in the presence of covariates