|
Abstract:
|
Investigating pharmacokinetics (PK) is an indispensable task in drug development in order to understand how externally administered compounds are absorbed, distributed, metabolised and excreted by an organism. PK behaviour is typically assessed by measuring the substance's concentration in blood, plasma or other tissues at a number of time points after administration. Based on the resulting profile, a variety of PK parameters such as the area under the concentration versus time curve (AUC) are considered. Two approaches, non-compartmental and modelling, are available for estimating PK. Modelling allows for unstructured sampling schemes, but comes at the cost of uncertainty about the correct model to use and technical difficulties in model fitting may arise. The non-compartmental approach uses minimal assumptions about the data-generating process. However, a more structured sampling design and some approximation of the curve between observed time points is needed. We discuss how optimal designs for the non-compartmental approach can be found using a simulation scheme. We explore the robustness of the design to its assumptions and introduce a mini-max design for increased robustness.
|
