ABSTRACT
This chapter describes longitudinal dose–response models that use data at all time points to describe the time-changing effect of a dosing regimen. The choice of a longitudinal dose–response model depends both on the purpose of the model and on the degree of scientific understanding of the drug action. Many drugs show the so-called linear pharmacokinetics. The chapter also describes the most common dose–time–response models: the direct-response and the indirect-response models. Empirical models may be sufficient if the main purpose is to descriptively fit the longitudinal data for the treatment regimens included in the dose-finding trial. The chapter discusses the use of such models for the design of clinical trials. Longitudinal data from one or more clinical trials allow people to develop dose–time–response models, estimate model parameters, and predict outcomes. The chapter considers here the analysis of a dose-finding clinical trial for the monoclonal antibody canakinumab in patients with acute gouty arthritis, a painful inflammatory disease.
