Statistical Analysis of Drug Disposition Data

Abstract

Pharmacokinetics has become a rigorous science that attempts to relate the interaction of a drug with the biological environment into which it is introduced. Such a relationship is often described by proposing a mathematical model of the system and then using the data to define values for the unknown model parameters. Aris1 points out that “being derived from ‘modus’ (a measure) the word ‘model’ implies a change in representation.” That is, a model attempts to organize observations or measurements of the system under investigation into a form useful for hypothesis testing. The model parameters, which frequently must be determined from the observed data, function as measures for comparison of results within and between experiments. Based on this criterion, at least three tasks confront the analysts when faced with a data set: (1) to define a parametric model that best describes the system under investigation, (2) to estimate values for unknown model parameters, and (3) to determine whether the proposed model is a good or bad prototype of the system under investigation.