|
Abstract:
|
Pharmacokinetic modeling studies the dose-concentration effects. Experimental data contains temporal measurements on the concentration of the drug in the body for individuals, in response to a drug dose, and there is usually covariates such as age and weight for each individual. Nonlinear mixed effects models are commonly used to fit the data, because the concentration often declines nonlinearly in time, and mixed effects adjust for individual variation. NONMEM, SAS NLMIX, WinBUGS and nlme in R, are most widely used software to analyze pharmacokinetic data sets. In this talk we describe ways to diagnose the fit of these complex models using graphics, both static and interactive using R and GGobi, which allow us to study individual responses, and compare model fits.
|
