Abstract

This one-day short course presents different methodologies associated with the analysis of linear and nonlinear models for repeated measurements. There are numerous applications in the biopharmaceutical and biological sciences and the healthcare industry which require the fitting of repeated measurements to linear and nonlinear models. Such fields as population pharmacokinetics (PK) and population pharmacodynamics (PD), bioassay, studies of biological or agricultural processes, and epidemiology all entail repeated measures applications. In this course, both population-averaged (PA) and subject-specific (SS) models for continuous and discrete data are presented. Different methods of estimation, hypothesis testing, and prediction are presented for these various models. A number of examples will be presented in which the various analytical techniques will be illustrated using the software program MIXNLIN. Participants are expected to have a working knowledge of linear models and matrix algebra.

Outline

• Introduction
• Linear Models for Repeated Measurements
• Nonlinear Regression Models for Repeated Measurements
• Generalized Nonlinear Mixed-Effects Models
• Some General Estimation Results
• Estimation Based on First-Order Approximation Methods
• Comparing Estimators
• Goodness-of-Fit
• Software
• Further Examples
• Missing Data
• Comments and Discussion

Edward. F. Vonesh, voneshe@baxter.com
Technical Director, Biometrics
Applied Statistics Center
Baxter Healthcare Corporation

Course Material

Much of the material appears in the book: Linear and Nonlinear Models for the Analysis of Repeated Measurements by Vonesh and Chinchilli (Marcel Dekker, 1997) which includes the SAS based software program MIXNLIN. Handouts of all overhead material will be provided.