|
Activity Number:
|
378
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #309526 |
|
Title:
|
Bayesian Spines To Infer Population Mean and Tolerance Region from Multivariate Longitudinal Data
|
|
Author(s):
|
Qianqiu Li and Yonggang Zhao*+
|
|
Companies:
|
Wyeth and Wyeth
|
|
Address:
|
500 Arcola Rd, Collegeville, PA, 19426,
|
|
Keywords:
|
bayesian ; confidence region ; tolerance region ; MCMC ; longitudinal data ; splines
|
|
Abstract:
|
Spine models are widely used for the analysis of longitudinal data. In this talk, we propose a high-dimensional spline model in mixed model framework to address two common issues arising in clinical safety studies: population mean inference and tolerance region construction for screening of individuals with unusual responses, based on multivariate longitudinal data. Bayesian analysis through MCMC samplers is employed for estimation and inference. Using multivariate central limit theorem and Monte Carlo technique, we obtain approximate confidence and tolerance regions through time.
|
