|
Activity Number:
|
24
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #310237 |
|
Title:
|
Longitudinal Analysis for Post-Transplant Serum Creatinine Data
|
|
Author(s):
|
Zailong Wang*+ and Luen Lee
|
|
Companies:
|
Novartis Pharmaceuticals and Novartis Pharmaceuticals
|
|
Address:
|
50 Ridge Drive, Livingston, NJ, 07039,
|
|
Keywords:
|
mixed-effect regression ; covariance structure analysis ; phi-divergence statistic ; goodness-of-fit test ; treatment effect ; serum creatinine data
|
|
Abstract:
|
In transplant clinical studies, serum creatinine (SCr) are measured longitudinally on the same individuals to monitor and assess immunosuppressive effects on renal function. Hence, longitudinal analysis methods should be appropriate to model the SCr data for identifying predictors among immunosuppressive regimens and potential covariates of renal function by testing the corresponding hypotheses. In this paper, we propose the use of mixed-effect regression model to analyze SCr data that were collected from a Novartis transplant clinical trial. The analysis covers the selection of fixed-effect factors and variance-covariance matrix structure, and testing on treatment effects. The residuals from the chosen "best" model are analyzed for goodness-of-fit test using phi-divergence statistic which is a generalized likelihood ratio or Pearson Chi-square statistic.
|
