|
Activity Number:
|
44
|
|
Type:
|
Invited
|
|
Date/Time:
|
Sunday, July 29, 2007 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #308096 |
|
Title:
|
Analysis of Clustered Longitudinal Data with Applications to Clinical Dental Research
|
|
Author(s):
|
Brian G. Leroux*+
|
|
Companies:
|
University of Washington
|
|
Address:
|
Departments of Biostatistics, Dental Public Health, Seattle, WA, 98195-7475,
|
|
Keywords:
|
clustered ; longitudinal ; multilevel ; regression ; estimating equation
|
|
Abstract:
|
Clinical dental data typically has a multilevel structure with clustering of teeth within patients and multiple observations over time for each tooth. The multilevel structure and the large number of observations per patient presents challenges for formulating valid and efficient methods for fitting regression models. Generalized Estimating Equations provides a useful framework for estimation but presents challenges in specifying a working correlation structure that will yield efficient estimates and accommodate missing or censored outcomes. We consider a new approach for modeling correlation structure of clustered longitudinal data and apply it to estimating equations for fitting regression models. Using computer simulation and application to real clinical dental data, the new approach is compared to standard GEE methods in terms of precision and susceptibility to missing data bias.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
