Activity Number:
|
264
|
Type:
|
Invited
|
Date/Time:
|
Wednesday, August 14, 2002 : 8:30 AM to 10:20 AM
|
Sponsor:
|
International Indian Statisticial Assoc, Reps. For Young Statisticians
|
Abstract - #300108 |
Title:
|
Methods for Prognostic Stratification in Clinical Trials
|
Author(s):
|
Mousumi Banerjee*+
|
Affiliation(s):
|
Wayne State University
|
Address:
|
121 Shiffman Library, 4325 Brush Avenue, Detroit, Michigan, 48201, USA
|
Keywords:
|
prognostic grouping ; clinical trials ; tree-based methods
|
Abstract:
|
Identifying groups of patients with differing survival outcomes is often of interest for understanding the prognostic structure of data or for designing clinical trials. Applications can include the development of stratification schemes for future randomized studies and identification of patient subgroups for targeted therapies. In this talk, we present various methods for constructing prognostic groups based on logical combinations of covariate values. The focus is on tree-based analysis that recursively partitions the data into patient groups leading to a fitted model that is piecewise constant over regions of the covariate space. We illustrate the methods using data from a retrospective study of breast cancer patients. The thrust is on qualitative as well as empirical comparisons between the various methods presented.
|
- 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 2002 program |