|
Activity Number:
|
149
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #305611 |
|
Title:
|
Endpoint Selection in the Presence of Missing Data
|
|
Author(s):
|
Guoxing (Greg) Soon*+
|
|
Companies:
|
FDA
|
|
Address:
|
CDER, 10903 New Hampshire Avenue, Silver Spring, MD, 20993-0002,
|
|
Keywords:
|
endpoint ; power ; robustness ; missing ; impute ; simulation
|
|
Abstract:
|
Minimizing Missing and collect off-treatment information are the keys to quality trials. However, other design and analysis considerations are also crucial. When there is no missing data, it is often advantageous to use a finer endpoint like change from baseline to maximize the power. However, when missing is present, it is usually much harder to find a credible way for imputation than for a crude endpoint like responder analysis. Fortunately selecting a crude endpoint does not necessarily lead to less power than a finer endpoint when strategies on handling missing data are take into consideration. In this talk I will focusing on examining the endpoint selection and their sensitivity and robustness when missing is present, using HIV trials as examples.
|
- 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 2009 program |
