Activity Number:
|
347
|
Type:
|
Topic Contributed
|
Date/Time:
|
Wednesday, August 14, 2002 : 2:00 PM to 3:50 PM
|
Sponsor:
|
Section on Statistics in Epidemiology*
|
Abstract - #300797 |
Title:
|
Response Profiles for Longitudinal Clinical Trial Data with Subject Dropout
|
Author(s):
|
Craig Mallinckrodt*+ and W. Clark and Raymond Carroll and Geert Molenberghs
|
Affiliation(s):
|
Eli Lilly and Company and Eli Lilly and Company and Texas A&M University and Center for Statistics, Belgium
|
Address:
|
Lilly Corporate Center, Indianapolis , Indiana, 46285, USA
|
Keywords:
|
Missing data ; Longitudinal data ; mixed-effects models
|
Abstract:
|
Treatment effects are often evaluated by comparing change over time in outcome measures. However, valid analyses of longitudinal data can be problematic, particularly when subjects dropout prior to completing the trial for nonrandom reasons. In choosing the primary analysis for confirmatory clinical trials, regulatory agencies have for decades favored the last observation carried forward (LOCF) approach for handling missing data. Many advances in statistical methodology and our ability to implement those methods have been made in recent years. For example, likelihood-based mixed-effects model analyses are robust to the biases from subject dropout and can be easily implemented with commercially available software. The objectives of this paper are to: 1.) examine characteristics of missing and nonmissing data that influence modeling choices; 2.) examine the desired attributes of confirmatory clinical trial analyses in light of regulatory considerations; and 3.) use the data characteristics, theoretical and empirical evidence, along with logistical and regulatory considerations to develop a robust approach for choosing the primary analysis in longitudinal clinical trials.
|