|
Activity Number:
|
106
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
ENAR
|
| Abstract - #310229 |
|
Title:
|
A Pattern Mixture Model To Analyze Longitudinal Quality-of-Life Data with Nonignorable Dropout
|
|
Author(s):
|
Qinfang Xiang*+ and Suna Barlas
|
|
Companies:
|
Endo Pharmaceuticals and Endo Pharmaceuticals
|
|
Address:
|
100 Endo BLVD, Chadds Ford, PA, 19317,
|
|
Keywords:
|
Pattern mxiture model ; Quality of Life ; Missing not at random ; Informative dropout ; Longitudinal data analysis
|
|
Abstract:
|
The analysis of longitudinal data that suffer from missing not at random poses a statistical challenge. Inappropriate handling of the missing data that are of non-ignorable type can result in biased estimates and lead to incorrect conclusions. This study examines QoL data from a randomized clinical trial comparing the efficacy of Lidocaine Patch 5% with celecoxib 200mg in treating pain from Osteoarthritis of the knee. The QoL assessments are based on the pain interference on activities of daily living using Question9 of Brief Pain Inventory and Beck Depression Inventory. The study suffered from a significant amount of missing data (49%). A random effects pattern mixture model is applied to compare the two treatment groups in terms of the QoL indicators. This model is compared to those based on data that used the last observation carried forward, complete cases, and all observed cases.
|
