Online Program Home
My Program

Abstract Details

Activity Number: 702
Type: Contributed
Date/Time: Thursday, August 4, 2016 : 10:30 AM to 12:20 PM
Sponsor: Biopharmaceutical Section
Abstract #318670 View Presentation
Title: Population-Averaged Model Versus Subject-Specific Model in the Longitudinal Binary Data Analysis
Author(s): Hua Li* and Biswajit Sen and David Ohlssen
Companies: Novartis and Novartis and Novartis
Keywords: population averaged model ; subject-specific model ; Generalized Estimation Equation (GEE) ; Generalized Linear Mixed Model (GLMM)

In the randomized controlled trials, longitudinal binary data are commonly observed and analyzed, for example, the responder analysis for the patient who achieved the Minimum Clinical Important Difference (MCID) in Asthma Control Questionnaire (ACQ) over time in respiratory area. Population averaged model via Generalized Estimation Equation (GEE) versus subject-specific model via Generalized Linear Mixed Model (GLMM) have been long debated and the discussion is still ongoing. In this presentation, we describe and review both models in the presence of missing response data. We demonstrate the application of these methods to real-world randomized controlled trials in respiratory area to compare them with each other to see which approach seems particularly advantageous with regards to treatment effects along with the corresponding parameter interpretation. The impact of the missing data is also evaluated via multiple imputations for Missing at Random (MAR) scenarios.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association