Online Program Home
My Program

Abstract Details

Activity Number: 122 - Clinical Trial Design and Missing Data
Type: Contributed
Date/Time: Monday, July 30, 2018 : 8:30 AM to 10:20 AM
Sponsor: Biopharmaceutical Section
Abstract #328786
Title: Impact on Statistical Power by Different Imputation Methods for Binary Endpoints with Missing Data
Author(s): Xiaomei Liao* and Jun Zhao and Bidan Huang
Companies: AbbVie Inc. and AbbVie and AbbVie Inc.
Keywords: Missing data; Non-Responder Imputation; Multiple imputation; Power

Non-Responder Imputation (NRI) is a common statistical approach for the analysis of binary efficacy variables in immunology clinical trials. These variables can take values of 'Achieved' or 'Not Achieved'. According to the NRI rule, all missing values, for any reason including discontinuation from study or switching to rescue medications, are considered as 'Not Achieved'. Even though NRI approach is simple to implement and perceived as conservative, it may cause biased estimation of treatment effect in certain situations. Multiple imputation (MI) is becoming a popular methodology as sensitivity analysis in clinical trials when dealing with missing data. In this research, impact on statistical power between the MI and NRI approaches has been evaluated, along with the comparison to observed case (OC) approach, in which missing data will not be imputed and only the observed data will be used. Different scenarios have been explored and simulation results show that the MI method appears to be an appropriate choice as a sensitivity analysis to NRI when dealing with ignorable missing data.

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

Back to the full JSM 2018 program