Sensitivity Analysis for Clinical Trials with Missing Outcome Data: a Bayesian Tipping Point Analysis

• Activity Number: 297
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Section on Medical Devices and Diagnostics
• Author(s): Ying Yang*+ and Yu Zhao and Meichun Ding
• Companies: FDA/CDRH and FDA and GlaxoSmithKline
• Keywords: missing data ; sensitivity analysis

Abstract

It is unavoidable that some outcome data are missing from confirmatory clinical trials. However, it is not an acceptable option to ignore them when planning, conducting or interpreting the analysis of a confirmatory clinical trial. The amount of missing data can be reduced by careful planning. Therefore it is important to specify a plausible approach to handling missing data in the statistical analysis plan. There is no universal approach that adjusts the analysis to take into account the missing data. Different approaches may lead to different conclusions. Whatsoever, all approached rely on unverifiable assumptions when there are missing data. Therefore it is necessary to investigate the robustness of study results through a range of sensitivity analyses based on different assumptions. In this presentation, I will present a Bayesian tipping point analysis which may incorporate site heterogeneity into the imputation of missing data using Bayesian hierachical model. An example will be used to demonstrate the approach.