Regulatory Experiences with the Use of Multiple Imputation for Missing Data in a Phase 3 Confirmatory Trial

The Mixed Model for Repeated Measures (MMRM) is widely used in clinical trials, however, its reliance on the Missing at Random (MAR) assumption and the exclusion of subjects lacking post-baseline data have been points of scrutiny, particularly due to misalignment with the Intent-to-Treat (ITT) principle. This paper presents an application of Multiple Imputation (MI) to address missing data in a hypertension clinical trial and discusses the subsequent interactions with regulatory authorities requesting additional analyses predominantly based on a Missing not at Random (MNAR) assumption. While MNAR-based approaches have been traditionally used for sensitivity analyses, we present an example demonstrating that regulatory agencies are increasingly expecting their integration into primary analyses.