Proper addressing missing data in clinical trial analysis remains complex and challenging, despite a great amount of research that has been devoted to this topic. Conventionally, under the missing at random (MAR) assumption, we often use maximum likelihood or multiple imputation based methods for inferences. However, the MAR assumption is unverifiable from data and has been deemed by regulatory agencies to be overly-simplistic and unrealistic. On the other hand, missing not at
random (MNAR) demands more sophisticated modeling treatments and estimation techniques. This tutorial covers various methods that have been advocated in dealing with missing data and illustrates how to carry out the analyses using SAS software.
The tutorial begins with an overview of missing data classifications and common analysis methods. We will use one case study of real clinical trials data and apply various analytical approaches to analyze the missing data. Methods include maximum likelihood methods, multiple imputation, generalized estimation equation approach, and Bayesian methods. The properties, advantages, and flexibilities of each method will be discussed. The last part of the tutorial is devoted to more
recently-developed methods, such as alternative and sensitivity analysis to assess robustness, control-based imputation and tipping point analysis,
Examples will be presented using SAS/STAT software, including PROC MIXED, PROC MI, PROC GEE, and PROC MCMC.
1. Review missing data analysis
2. Common analysis methods with example:
* Restricted maximum likelihood (REML)
* Mixed-effects Model Repeated Measure (MMRM),
* constrained Longitudinal Data Analysis (cLDA)
* GEE and wGEE
* Multiple imputation (MI)
* Bayesian approach
3. Recently-developed methods
* Alternative and sensitivity analysis models
* Control-based imputation
* Tipping point analysis
Guanghan Frank Liu is a distinguished scientist at Merck & Co. Inc. For the past 20 years, he has worked in a variety of therapeutic areas, including neuroscience, psychiatry, infectious disease, and vaccines. His research interests include methods for longitudinal trials, missing data, safety analysis, and noninferiority trials. Frank is a Fellow of ASA, and an associate editor for Journal of Biopharmaceutical Statistics and for Statistics in Biosciences. He also co-led a Bayesian missing data analysis subteam in the DIA Bayesian Working Groups. He has taught short courses previously at Deming conference and Regulatory-Industry workshop.
Fang Chen is a senior manager of Bayesian statistical modeling in the Advanced Analytics Division at SAS Institute Inc. Among his responsibilities are development of Bayesian analysis software and the MCMC procedure. Before joining SAS Institute, he received his PhD in statistics from Carnegie Mellon University in 2004. He has taught many short courses previously at statistical meetings including JSM, ENAR,ICSA Applied Statistics Symposium, and Regulatory-Industry workshop.