New books about 'missing data in clinical trials'

Missing data is one of the critical issues in statistics and in clinical trials. Three new books about the missing data were written by the biostatisticians working directly in pharmaceutical and drug development fields. These three books are worth recommending.

• Michael O'Kelley and Baohdana Ratitch (2014) Clinical Trials with Missing Data: A guide for Practitioners - also see O'kelley's recent talk "Missing data in clinical trials: developments in practice"

• Craig H. Mallinckrodt (2013) Preventing and Treating Missing Data in Longitudinal Clinical Trials

• James R. Carpenter, Michael G. Kenward (2013) Multiple Imputation and Its Application (Statistics in Practice) 

For free books regarding the missing data, the following resources are available: 

• Panel on Handling Missing Data in Clinical Trials Committee on National Statistics, Division of Behavioral and Social Sciences and Education, NATIONAL RESEARCH COUNCIL (2010) The Prevention and Treatment of Missing Data in Clinical Trials

• James R. Carpenter & Michael G. Kenward (2007) Missing data in randomised controlled trials — a practical guide

Does FDA or EMA have a preferred method for dealing with missing data for when companies submit new drugs for approval?

• The EMA has a guideline document on the handling of missing data in trials titled "Guideline on Missing Data in Confirmatory Clinical Trials" 
• The FDA does not have a single document which contains its recommendations for handling missing data in clinical trials. It appears that the FDA does not plan to issue a new guidance on missing data any time soon. However, the FDA often cited NRC’s “The Prevention and Treatment of Missing Data in Clinical Trials” for directions in dealing with the missing data. 

The general trends in handling the missing data:

• Shifting from handling the missing data to preventing the missing data
• Single imputation methods are being criticized
• None of the missing data handling approach is perfect (for example, MMRM method is criticized in recent FDA advisory committee meeting)
• New missing data handling approaches are proposed and used (for example, the jump to reference (or J2R) imputation approach used by FDA statistician in Pulmonary-Allergy Drugs Advisory Committee Meeting September 10, 2013)
• When missing data becomes an issue, sensitivity analyses with different imputation approach should be on must-to-do list especially for phase III studies – it will be a problem if the contradictory results are found with different imputation methods. See Olanzapine NDA review memos