Facing Your Fear of the Unknown - Handling Missing Data

Ideally, data collected during a clinical trial would be complete, with subjects not missing any visits or dropping out. However, this is never the case in reality. In fact, missing data is a common and serious issue for clinical trials as it can limit the conclusions drawn or camouflage safety issues with the treatment. The focus of this paper is to describe various techniques used to handle missing data, to help statisticians and statistical programmers make an informed choice on the most suitable technique to use for the data being analyzed. The paper is targeted at statisticians and statistical programmers within the pharmaceutical industry who have a basic understanding of statistical analysis.