Abstract

Missing Data in Clinical Trials Forum

Author(s): Terry L Katz

Missing data, defined as values not available that would have been meaningful for analysis if they were observed, are commonplace in academic, government and industry run clinical trials. However, this data loss can result in an inadequate basis for study inferences. In 2014, multiple US FDA advisory committees challenged the quantity of missing data and the conclusions drawn from their study outcomes, resulting in approval delays or rejection of product registration. The Missing Data in Clinical Trials Forum invited experts from academia, government and industry to explore methods to avoid missing data by design, and present on tools to detect and rectify missing data during the trial. Statistical methodology to impute or model data missing at random, or not at random, was also discussed.