Missing Data in a Large Long-Term Clinical Trial: A Case Study

• Activity Number: 391
• Type: Topic Contributed
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Dacheng Liu*+
• Companies: Boehringer Ingelheim Pharmaceuticals, Inc.
• Keywords: Missing data ; multiple imputation ; missing at random ; clinical trial ; longitudinal data

Abstract

One of the common issues in clinical trials is missing data. In particular the monotone missing arises in many longitudinal studies when patients discontinue a study early without further measurements. In this case study we give an example of missing data from a large long-term trial in patients with chronic obstructive pulmonary disease (COPD). Incorporating the longitudinal feature of the data, we used a simple parametric model under the assumption of missing at random to analyze an endpoint with large amount of missing data. Multiple imputation was also considered and compared with the parametric model.