Issues of Missing Data in Orthopedic Implant Clinical Trials: A Regulatory Reviewer's Perspective

• Activity Number: 61
• Type: Topic Contributed
• Date/Time: Sunday, August 1, 2010 : 4:00 PM to 5:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Jianxiong Chu*+
• Companies: FDA
• Keywords: missing data ; LOCF ; orthopedic implant

Abstract

Missing data is inevitable in most device trials despite carefully designed and executed investigational plans. In this talk, I would discuss the common missing patterns in terms of missing percentage, timing and reasons for missingness as observed from orthopedic implant trials. Potential impacts on the missing patterns from the open-label nature and the involvement of the surgeon to re-fix the implant will be emphasized. Current common practices for handling missing data will also be discussed followed by a proposal that, in contrast to a single method treating all missing data uniformly regardless of the different missing patterns, missing data could be classified into several categories of most plausible missing mechanisms and handled with a hybrid approach, i.e., apply several approaches (sensible LOCF, multiple imputation, etc.) to the different subsets of missing data.