Sensitivity Analysis Using Enhanced Tipping-Point Displays for Studies with Dichotomous Treatment and Partially Missing Outcomes

• Activity Number: 582
• Type: Topic Contributed
• Date/Time: Wednesday, August 7, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Victoria Liublinska*+ and Donald B. Rubin
• Companies: Harvard University and Harvard University
• Keywords: Clinical trials ; Missing data ; Multiple imputation ; Sensitivity analysis ; Tipping-point analysis

Abstract

Although recent guidelines for dealing with missing data emphasize the need for sensitivity analysis, universal recommendations for conducting and displaying these analyses are scarce. Enhanced tipping-point (ETP) displays were recently introduced to help visualize and systematize the results of sensitivity analyses for experiments with dichotomous treatment and binary outcomes. We generalize the idea to a broader range of outcome distributions and identify families of distributions for which the application of ETP displays is straightforward. We also describe a systematic way of exploring missing-not-at-random models using ETP displays, based on a pattern-mixture factorization of the outcome distribution, and present a set of sensitivity parameters that arises naturally form such a factorization.