Global Sensitivity Analysis for Repeated Measures Studies with Informative Drop-Out

• Activity Number: 8
• Type: Invited
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: ENAR
• Author(s): Daniel Scharfstein and Aidan McDermott*+ and William Olson and Frank Weigand
• Companies: Johns Hopkins Bloomberg School of Public Health and Johns Hopkins Bloomberg School of Public Health and Janssen and Janssen
• Keywords: Explainable drop-out ; Exponential Tilting ; G-computation; Identification ; Missing at Random ; Pattern-Mixture Model ; Selection Model

Abstract

We present a global sensitivity analysis methodology for drawing inference about the mean at the final scheduled visit in a repeated measures study with informative drop-out. We review and critique the sensitivity frameworks developed by Rotnitzky et al. (1998, 2001) and Daniels and Hogan (2008). We identify strengths and weaknesses of these approaches and propose an alternative. We illustrate our approach via a comprehensive analysis of the RIS-INT-3 trial.