Sensitivity Analyses for Recurrent Event Data Trials Subject to Missing Data

• Activity Number: 395
• Type: Contributed
• Date/Time: Tuesday, August 5, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Mouna Akacha*+ and Emmanuel Ogundimu
• Companies: Novartis and University of Oxford
• Keywords: Missing Data ; Recurrent Event Data ; Informative Missingness ; Sensitivity Analyses ; Pattern-Mixture Models

Abstract

Recurrent events involve the occurrence of the same type of event repeatedly over time and are commonly encountered in longitudinal studies. Examples include seizures in epileptic studies or occurrence of cancer tumors.

One considerable challenge in analyzing such data arises when a large proportion of patients discontinues early, e.g. due to adverse events, leading to partially observed data. Any analysis of such data relies on untestable assumptions regarding the post-discontinuation behaviour of patients that drop out. Regulatory agencies are therefore increasingly asking for sensitivity analyses which assess the robustness of conclusions across a range of different assumptions.

Sophisticated sensitivity analyses for continuous data are being increasingly performed, however, this is less the case for recurrent event data. In this talk, we will present a sensitivity analysis approach for recurrent event data which bases on the pattern-mixture model framework. Different assumptions about the future behavior of dropouts dependent on reasons for dropout and received treatment can be made. We illustrate our approach with a clinical trial in patients who suffer from bladder cancer.