How Robust Are Your Trial Results? An Applied Comparison of Time-To-Event Missing Data Analyses

• Activity Number: 624 - Trial Design and Analysis Issues in Medical Devices
• Type: Contributed
• Date/Time: Thursday, August 3, 2017 : 8:30 AM to 10:20 AM
• Sponsor: Section on Medical Devices and Diagnostics
• Author(s): Joshua Rapkin*
• Companies: Abbott
• Keywords: Sensitivity analysis ; missing data ; multiple imputation ; tipping point ; clinical trial ; time-to-event data

Abstract

Background: Sensitivity analyses for missing data are necessary to assess the robustness of clinical trial results. Two missing data methodologies, proposed by a medical journal and a regulatory body, were applied to primary endpoint data from the RESPECT trial.

Methods: In the modified tipping point analysis (MTP) analysis, discontinued subjects were sequentially failed at the time of their discontinuation in both randomization groups and analyzed using the log-rank test. The Kaplan-Meier multiple imputation (KMMI) analysis imputed missing follow-up time for discontinued subjects based on systematically varied group-specific conditional cumulative incidence functions (GSCCIF).

Results: In the MTP analysis, 66% of scenarios observed device success. In the missing at random (MAR)-like case of the KMMI analysis, there was a statistically significant reduction in the primary endpoint favoring the device (HR: 0.50, 95% CI: 0.28 - 0.89, p=0.022), and, in non-MAR-like cases, the assumed GSCCIF had to be greater than 1.5 times in the device group than medical management group to fail the endpoint.

Conclusion: Both missing data analyses support the RESPECT trial efficacy results.