Sample Size Calculation and Data Analysis Given Endpoint Choice in Studies of Continuous Glucose Monitors

• Activity Number: 281
• Type: Topic Contributed
• Date/Time: Tuesday, July 31, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Chava Zibman*+
• Companies: FDA
• Keywords: medical devices ; sample size calculation ; diabetes

Abstract

In studies of continuous glucose monitors (CGMs) and similar devices, multiple observations per subject are obtained at regular or unevenly spaced intervals. It is therefore the case that the sample size calculations and the analyses performed for these studies must account for the auto-correlation structure of the data. Adjustment for repeated measures, however, poses a challenge to many device manufacturers. This presentation will offer a closer look at the design and analysis of studies of monitoring devices and will provide an overview of how sponsors performing these studies might proceed in accounting for repeated measures. Since CGM accuracy may be expressed in terms of the mean absolute difference, mean absolute relative difference, or a binary variable indicating whether the difference between the measurements made by the test device and the reference standard fall within a specified range, we will address cases involving each of these possible endpoints.