Analysis of Gender Differences in a Medical Device Trial That Was Not Constructed for Gender Comparisons

• Activity Number: 592
• Type: Topic Contributed
• Date/Time: Thursday, August 6, 2009 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): David C. Naftel*+ and Margaret T. Foushee and Susan L. Myers
• Companies: The University of Alabama at Birmingham and The University of Alabama at Birmingham and The University of Alabama at Birmingham
• Keywords: gender differences ; clinical trials ; FDA

Abstract

Introduction: Most medical device trials are designed to power a single comparison of treatment vs. control for the primary endpoint. The differential treatment effect across gender is of great interest to the FDA, the company, clinicians and the public. This paper presents a pre-planned strategy for gender analyses. Gender Comparison Strategy: The goal is to produce a plan so that the clinician and the statistician can produce objective evidence for assessing gender differences. 1. What is the disease prevalence and incidence in both men and women? 2. Do the inclusion/exclusion criteria affect men and women differently? 3. Analyze the gender effect on the primary endpoint. Estimate the gender affect before and after adjustment for other risk factors. 4. Plot the effect size and confidence limits. Conclusion: Evaluate all evidence to produce a clinical/statistical inference.