Methods for Confidence Intervals for Vaccine Efficacy

• Activity Number: 572
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Tian Dai*+ and Kerry Go and Fabrice Bailleux and Xiao Geng and Andrew Dunning and Huiling Xiong
• Companies: Emory University and Sanofi Pasteur and Sanofi Pasteur and Michigan State University and Sanofi Pasteur and Sanofi Pasteur
• Keywords: Confidence Interval ; Vaccine Efficacy ; Efficiency

Abstract

The large size and high cost of vaccine efficacy trials means that small improvements in statistical efficiency might have appreciable benefits in terms of sample size and power. Ewell published Comparing Methods for Confidence Intervals for Vaccine Efficacy in 1996, comparing eight methods. Since then, a number of other possible approaches and modifications of previous methods have been developed or become feasible. In this presentation, a review of methods for constructing exact and large sample intervals for vaccine efficacy is provided as a framework for comparison. Theoretical comparisons and simulation study are used to determine when and under what circumstances each method yields conservative, unbiased or anti-conservative lower limits of the confidence interval. Finally, it discusses the relative merits of the methods and provides suggestions on using these methods under different circumstances in terms of efficiency.