On Estimation of Vaccine Efficacy Using Validation Samples with Selection Bias

• Activity Number: 92
• Type: Invited
• Date/Time: Monday, August 7, 2006 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Michael Daniels*+ and Daniel Scharfstein and M. Elizabeth Halloran and Haitao Chu
• Companies: University of Florida and Johns Hopkins Bloomberg School of Public Health and Fred Hutchinson Cancer Research Center and Johns Hopkins Bloomberg School of Public Health

Abstract

Using validation sets for outcomes can improve greatly the estimation of vaccine efficacy (VE) in the field. Most statistical methods for using validation sets rely on the assumption that outcomes on those with no cultures are missing at random. However, often the validation sets will not be chosen at random. For example, confirmational cultures often are done on people with influenza-like illness as part of routine influenza surveillance. Vaccine efficacy estimates based on such nonrandom validation sets could be biased. Here, we propose an approach to sensitivity to selection bias when using validation sets in estimating vaccine efficacy, building on the work of Rotnitzky et al. (1998, 2001), Scharfstein et al. (1999, 2003), and Robins et al. (2000). We develop both frequentist and Bayesian approaches.