The analysis for case-control study in influenza

Abstract

This thesis is a simulation study for Serious Outcomes Surveillance (SOS) in influenza. We introduce some statistical methods appropriate for matched and unmatched data and describe an algorithm to simulate from a point process, and reports the results of some simulation studies which examining the performance of matched and unmatched analysis methods to assess the presence of a vaccine effect. We report further simulations which address the ability of matched and unmatched methods to accommodate additional predictor variables, and also investigates the effect of missing data in the form of missing matched controls.