On relative risk estimation under multiple matching

Abstract

The estimation of relative risk from a rare disease in a retrospective design with multiple matching is investigated. This paper proposes a new noniterative estimator by averaging the logarithms of cell odds ratios.