Odds Ratio Estimation in 1:N Incomplete Matched Case-Control Studies

• Activity Number: 183
• Type: Contributed
• Date/Time: Monday, August 10, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Chan Jin* and Stephen Looney
• Companies: Georgia Regents University and Georgia Regents University
• Keywords: matching ; conditional logistic regression ; cross-product ratio ; simulation ; bias ; mean-squared error

Abstract

A 1:n matched case-control design, in which each case is matched to n controls is commonly used to evaluate the association between exposure to a risk factor and a disease. The odds ratio (O.R.) is typically used to quantify such an association. Difficulties in estimating the true O.R. arise when the exposure status is unknown for at least one individual in a matched case-control grouping. In the case where the exposure status is known for all individuals in the group, the true O.R. can be estimated using conditional logistic regression, among other methods. In the case where the case-control data are independent, the O.R. is estimated using the cross-product ratio from the exposure-by-disease contingency table. In this paper we suggest a simple method for estimating the O.R. when the sample consists of a combination of matched and unmatched observations, resulting from incomplete 1:n matching. This method uses a weighted average of traditional methods for estimating the O.R. with matched and unmatched data. We illustrate our method with a hypothetical example.