Analysis of Case-Control Studies with Sample Weights

• Activity Number: 250
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Victoria Landsman*+ and Barry Ira Graubard
• Companies: National Cancer Institute and National Cancer Institute
• Keywords: Informative sampling, ; Modeling sample weights ; Weighted estimating equations ; Sample distribution

Abstract

Analysis of population-based case-control studies with complex sampling designs is challenging because the sample selection probabilities (and, therefore, the sample weights) depend on the response variable and covariates. In this paper we propose a new semi-parametric weighted estimator which incorporates modeling of the sample expectations of the weights into a design-based framework. The estimator has a theoretical basis and is robust to model misspecification. We describe also the sample pseudo maximum likelihood estimator for inference from case-control data. This approach generalizes the Breslow and Cain (1988) estimator and it can be applied to the data from complex sampling designs, including cluster samples. We discuss benefits and limitations of each of the two proposed estimators emphasizing efficiency and robustness. We compare the final sample properties of the two new estimators and the existing weighted estimators in simulations under different sampling plans. We apply the methods to the National Cancer Institute's U.S. Kidney Cancer Case-Control Study in order to identify risk factors for kidney cancer.