Analysis of case-cohort designs.

Title	Analysis of case-cohort designs.
Authors	L Ichikawa, John Barlow, D Rosner, S Izumi
Magazine	Journal of Clinical Epidemiology
Date	11/30/1999
DOI	10.1016/S0895-4356(99)00102-X
Introduction	The case-cohort design is valuable for studying rare failures in large cohorts. It collects data from all failures and a representative sample of censored observations, offering flexibility over nested case-control designs. Despite its efficiency, it is less commonly used due to perceived complexity. This article illustrates a simple variance estimator and model fitting techniques using SAS. It evaluates three weighting methods and applies them in an occupational study of nickel refinery workers. Compared to nested case-control designs, case-cohort sampling from the full cohort proved more efficient in analysis and simulation.
Quote	L Ichikawa, John W Barlow and D Rosner et al. Analysis of case-cohort designs. 1999. DOI: 10.1016/S0895-4356(99)00102-X
Element	Nickel (Ni)