Abstract:
|
A critical need exists for efficient re-purposing of existing data, particularly for questions of interest where intervention studies are not practical, ethically feasible, or where preliminary evidence is needed. The nested case-control (NCC) design has traditionally been used to assess biomarker exposures from a large cohort. Our primary interest lies in exploring best practices surrounding the re-use of existing data for more than one NCC for diseases other than that of the original design. Methodologies will be illustrated by our motivating example with data from the Women's Health Initiative Observational study, exploring the relationship between long-chain omega-3 fatty acid, as measured by dietary self-report and by blood, with breast cancer risk.
|