|
Activity Number:
|
292
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #309277 |
|
Title:
|
Weighting for Case-Control Studies with Partial Matched Controls
|
|
Author(s):
|
Ronaldo Iachan*+ and Aliza Fink
|
|
Companies:
|
Macro International and Macro International
|
|
Address:
|
20705 Beltsville Dr, Calverton, MD, 20705,
|
|
Keywords:
|
|
|
Abstract:
|
This analysis was motivated by a matched case control study where controls were not recruited for a large proportion of the cases. An analytic solution was explored that capitalized on the known distribution of the matching factors in the cancer registry population used to select both cases and controls. The method applied post-stratification weights to the participants to make the weighted sample reflect the total eligible population. This analysis will determine whether this method can reduce the selection bias created by matching. Our simulation study will use data on case/control status and matching factors taken from the state cancer registry. The analysis, comparing odds ratios, will be repeated for varying a) response rates and ns (controls per case), b) the association between the exposure and the outcome, c) the association between the exposure and the matching factors.
|
