|
Activity Number:
|
231
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 5, 2008 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #301247 |
|
Title:
|
Mapping Empirical Bayes Estimates of Cancer Mortality Rates Adjusted for Smoking
|
|
Author(s):
|
Huilin Li*+ and Barry I. Graubard and Mitchell H. Gail
|
|
Companies:
|
National Cancer Institute and National Cancer Institute and National Cancer Institute
|
|
Address:
|
6120 Executive Boulevard, Rockville, MD, 20852-7242,
|
|
Keywords:
|
Empirical Bayes ; SMR ; cancer mortality rates ; Poisson Gamma random effects Model ; disease mapping ; Health Service Areas (HSA)
|
|
Abstract:
|
The purpose of our research is to use information on known risk factors to produce adjusted cancer rates that might reveal other sources of spatial variation. We extend Clayton and Kaldor's Poisson-Gamma random effects model to analyze cancer mortality rates with regression adjustment for smoking prevalence. In this way we can determine how much of the geographic variation can be explained by smoking and provide an estimate of the residual geographic variability after accounting for smoking. We compare conventional Standardized Mortality Ratio (SMR) estimates with SMR estimates adjusted for smoking and with empirical-Bayes SMR estimates from the Poisson-Gamma model that are also adjusted for smoking and are more stable for small areas. We analyze nationwide lung cancer and cardiovascular disease data in 805 geographic areas defined by Health Service Areas (HSA) level.
|
