|
Activity Number:
|
344
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 11, 2004 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #301476 |
|
Title:
|
Spatial Summaries of Small-area Temporal Effects from an Age-period-cohort Model
|
|
Author(s):
|
Theodore R. Holford*+ and Linda W. Pickle
|
|
Companies:
|
Yale University School of Medicine and National Cancer Institute
|
|
Address:
|
60 College St., New Haven, CT, 06511,
|
|
Keywords:
|
spatial models ; age-period-cohort analysis ; Bayes method ; MCMC methods ; rates ; lung cancer
|
|
Abstract:
|
The age-period-cohort (APC) model provides useful insights into the analysis of time trends for the disease rates when one considers estimable functions of the parameters. However, when one wishes to consider whether these temporal trends are spatially consistent, the precision of the estimated effects suffer due to small numbers of cases, especially in regions with small populations. A temporal-spatial model is described in which a conditional autoregressive prior distribution is applied to the region-specific temporal effects, thereby providing a means of obtaining Bayesian estimates of estimable functions of the model parameters using Markov chain Monte Carlo techniques. Methods for extending this spatial APC model to the analysis of trends in regions with relatively small populations are described. The estimable functions that can be used to describe the temporal trends include not only the temporal-spatial smoothed rates, but estimates of drift which describe the net temporal change that is occurring in a population. This model will also be extended to allow for the inclusion of covariate information that seeks to identify factors related to the temporal trend.
|
