Gm. Jacquez et Li. Kheifets, SYNTHETIC CANCER VARIABLES AND THE CONSTRUCTION AND TESTING OF SYNTHETIC RISK MAPS, Statistics in medicine, 12(19-20), 1993, pp. 1931-1942
Citations number
14
Categorie Soggetti
Statistic & Probability","Medicine, Research & Experimental","Public, Environmental & Occupation Heath","Statistic & Probability
Cancer cluster investigations are usually univariate in nature; they f
ocus on a particular cancer, such as leukaemia, and attempt to determi
ne whether excess risk is associated with a suspected cancer-causing a
gent. Although several causes of death (such as leukaemia, lymphoma, H
odgkin's) may be considered, the approach is univariate because the ca
uses of death are analysed sequentially and independently of one anoth
er. This approach is consistent with a one-cause one-effect model. Rar
ely, however, is the action of a carcinogen manifested at only one bod
y site, and correlations among causes of death are the norm rather tha
n the exception. A multiple effects model is therefore appropriate, an
d the multivariate nature of cancer mortality data should be exploited
when exploring geographic pattern in cancer risks. This paper describ
es such an approach. We construct maps based on a principal components
analysis of cancer mortality rates from different geographic areas. T
he resulting principal components are called synthetic cancer variable
s (SCVs), and maps of the SCV scores are synthetic risk maps (SRMs). T
hese maps quantify geographic variation in cancer risk at several body
sites simultaneously, and may be analysed for (1) spatial structure a
nd (2) geographic association with potential risk factors. As an examp
le, we use synthetic risk maps to determine whether high-risk counties
in Illinois cluster near nuclear facilities. Much work remains to be
done, but synthetic cancer risk maps appear to be a useful tool for qu
antifying geographic pattern and multivariate structure in cancer mort
ality.