Abstract

There is a wealth of information available when combining the US census data with state cancer registries. The article by Wan et al in this month's issue of the Southern Medical Journal studies the effect of using different geographic groupings on the study of prostate cancer mortality in African Americans versus whites in Texas. Most of us are unaware of the exact mechanism of census taking or the valuable data made available through this effort. The census data are presented by state, county, ZIP code, blocks (for rural areas), and tracts (for metropolitan areas). The census tracts are designed by local committees to represent groupings of population characteristics, economic status, and living conditions (http://www.census.gov/geo/www/cen_tract.html). Using census tract data, other studies have shown a link between prostate cancer mortality and racial status in Texas (eg, White et al. Cancer 2011;117(5):1080-8). In this study, the authors looked into the relationship of geographic groupings of people in the census with respect to mortality figures for prostate cancer. The intent was to see how health disparities (prostate cancer mortality) varied with the scale of geographic groupings. Differences in mortality by race at the block grouping and census tract level diminished and even disappeared at the county level. The implication for design of future studies is that health disparities researchers must choose the appropriate-size geographic scales when designing comparative studies. The authors point out that the scale may be different for different disease processes as well.