Abstract
BACKGROUND: Racial disparities in survival from breast and prostate cancer are well established; however, the roles of societal/socioeconomic factors and innate/genetic factors in explaining the disparities remain unclear. One approach for evaluating the relative importance of societal and innate factors is to quantify how the magnitude of racial disparities changes according to the geographic scales at which data are aggregated. Disappearance of racial disparities for some levels of aggregation would suggest that modifiable factors not inherent at the individual level are responsible for the disparities. METHODS: The Michigan Cancer Surveillance Program compiled a dataset from 1985 to 2002 that included 124,218 breast cancer cases and 120,615 prostate cancer cases with 5-year survival rates of 78% and 75%, respectively. Absolute and relative differences in survival rates for whites and blacks were quantified across different geographic scales using statistics that adjusted for population size to account for the small numbers problem common with minority populations. RESULTS: Whites experienced significantly higher survival rates for prostate and breast cancer compared with blacks throughout much of southern Michigan in analyses conducted using federal House legislative districts; however, in smaller geographic units (state House legislative districts and community-defined neighborhoods), disparities diminished and virtually disappeared. CONCLUSIONS: The current results suggest that modifiable societal factors are responsible for apparent racial disparities in breast and prostate cancer survival observed at larger geographic scales. This research presents a novel strategy for taking advantage of inconsistencies across geographic scales to evaluate the relative importance of innate and societal-level factors in explaining racial disparities in cancer survival.