Background: In the U.S. there is a wide geographic and racial variation in breast cancer mortality, but there exist counties with large African American populations where breast cancer mortality rates among the elderly are low and do not differ for white and black women. The objective of this study is to better understand how successful communities compare with less successful ones.

Methods: We obtained county level age-adjusted breast cancer mortality rates (BCMR) in black and white women 65+ in 1999–2005 from the Compressed Mortality File of the National Center for Health Statistics. We selected 203 counties with ≥ 14 deaths in black women and classified them in 4 types: Failing, i.e., with black and white BCMR greater than the average white BCMR; With Disparity, i.e., with black BCMR greater and white BCMR lower than the average white BCMR; With Reverse Disparity, i.e., with black BCMR lower and white BCMR greater than the average white BCMR; and Successful, i.e., with black and white BCMR lower than the white average BCMR. Using data from the Area Resource File, we identified economic factors and medical resource availability that may be associated with type of county. We used a multinomial logistic regression to identify independent factors associated with county type.

Results: Of 203 counties, the average BCMR was 137.5 per 100,000 (range: 55.1–296.8) for black, and 117.2 (range: 70.6– 230.5) for white women. Counties were: 79 (38.9%) Failing, 72 (34.5%) With Disparity, 21 (10.3%) With Reverse Disparities, and 31 (15.3%) Successful. Of counties of all types, Reverse Disparities counties had the highest average median household income ($48,371), and lowest proportion of people living below poverty (11.6%), of blacks with < 9 years of school (6.3%), of persons working in the county (68.6%). They also had the highest number of medical doctors (MDs) per 1000 females (8.2), proportion of hospitals associated with a medical school (50%), and the lowest change in MDs between 1999 and 2004 (7.4%). Successful counties had the lowest average median income ($39,222), the highest proportion of people living in poverty (14.2%); of blacks 25 and older with < 9 years of school (9.7%); of persons working in the county (82.2%); and had the lowest number of MDs (6.0). Counties With Disparities had the lowest number of MDs (6.0) but the highest increase in MDs (27.2%); the lowest proportion of hospitals associated with a medical school (28%), highest proportion of uninsured (15.5%), and the highest increase in people living under poverty from 1999 to 2002 (5%). Other variables not significantly different across county type were: number of primary care MDs and radiologists per 1000 women, HMO penetration, unemployment rate, and proportion of owner occupied homes, of blacks living below the poverty line, of black women and of women 65 and older. In multivariable analysis the proportion of people uninsured, working in the county, living below the poverty line, and of blacks 25 and older with < 9 years of school, were found to be significant independent predictors of county type.

Conclusions: Counties with different levels of breast cancer mortality and disparities among elderly women differ in socioeconomic status, availability and type of medical resources. Further understanding of county differentiation may address both sources and solutions pertaining to disparities in elderly breast cancer mortality.

Second AACR International Conference on the Science of Cancer Health Disparities— Feb 3–6, 2009; Carefree, AZ