Association of contextual factors and breast cancer screening: finding new targets to promote early detection.

BACKGROUND: Disparities in breast cancer screening (BCS) exist within the United States. Although such factors as a woman's income and insurance status explain some differences, additional contributions related to local healthcare system characteristics or the social and economic context in which women live have not been fully explored.
METHODS: Using data from a cross-sectional survey of Ohio residents, we assessed BCS in a state-representative sample of 2231 women between the ages of 50 and 69 years. Urbanization, the proportion of female-headed households (FHH), managed care activity, the number of primary care physicians (PCPs) per capita, and county designation as being medically underserved represented some of the contextual characteristics we examined. Using nested hierarchical logistic regression models, we evaluated the association of these characteristics with BCS before and after adjusting for respondents' characteristics.
RESULTS: The proportion of age-eligible women screened for breast cancer was 61.9% (n = 1383); county screening rates varied from 12.9% to 100% (mean 60.3%). Failure to complete high school, lower family income, and absence of continuous insurance, a usual source of care, or current employment were associated with lower BCS. After accounting for these characteristics, per capita PCPs (adjusted odds ratio [AOR] 1.05 (1.01, 1.10), p = 0.02) and the proportion of FHH (AOR 0.66 (0.44, 0.99), p = 0.045) remained independently associated with BCS.
CONCLUSIONS: Contextual characteristics independently associated with BCS identify areas in which women are at increased risk for delayed breast cancer diagnosis. The approach described here can inform the planning phase of regional, state, or federal initiatives to enhance BCS and reduce subsequent disparities in treatment outcomes.