Abstract
OBJECTIVES: The incidence of related sentinel events--breast cancer mortality and neighborhood-specific morbidity for advanced stage at diagnosis--were calculated for women likely to use a community health center in Denver, Colo. METHODS: For the center's service area, neighborhoods (n = 37) were defined by program use. Mortality rates and proportional hazards regression models were estimated for 4189 breast cancer cases recorded between 1979 and 1990. Neighborhood-specific standard morbidity ratios of advanced-stage tumors were based on age-specific rates applied to the entire community. RESULTS: Service area residents were more likely to present with advanced tumors (odds ratio [OR] = 1.4; 95% [CI] = 1.2, 1.5). After adjustment, advanced-stage disease and socioeconomic-demographic status, but not race-ethnicity, contributed significantly to survival. Two neighborhoods (6.5% of the population at risk) with standard morbidity ratios of 2.1 (95% CI = 1.3, 3.4) and 1.7 (95% CI = 1.2, 2.5) accounted for 42% of the excess cases of advanced-stage tumors between 1986 and 1990. CONCLUSIONS: Neighborhood variation in advanced-stage cancer can serve as the basis for efforts to improve access to breast cancer screening.