Hunt Bijou R, Silva Abigail, Lock Derrick, Hurlbert Marc
Sinai Urban Health Institute, Sinai Health System, 1500 S. Fairfield Ave., Room K438, Chicago, IL, 60608, USA.
Department of Public Health Sciences, Loyola University Stritch School of Medicine, 2160 S. First Avenue, Maywood, IL, 60153, USA.
Cancer Causes Control. 2019 Feb;30(2):149-164. doi: 10.1007/s10552-018-1125-x. Epub 2019 Jan 17.
We employed a city-level ecologic analysis to assess predictors of race-specific (black and white) breast cancer mortality rates.
We used data from the National Center for Health Statistics and the US Census Bureau to calculate 2010-2014 race-specific breast cancer mortality rates (BCMR) for 47 of the largest US cities. Data on potential city-level predictors (e.g., socioeconomic factors, health care resources) of race-specific BCMR were obtained from various publicly available datasets. We constructed race-specific multivariable negative binomial regression models to estimate rate ratios (RR) and 95% confidence intervals (CIs).
Predictors of the white BCMR included white/black differences in education (RR 0.95; CI 0.91-0.99), number of religious congregations (RR 0.87; CI 0.77-0.97), and number of Medicare primary care physicians (RR 1.15; CI 1.04-1.28). Predictors of the black rate included white/black differences in household income (RR 1.03; CI 1.01-1.05), number of mammography facilities (RR 1.07; CI 1.03-1.12), and mammogram use (RR 0.93; CI 0.89-0.97).
Our ecologic analysis found that predictors of breast cancer mortality differ for the black and white rate. The results of this analysis could help inform interventions at the local level.
我们采用城市层面的生态分析来评估特定种族（黑人和白人）乳腺癌死亡率的预测因素。
我们使用了美国国家卫生统计中心和美国人口普查局的数据，计算了美国47个最大城市2010 - 2014年特定种族的乳腺癌死亡率（BCMR）。特定种族BCMR潜在的城市层面预测因素（如社会经济因素、医疗资源）的数据来自各种公开可用的数据集。我们构建了特定种族的多变量负二项回归模型来估计率比（RR）和95%置信区间（CI）。
白人BCMR的预测因素包括白人与黑人在教育方面的差异（RR 0.95；CI 0.91 - 0.99）、宗教场所数量（RR 0.87；CI 0.77 - 0.97）以及医疗保险初级保健医生数量（RR 1.15；CI 1.04 - 1.28）。黑人死亡率的预测因素包括白人与黑人在家庭收入方面的差异（RR 1.03；CI 1.01 - 1.05）、乳房X光检查设施数量（RR 1.07；CI 1.03 - 1.12）以及乳房X光检查的使用情况（RR 0.93；CI 0.89 - 0.97）。
我们的生态分析发现，乳腺癌死亡率的预测因素在黑人和白人发病率方面存在差异。该分析结果有助于为地方层面的干预措施提供参考。
