Racial differences in breast cancer survival in the Detroit Metropolitan area.

African American (AA) women have poorer breast cancer survival compared to Caucasian American (CA) women. The purpose of this analysis was to determine whether socioeconomic status (SES) and treatment differences influence racial differences in breast cancer survival. The study population included 9,321 women (82% CA, 18% AA) diagnosed with local (63%) or regional (37%) stage disease between 1988 and 1992, identified through the Metropolitan Detroit SEER registry. Data on SES were obtained through linkage with the 1990 Census of Population and Housing Summary Tape and cases were geocoded to census block groups. Pathology, treatment and survival data were obtained through SEER. Cox proportional hazards models were used to compare survival for AA versus CA women after adjusting for age, SES, tumor size, number of involved lymph nodes, and treatment. AA women were more likely to live in a geographic area classified as working poor than were CA women (p<0.001). AA women were less likely to have lumpectomy and radiation and more likely to have mastectomy with radiation (p<0.001). After multivariable adjusted analysis, there were no significant racial differences in survival among women with local stage disease, although AA women with regional stage disease had persistent but attenuated poorer survival compared to CA women. After adjusting for known clinical and SES predictors of survival, AA and CA women who are diagnosed with local disease demonstrate similar overall and breast cancer-specific survival, while race continues to have an independent effect among women presenting at a later stage of disease.

Race, socioeconomic status and stage at diagnosis for five common malignancies.

BACKGROUND: African-Americans are more likely than Caucasians to be diagnosed at an advanced stage of colorectal, lung, breast, cervical, and prostate cancers. This study explores if racial differences in stage at diagnosis can be explained by socioeconomic status (SES) differences. Previous studies investigating this association have used aggregate SES indicators from census tract of residence; we used census block-group data, representing a smaller, potentially more homogenous group. METHODS: We included all African-American and Caucasian invasive cancers of the colon and rectum, lung and bronchus, female breast, cervix uteri, and prostate that were diagnosed between January 1, 1988 and December 31, 1992 in the Detroit area. Stage of disease at diagnosis was grouped as local or non-local. An SES value was calculated for each case using aggregate 1990 US Census data for education, poverty status, and occupation specific to each case's census block-group. Logistic regression analysis was used to model the probability of non-local stage using SES, race, age group, and sex as covariates. RESULTS: SES was an independent predictor of stage at diagnosis for each cancer site, with cases from the highest SES block-group more likely to present with local stage disease than those from the lowest SES group. Race independently predicted stage only for breast and prostate cancers; African-Americans presented with more advanced stage than Caucasians. CONCLUSIONS: Based on census block-group aggregate data, SES is an important predictor of stage at diagnosis, most likely accounting for much of the disparity in stage between African-Americans and Caucasians for colorectal, lung, and cervical cancers. Biological factors may play a role in racial disparities for breast and prostate cancer stage at diagnosis.