Racial disparities in breast cancer survival after treatment initiation in Brazil: a nationwide cohort study

Breast cancer is a public health priority in Brazil and ensuring equity in health care is one of the cancer control plan goals. Our aim was to present the first assessment on the influence of race or skin colour on breast cancer survival at the national level. In this nationwide cohort study, data on women who initiated treatment for breast cancer in the public health-care system (Sistema Unico de Saúde), Brazil, were assembled through record linkage of administrative and mortality information systems. The administrative information systems were the Outpatient Information System (data from high complexity procedure authorisations) and the Hospital Information System (data from hospitalisation authorisations). We included women aged 19 years or older who started treatment between Jan 1, 2008, and Nov 30, 2010; self-identified as having White, Black, or Brown race or skin colour; had tumour stage I–IV; and were treated with chemotherapy or radiotherapy, or both. Patients were followed up until Dec 31, 2015. Patients with only hormone therapy records or who underwent only surgery were excluded. The Kaplan-Meier method was used to estimate crude overall survival for race or skin colour by time since treatment initiation, and Cox regression to estimate all-cause mortality hazard ratios (HRs) before and after adjustment for other covariates. We identified 59 811 women treated for stage I–IV breast cancer. 37 318 (62·4%) women identified themselves as White, 18 779 (31·4%) as Brown, and 3714 (6·2%) as Black. 5-year overall survival probability was higher for White women (74% [95% CI 73–74]) than Black women (64% [62–65]; p