Presentation (CTI)

OBJECTIVES: Early diagnosis of breast cancer is associated with improved outcomes. Black women are generally more likely than White women to be diagnosed at a later stage. Some studies have found that racial disparities in diagnostic timing persist despite adjustment for SDoH and other factors. We aim to explore the associations between race, SDoH, and stage at diagnosis using EHR data.
METHODS: TriNetX Dataworks-USA is a federated research network of de-identified EHR data from 69 U.S. healthcare organizations (HCOs), a subset of which allows linkage of their data to other sources in a secure way. This study utilizes TriNetX Dataworks-USA data (9 HCOs) linked with third-party SDoH information to explore the relationships between race, SDoH, and stage at diagnosis (early stage: 0-2; late: 3-4). Female Black and White patients diagnosed with breast cancer between 2000-2024 were included. Descriptive statistics (e.g., standardized mean differences) were examined. Additional analyses are planned.
RESULTS: Differences in SDoH were observed between Black and White patients with breast cancer. Relative to White women, a higher proportion of Black women were in the lower household income category (<$50,000 annually; SMD 0.614). Black women had higher mean BMI at diagnosis than White women (32.4 vs 29.0, respectively; SMD -0.509), and there were differences in highest level of educational attainment (SMD 0.32).Among those with a lower household income, a higher proportion of black women had a late stage at diagnosis (18.5%) than white women (12.0%); however, the difference was reduced (12.0% vs. 10.4%) among the higher income group ($50,000+). Additionally, a higher proportion of black women had late stage diagnosis among both educational attainment strata.
CONCLUSIONS: EHR data linked with supplementary data can be leveraged to elucidate racial disparities in timely cancer diagnosis. Additional analyses exploring the extent to which disparities persist when accounting for SDoH are pending.