Improving our understanding of breast cancer mortality disparities through recurrence: a multi-level approach among women in Georgia

PROJECT SUMMARY/ABSTRACT Significance. Although national and philanthropic efforts have sought to reduce and eliminate breast cancer (BC) mortality disparities over the past few decades, they have not only persisted—but widened. Additionally, due to incomplete capture of recurrence data, no previous investigation has identified drivers of disparities in BC

recurrence following a diagnosis of early-stage (I–IIIA) disease. In Georgia, where economic and racial/ethnic disparities are among the greatest in the United States, the sources of BC outcome disparities are unresolved, and likely arise from the interplay of causal and contributing factors at multiple levels—from cell to society.

Approximately 40% of all BC survivors will suffer a recurrence during their lifetime, and clinical data suggest a higher risk of recurrence in minority and low-income women. Given the high lifetime risk of recurrence, posited race/ethnic disparities in recurrent BC, and documented mortality disparities across demographic domains, now

is the pivotal time to characterize underlying pathways contributing to inequities in BC prognosis. Innovation. Our proposal is innovative in that it will be the first to estimate risks and rates of BC recurrence by demographic characteristics, consider intersectionality in BC outcome disparities, and use a multilevel decomposition

approach to identify potential targets for intervention. Approach. Integrating multiple data streams (e.g., discharge, administrative claims, hospital, and census data) with cancer registry data from a large, diverse population, we will identify proximal, intermediate, and distal determinants of race/ethnic, SES, and urban/rural

disparities in both recurrence and BC-specific mortality, as well as examine recurrence and its treatments as mediators of disparities in mortality rates by race, SES, and urban/rural characteristics. Data will be from approximately 30,000 women diagnosed with a first primary stage I–IIIA BC in Georgia (2013–2017) and followed

for up to 12-years. Impact. Previous research in this area has had consistent shortcomings including (1) insufficient ascertainment of recurrence at the population level; (2) examining one or few factors without accounting for shared contributions across multiple levels; and (3) inadequate power to explore intersections of

identity. Our study, for the first time, will examine multi-level contributors to race/ethnic, SES, and urban/rural disparities in both BC recurrence and mortality among women with early-stage disease. Our innovative multi- level decomposition approach will move us beyond merely documenting disparities, to identifying modifiable

targets within the social contexts of affected communities, facilitating prioritization of interventions.