From food deserts to clinical trial deserts: Challenges in access to breast cancer trials.

Abstract

1525 Background: Patients who live in food deserts have high mortality rates from breast cancer and stand to benefit from participating in research studies, yet they may face complex barriers to doing so. This study explores the relationship between living in a food desert and breast cancer clinical trial enrollment and assesses the contribution of transportation barriers and distance. Methods: The national Vizient Clinical Database (which includes 98% of academic and 1,000 community hospitals) was queried for women treated for breast cancer between January 2022 and June 2024. Patients who traveled > 4 hours to get to treatment were excluded (“destination care”). The outcome of interest was participation in a clinical trial. A “clinical trial desert” was defined as > 2 hour driving time to the nearest hospital enrolling patients in clinical trials. Multivariable analysis evaluated the association between living in a level 1 (most severe) USDA food desert census tract, a clinical trial desert, a census tract with high transportation vulnerability (households with limited access to a car or public transit), and breast cancer clinical trial participation. Interaction analysis was performed between food desert and clinical trial desert status. Results: Of 1,317,269 patients, 103,790 (7.9%) lived in a food desert, and 22,779 (1.7%) participated in a clinical trial. Patients living in a food desert comprised 6.2% of patients enrolled in a trial vs 7.9% of patients not enrolled, p < .0001. 41.0% of patients living in a food desert also resided in a clinical trial desert. On multivariable analysis, living in a food desert was associated with decreased odds of trial participation (aOR 0.87, 95% CI 0.82-0.92, p < .0001), as was living in a clinical trial desert (aOR 0.89, 95% CI 0.84-0.94, p < .0001), and living in the most vulnerable quartile for neighborhood transportation (aOR 0.89, 95% CI 0.85-0.93, p < .0001; ref: least vulnerable). Medicaid insurance also decreased the odds of enrollment (aOR 0.84, 95% CI 0.80-0.89, p < .0001; ref: private). Conversely, receiving care at an academic hospital increased the odds of enrollment (aOR 2.98, 95% CI 2.62-3.15, p < .0001; ref community hospital). Living in both a food desert and a clinical trial desert decreased the odds of clinical trial participation by 27% (aOR 0.73, 95% CI 0.70-0.76, p < .0001) compared to 19% if living in a food desert alone (aOR 0.81, 95% CI 0.78-0.84, p < .0001); p < .0001 for interaction. Conclusions: Neighborhood transportation barriers, clinical trial deserts, and food deserts all independently confer a similar lower likelihood of participation in a clinical trial. Living in a clinical trial desert compounds the negative impact of living in a food desert alone, further taxing already disadvantaged populations. Interventions such as patient navigation, food banks, and opening clinical trials near communities experiencing food insecurity may mitigate these challenges.