Travel And Treatment Among Breast Cancer Patients: A Population-Based Analysis

Abstract

Treatment choices for patients with breast cancer require balancing a variety of considerations, but travel distance may create barriers to accessing specific treatments and impact patient choices and outcomes. This dissertation uses novel methods to explore travel burden and evaluate the relationship between travel distance and: 1) surgical choice (mastectomy or breast-conserving surgery (BCS), 2) optimal receipt of radiation, and 3) breast reconstruction among a population-based national sample of newly diagnosed breast cancer patients. We use data from the Surveillance, Epidemiology and End Results (SEER)-Medicare linked database. Our study cohort consists of breast cancer patients newly diagnosed between 2004 and 2013. We use Google Maps and ArcGIS to calculate driving distances and driving times to the nearest available treatment facility. We find that patients living in rural areas traveled on average nearly three times as far for radiation treatment as those from urban areas (40.8 miles vs. 15.4 miles), and their nearest facility was more than four times farther away (21.9 miles vs. 4.8 miles). Disease severity (stage, grade, etc.) was not significantly associated with actual or minimum travel distance. We also find that women living farther from radiation facilities (>50 miles vs. <10 miles) were more likely to undergo mastectomy vs. BCS (OR: 1.48, 95% CI: 1.22, 1.79). Among those who underwent BCS, women living farther from radiation facilities were less likely to complete any RT (OR: 1.72, 95% CI: 1.32, 2.23) or the full recommended course of RT (OR: 1.79, 95% CI: 1.24, 2.60), and were thus more likely to receive guideline-discordant treatment. Women receiving guideline-discordant treatment had worse overall (HR: 1.50, 95% CI: 1.42, 1.57) and breast-cancer specific survival (HR: 1.44, 95% CI: 1.29, 1.60) compared to women receiving guideline-concordant treatment. Finally, we find that increased distance to the nearest reconstruction provider was associated with decreased odds of reconstruction (p<0.001) among mastectomy patients and increased odds of delayed vs. immediate reconstruction among reconstruction patients (p=0.0003). Women living >50 miles away from a reconstruction provider had 51% lower odds of immediate reconstruction (OR: 0.49, 95% CI: 0.39, 0.62) and 93% higher odds of delayed reconstruction (OR: 1.93, 95% CI: 1.03, 3.60) compared to women living within 10 miles of a reconstruction provider. Method of reconstruction was also highly correlated with geography. Increases in reconstruction rates were greater among urban patients, widening the disparity in reconstruction rates among urban vs. rural patients. Taken together, these results suggest that travel burden may be contributing to patients making suboptimal treatment decisions, which in turn contribute to suboptimal survival and patient-centered (quality of life) outcomes. Clinicians, policymakers, and patient advocates should explore social support models, such as lodging and transportation support, and service delivery models, such as shorter treatment regimens, aimed at reducing travel burden and improving guideline-concordant treatment among this patient population.