Purpose: Women with breast cancer (BC) are increasingly diagnosed and treated within the VHA. Breast cancer requires specialized care in tertiary settings such as VAMCs, typically located in urban settings, placing BC patients in rural areas at a disadvantage. Assigning rural-urban status is complicated by the presence of multiple classification plans. In this report, we compare rural-urban status of BC patients in the VHA and its association with distance to nearest VAMC, using 2 plans: USDA Economic Research Service (ERS) Rural Urban Continuum (RUC) and University of Washington’s Rural Urban Commuting Areas 2.0 (RUCA).
Methods: Between 2000 and 2012, 3,622 women were diagnosed with and/or treated for BC within the VHA and recorded in the VA Central Cancer Registry (VA CCR). The patient’s zip code of residence at the time of diagnosis and rural-urban status according to USDA ERS RUC were obtained from the VA CCR. Rural urban commuting status was aggregated into 3 categories: metropolitan, large nonmetropolitan, and rural. Using zip code of residence, rural-urban status of all but 63 women was determined using the University of Washington’s (RUCA) plan and aggregated into 3 categories: urban (metropolitan), large rural or micropolitan, and small rural/isolated small rural. The VHA is organized into 21 regional administrative service networks, or VISNs. The geographic distribution of BC in VHA was determined using the RUC and RUCA scheme, then reported by VISN Census Bureau geographic region: Northeast, Midwest, South, and West. The two plans were compared, using Cohen’s Kappa statistic. The distance between zip code of residence and the nearest within-VISN VAMC was obtained from the VA Planning Systems Support Group database. The association between rural-urban status according to RUC and RUCA and the distance to the nearest VAMC was determined using analysis of variance (ANOVA).
Results: Rural-urban status according to RUC and RUCA were strongly associated (Cohen’s Kappa 0.74, P < .001). About 80% of women with BC in VHA resided in metropolitan areas; the remaining women were split evenly between large nonmetropolitan/micropolitan and rural/small, isolated rural. The Midwest had the highest percentages of both large rural (14%) and small/isolated rural patients (17%), whereas patients in the Northeast had the smallest percentages of large rural (8%) and small/isolated rural patients (7%). Patients living in the Northeast had the shortest travel distances to the nearest within-VISN VAMC, whereas patients in the West had the longest distances. In the Northeast, the average distance to nearest VAMC increased from 11 miles for patients living in metropolitan areas, to 44 miles in small/isolated rural areas. In the West, patients living in metropolitan areas were on average, 37 miles from nearest VAMC. This increased to 124 miles for patients in small/isolated rural areas in the West. Both classifications were significantly associated with increased distance to nearest VAMC (P < .001). On multivariate analysis, rural residence remained significantly associated with increased distance to nearest VAMC (P = .01) even after adjusting for RUCA.
Conclusions: Women with BC living in rural areas must travel longer distances to their VHA facility to receive specialized cancer care. Various plans define rural-urban status, using different methodologies. The rural-urban status of women with BC in VHA was similar using either RUC or RUCA. Rural residence defined by RUC was significantly associated with longer distances to VAMC even after adjusting for RUCA. This suggests that the 2 methodologies are not identical but are highly related when being compared with distance from tertiary care. The choice of rural classification methodology should be considered carefully when researching rural status and cancer outcomes.
