Rethinking rurality: using hospital referral regions to investigate rural-urban health outcomes

Abstract

Abstract Background Rural residents in the United States face disproportionately poorer health outcomes compared to urban residents. This study aims to establish a continuous rural-urban measure for the 306 hospital referral regions (HRRs) in the U.S. and to investigate the relationship between the proportion of rural population served in each HRR and health outcomes, healthcare spending and utilization, and access to and quality of primary care. Methods Cross-sectional analysis using data from The Dartmouth Atlas and the U.S. Census. The sample is limited to fee-for-service Medicare beneficiaries aged 65–99 years and living during 2015. The primary outcomes were measured at the HRR-level: mortality rates, Medicare reimbursements, percent Medicare enrollees who have at least one visit to a primary care physician, diabetic hemoglobin A1c testing rates, and mammography rates. We calculate a population-weighted rural proportion and population-weighted area deprivation index (ADI) for each HRR by aggregating zip-code level data. Results The most rural quartile of HRRs had significantly greater mean mortality rate of 4.50%, compared to 3.95% in most urban quartile of HRRs (p < 0.001). Increasing rural proportion was associated with decreasing price-adjusted Medicare reimbursements. In the multivariate, linear regression model, increasing area deprivation (ADI) was associated with increasing rates of mortality and greater utilization. Conclusion Disparities in rural mortality are driven by sociodemographic disadvantage, rather than the quality of care provided at hospitals serving rural areas. After accounting for sociodemographic disadvantage, rural areas achieve similar quality of primary care in measured domains at an overall lower cost.