Association Between Radiation Oncologist Density and Medicare Spending

Abstract

The accessibility of radiation oncology services impacts the likelihood of radiation therapy receipt during cancer care. The density of radiation oncologists (ROs) varies nationally, impacting the geographic access to radiation therapy. We aimed to assess the relationship between RO density and Medicare spending to better evaluate this impact of this relationship.We used the Physician and Other Supplier Public Use File from the Centers for Medicare & Medicaid Services to ascertain the total Medicare reimbursements to each provider marked as a "Radiation Oncologist" for the year 2017. We then placed each radiation oncologist within a Hospital Referral Region (HRR), designed by the Dartmouth Atlas to represent regional healthcare markets for tertiary medical care, using a publicly available zip code-to-HRR crosswalk from the Dartmouth Atlas. We summed the total Medicare reimbursements to ROs in each HRR and compared this to the total Medicare reimbursements (part A+B) in each HRR for the year 2017, as reported by the Dartmouth Atlas group. After winsorization of the top and bottom 2% of HRRs by RO density, linear regression was used to evaluate the relationship between RO density and percentage of Medicare expenditure on RO. Modeling without a constant was used as in an HRR without an RO, there can be no RO services billed.We identified 4,533 ROs in the United States that received a total of $1.38 billion in Medicare reimbursements in 2017. This represented 0.5% of total Medicare reimbursements in 2017. There were 17.0 ROs per 100,000 Medicare enrollees in the US in 2017. On an HRR level, there was an average of 15.4 ROs per 100,000 Medicare enrollees (range 0 - 139.4, standard deviation 10.6). On linear regression, there was a positive relationship between the percentage of Medicare spending on RO and the number of ROs per 100,000 Medicare enrollees: for every additional RO per 100,000 Medicare enrollees, an additional 0.027% (95% CI = 0.025% - 0.030%; P < 0.001) of Medicare expenditures in a given HRR were spent on RO.Using density of RO by HRR as a proxy for access to care, this analysis suggests that with increasing access to care, an increasing proportion of Medicare dollars are spent on RO services at the HRR level.