Date of Award

January 2013

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

Department

Medicine

First Advisor

James B. Yu

Subject Area(s)

Medicine, Public health, Oncology

Abstract

Previous analyses of the radiation oncology (RO) workforce have focused on gross numbers and not geographic distribution. We investigated trends in the geographic distribution of the radiation oncology workforce across the United States. Additionally, we assessed the impact of geographic variations in the RO workforce on prostate cancer management and outcomes. We hypothesized that geographic variations in the workforce would be associated with prostate cancer management and prostate-cancer mortality.

We used the Area Resource File to calculate and map the ratio of radiation oncologists to the population aged 65 or older (ROR) within different health service areas (HSA) across the United States from 1995-2007. Multivariate regression models were built to test the association between ROR and socioeconomic variables (income, minority population, unemployment rate, population education). Using patient data from the Surveillance Epidemiology End Results Program (SEER) we built multivariate logistic regression models to test associations between variations in the RO workforce and patient decisions to observe, undergo a radical prostatectomy, or undergo radiation therapy. Using mortality data from the State Cancer Profiles dataset, we built multivariate linear regression to test the association between RO workforce and count-level age-adjusted prostate cancer mortality.

Despite a 24% increase in the workforce from 1995 to 2007, there remained consistent geographic maldistribution of radiation oncologists, specifically affecting the rural HSAs. Regression analysis found higher ROR associated with more educated (p=.001), affluent (p<.001) HSAs with lower unemployment rates (p<.001), and higher minority populations (p=.022). Of the 108,612 prostate cancer patients queried from the SEER dataset, patients with low-risk disease (p<.001) residing in HSAs with fewer radiation oncologists (p=.001-.041), fewer urologists (p<.001), and more primary care physicians (p<.001) were most likely to observed in lieu of curative treatment. Of the 91,643 patients who underwent some form of curative treatment, older, single (p<.001), African American patients (p<.001) with low-risk disease (p<.001) residing in HSAs with more radiation oncologists (p=.007-.001) and primary care physicians (p<.001) were more likely to receive radiation therapy. The presence of at least one radiation oncologist was associated with between 5.74% and 1.48% reduction in prostate cancer mortality (p=.001-.045) even when adjusting for county-level prostate cancer incidence.

Despite a modest growth in the radiation oncology workforce, there exists persistent geographic maldistribution of radiation oncologists allocated along socioeconomic and racial lines. Regional variations in the RO workforce are associated with variations in the management of prostate cancer. The presence of at least one radiation oncologist is associated with a reduction in county-level prostate cancer mortality. There is a need for geographically aware policy in order to optimize the RO workforce and improve prostate cancer outcomes.

Share

COinS