Date of Award

January 2014

Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)



First Advisor

James B. Yu

Subject Area(s)



THE INFLUENCE OF REGIONAL HEALTH SYSTEM CHARACTERISTICS ON THE MANAGEMENT OF GLIOBLASTOMA MULTIFORME. Dhruv Khullar, Sanjay Aneja, James B. Yu. Department of Therapeutic Radiology, Yale University, School of Medicine, New Haven, CT.

Despite a known optimal treatment protocol for the management of glioblastoma multiforme (GBM), many patients fail to receive complete surgical resection or post-operative radiation therapy (PORT). The underlying reasons behind this disparity are unclear. We hypothesize that regional health system resources influence the surgical management and PORT in patients with GBM.

Surgical intervention, PORT receipt, and patient data for patients diagnosed with GBM from 2004 to 2008 were obtained from the NCI Surveillance, Epidemiology, and End Results (SEER) database and combined with the health system data from the Area Resource File. Health system characteristics studied included radiation oncologist density, neurosurgeon density, primary care provider (PCP) density, general radiation therapy (RT) and/or medical oncology (MO) equipped hospital density, and median household income. The geographic units of analysis were NCI-defined Health Service Areas (HSA) within the SEER registry. Four logistic models were constructed to test the effect of health system characteristics on surgical treatment choice and PORT receipt.

Of the 8,337 patients in our sample that were diagnosed with GBM, 71.45% received PORT. We found that younger, married patients in HSAs with higher median incomes were significantly more likely to receive both gross total resection (p < .001, p < .001, p = 0.002) and PORT (p < .001, p < .001, p = .008). For every $10,000 increase in the median income of a HSA, a patient's likelihood of receiving gross resection and PORT increased by 7% and 6.3%, respectively. The density of primary care providers and radiation oncology equipped hospitals were also significant predictors of PORT receipt (p = .024, p = .002). Patient race, radiation oncologist, and neurosurgeon densities were not associated with likelihood to receive PORT

Our findings suggest that regional variations in neuro-oncology services and income may have impact on GBM management. The presence of hospitals with oncology services within an HSA was more predictive of PORT receipt than the density of radiation oncologists and neurosurgeons themselves, suggesting that hospital-level infrastructure is needed to optimize care of GBM, independent of physician staffing levels. Policies aimed at narrowing disparities in treatment may need to focus on addressing regional variations in oncology resources.