Geographic Distance to Treatment Is Not Associated with Overall Survival

Presentation: AHNS036
Topic: Survivorship
Type: Oral
Date: Wednesday, April 18, 2018
Session: 3:40 PM - 4:40 PM Survivorship
Authors: Maheer Masood1, Douglas Farquhar1, Angela Mazul1, Philip McDaniel1, Trevor Hackman1, Jose Zevallos2, Andrew Olshan1
Institution(s): 1University of North Carolina, 2Washington University at St. Louis

Introduction: Patients diagnosed with head and neck squamous cell carcinoma (HNSCC) often choose between receiving treatment in their community or traveling to a larger medical center. Outcomes between treatment at different centers has been controversial. Our aim in this study was to examine the impact of geographic distance traveled by a patient to reach their provider on overall survival in HNSCC. We examined survival by the distance to both diagnosing and treating providers, with adjustment for demographics, socioeconomic status, tobacco and alcohol use, and stage at presentation.

Methods: Data for analysis was obtained from the Carolina Head and Neck Cancer Epidemiology Study (CHANCE); a population-based case-control study in 46 counties in North Carolina (NC). Cases were identified from 2001 to 2006 through rapid case ascertainment with the North Carolina Central Cancer Registry. An in-person interview obtained a complete residence history and other factors, and medical records were used to ascertain stage at diagnosis and the locations for cancer diagnosis and treatment. Linear distances between the patient’s home address and biopsy and treatment site(s) were calculated in ArcMap 10.5 (ESRI, 2017). Distances were divided into quartiles for analysis. Multivariable Cox proportional hazard regression was used to calculate hazard ratios for overall survival, before and after adjustment for age, sex, race, income, insurance status, tobacco use, alcohol use, T-stage, and N-stage at presentation. Separate models were used to examine distance to diagnosing provider, distance to surgeon (if tumor was surgically resected), distance to radiation oncologist, and distance to medical oncologist. Patients with distant metastases at presentation were excluded.

Results: A total of 935 patients were included in this analysis. The median distance between patients and diagnosing providers was 10 mi (range < 1 to 213). The median distance between patients and treating providers was 108 mi for surgery (range 3-300), 102 mi for radiation (range 3-291), and 104 mi for chemotherapy (range 3-269). There were no significant associations between the distance to any provider and survival, in both adjusted and unadjusted models. For example, the HR for patients in the quartile farthest from the diagnosing provider was 1.1 relative to patients in the closest quartile (95% CI 0.9 – 1.5). The only variables in the adjustment set that associated with survival were insurance status (relative to private insurance, HR of 2.29, 95% CI 1.6-3.2 for Medicare/Medicaid, and HR of 2.30, 95% CI of 1.6-2.4 for no insurance), T-stage (HR 1.7 95% CI 1.2 - 2.4 for T4 vs. T1), N stage (HR 1.5, 95% CI 1.2 - 1.8 for N1 vs. N0), and annual income (HR 1.5, 95% CI 1.1 - 2.1 for income < $20,000 relative to > $50,000).

Conclusion: The geographic distances between patients and diagnosing or treating providers are not associated with overall survival.