Proximity to Radiotherapy Center, Population, Average Income, and Health Insurance Status as Predictors of Cancer Mortality at the County Level in the United States

• Published in: JCO global oncology Vol. 9; no. 9; p. e2300130
• Main Authors: Beckett, Matthew, Goethals, Luc, Kraus, Ryan D., Denysenko, Kseniya, Barone Mussalem Gentiles, Maria Fernanda, Pynda, Yaroslav, Abdel-Wahab, May
• Format: Journal Article
• Language: English
• Published: Wolters Kluwer Health 01.09.2023; American Society of Clinical Oncology
• Subjects: ORIGINAL REPORTS
• ISSN: 2687-8941; 2687-8941
• DOI: 10.1200/GO.23.00130

Toward further understanding of the geographic disparities in oncologic outcomes faced by American patients. PURPOSE Sufficient radiotherapy (RT) capacity is essential to delivery of high-quality cancer care. However, despite sufficient capacity, universal access is not always possible in high-income countries because of factors beyond the commonly used parameter of machines per million people. This study assesses the barriers to RT in a high-income country and how these affect cancer mortality. METHODS This cross-sectional study used US county-level data obtained from Center for Disease Control and Prevention and the International Atomic Energy Agency Directory of Radiotherapy Centres. RT facilities in the United States were mapped using Geographic Information Systems software. Univariate analysis was used to identify whether distance to a RT center or various socioeconomic factors were predictive of all-cancer mortality-to-incidence ratio (MIR). Significant variables ( P ≤ .05) on univariate analysis were included in a step-wise backward elimination method of multiple regression analysis. RESULTS Thirty-one percent of US counties have at least one RT facility and 8.3% have five or more. The median linear distance from a county's centroid to the nearest RT center was 36 km, and the median county all-cancer MIR was 0.37. The amount of RT centers, linear accelerators, and brachytherapy units per 1 million people were associated with all-cancer MIR ( P < .05). Greater distance to RT facilities, lower county population, lower average income per county, and higher proportion of patients without health insurance were associated with increased all-cancer MIR ( R-squared, 0.2113; F, 94.22; P < .001). CONCLUSION This analysis used unique high-quality data sets to identify significant barriers to RT access that correspond to higher cancer mortality at the county level. Geographic access, personal income, and insurance status all contribute to these concerning disparities. Efforts to address these barriers are needed.