The Effect of Distance from Cancer Facility on Advanced Clinical Stage at Diagnosis in Cervical Cancer

Abstract

In the United States, cervical cancer remains a significant cause of mortality especially among young women. The effect of distance from cancer facility is not well understood in cervical cancer and has been shown to have a complicated relationship with disease characteristics and outcomes in other cancers. The purpose of this study is to investigate the relationship between distance from cancer facility on clinical stage at diagnosis in patients with cervical cancer. Data were obtained from the National Cancer Database which includes patient demographics, disease characteristics, and first course treatment decisions. Patients diagnosed with cervical cancer between 2004 and 2015 were included. Patients were excluded if they had missing demographic or disease information, variant histology, or lived >1000 miles from their facility resulting in 51,413 patients. Distance was categorized by quartiles. Univariate comparisons were performed using analysis of variance and chi-square. Disease was classified as localized (surgically curable, stage 1a-2a) or advanced (not surgically curable, stage 2b-4b). A multivariable binomial logistic regression model was used to investigate the effect of distance on stage at diagnosis while adjusting for age, insurance status, race and ethnicity, income quartile, education, urbanization, Charlson Comorbidity Score, and year of diagnosis. The mean age was 51.0 years, 16.9% of patients were black, 14.7% were Hispanic, 45.0% of patients had private insurance, 10.7% were uninsured, 15.7% were in the highest education quartile nationally, and 98.4% of patients lived in urban or metro counties. Overall, 50.9% of patients presented with advanced disease. On multivariable analysis, patients living in the 2nd (OR 0.96, p=0.098), 3rd (OR 0.90, p<0.001) and 4th distance quartile (OR 0.66, p<0.001) demonstrated a stepwise reduction in risk of advanced disease by increasing distance compared to those living closest to their facility. Black patients had increased risk of advanced disease (OR 1.18; p<0.001) while Hispanic patients had decreased risk (OR 0.81; p<0.001) compared to white patients; patients with Medicaid (OR 1.98; p<0.001), Medicare (OR 1.14; p<0.001), and who were uninsured (OR 2.31; p<0.001) experienced increased rates of higher stage disease relative to patients with private insurance. To our knowledge this is the first study to analyze the effect of distance on cervical cancer stage at diagnosis in a nationally representative cohort. Greater distance from cancer facility resulted in lower risk of advanced stage disease at diagnosis. Intuitively, greater distance might pose a barrier to care, yet has demonstrated a protective effect among patients in other disease sites. Additional research is needed to understand the nuanced relationship between distance, disease characteristics, and outcomes in cervical cancer.