S860 Inverse Association Between Measures of Access to Endoscopy and Stage at Diagnosis of Esophageal Adenocarcinoma in Ohio

Introduction: Esophageal adenocarcinoma (EAC) rates in the United States have increased over the past 6 decades. Understanding the association of the regional distribution of gastroenterologists and EAC diagnoses could inform public health efforts to address the growing incidence of EAC. This study examined the relationship between geographic proximity to medical providers in Ohio capable of performing upper endoscopy, stage at diagnosis and survival in patients diagnosed with EAC. Methods: Data from the Ohio Cancer Incidence Surveillance System (OCISS) was used to examine geographical distribution and stage at diagnosis of newly diagnosed EAC. Spatial access scores to gastroenterology and general surgery practices were created for U.S. census tracts using Ohio public listings, American College of Gastroenterology provider file, and the American Medical Association’s Physician Masterfile. Relationship between spatial access scores and stage at diagnosis was explored through univariate and multivariate logistic regression adjusted for age, sex, race, ethnicity, smoking status, and measures of social deprivation. Results: There were 1,048,575 cases of cancers of all sites in OCISS registry identified. Esophageal adenocarcinoma was diagnosed in 7,467 individuals. Mean age at diagnosis was 67.1± standard deviation (SD) 11.6 years. 6,294/7,467 (84.3%) of cancers were diagnosed in men. Majority of the patients with diagnosis were White (7,202/7,467; [96.5%]), only 207/7,467 (2.8%) of patients in the registry were Blacks with EAC. Fifty-two point one percent of persons diagnosed with EAC resided in areas with poverty level exceeding 10%. Thirty-two percent of cases presented as distant disease. Residing in an area with best spatial access to care (i.e., highest quartile of spatial access ratio), decreased the odds of being diagnosed with advanced EAC stage in multivariate analysis (odds ratio [OR] 0.73; 95% confidence interval [CI] 0.53-0.99). Odds of dying from EAC were increased for patients residing in census tracts with highest poverty levels (OR 1.83; 95% 1.13-2.95). Residing in a non-rural area also showed inverse association with disease stage at diagnosis (OR = 0.85; 95% CI 0.67-1.08). Conclusion: We identified an inverse association between geographic proximity to gastroenterology providers and advanced stage of diagnosis of EAC, suggesting a relationship of access to care by endoscopists and clinical outcomes. Future studies could expand this analysis to national cancer registry data and include primary care providers in the analysis.