EPID-34. GEODEMOGRAPHIC CLASSIFICATION OF GLIOBLASTOMA RISK IN A RURAL HOSPITAL SYSTEM: RELATIONSHIP TO TUMOR INCIDENCE AND OVERALL SURVIVAL

Abstract In 2019, a projected 86,000 persons in the US will be diagnosed with a primary brain tumor, with an estimated prevalence of 700,000. Glioblastoma (GBM) is the deadliest and most common of these tumors in adults with an overall survival ~15 months. Exposure to ionizing radiation is the single known environmental risk factor for GBM. Most research in the field is directed at treatment and tumor biology/classification. There is growing interest in geodemographics, disease prevention, and measuring quality-survival in GBM. We evaluated 622 GBM patients (2006 to 2018) in Northeastern/Northcentral Pennsylvania treated at one of the nation’s largest rural hospital systems. Each patient was geocoded (plotted longitude/latitude) address at time of diagnosis and we used medical records, projected 2015 Census, and Environmental Protection Agency data to develop risk maps. Preliminary evaluation of overall survival (OS) and incidence related to distance from hospital, age, sex, level of education, insurance, income, and income disparity (Gini coefficient measure of income inequality) was completed. Interestingly, all significant clusters of high GBM incidence lay along water-ways or overlapped with high coal mine reprocessing. Contrary to recent studies, there was no significant sex or distance (0mi to 150mi) impact on survival. However, each age-decile increase resulted in an average 9mo decrease in OS (37mo[< 40yo], 26mo [41-50yo], 17mo [51-60yo], 12mo [61-70yo], and 3mo [>71yo]). Counties with the lowest OS shared geodemographic characteristics such as fewer insured and lower incomes (low income inequality), and counties with more college educated persons had longer survival. Counties with higher GBM incidences had better insured percentages, lower incomes (high income inequality), and large polarity between high school uneducated and college educated persons. Rural populations potentially demonstrate unique geodemographic characteristics that might better predict disease outcomes. Studies in this area could influence improvements in public policy and prevention-focused health literacy opportunities.