Abstract 426: The Rural Experience of Cardiac Arrest in Maine: Geography, EMS Access, and Social Determinants of Health

Abstract

Listen

Translate article

Introduction: Resuscitation resources and outcomes differ between rural and urban areas, largely due to patient risk factors and limited emergency medical service (EMS) resources. The current definitions of rural areas are inconsistent and do not encompass all the factors that may affect post-resuscitation care, such as healthcare access, and social determinants of health (SDOH). Hypothesis: In addition to rurality category, measures of healthcare access and SDOH further explain variability in cardiac arrest outcomes across the state of Maine. Goals: Evaluate the difference in model performance in determining survival at three months, using EMS-data, measures of geographic rurality, healthcare access, and SDOH. Approach: We utilized Maine EMS & Fire Information Systems to identify non-traumatic out-of-hospital cardiac arrests in Maine from 2018-2021. We merged this data with state Center for Disease Control data to determine 30-day survival. Risk adjusted models were developed and measures of healthcare access and SDOH were added for all patients and for a cohort of patients who were transported to an Emergency Department (ED). We measured model performance and compared models using likelihood ratio tests. Results: Inclusion criteria was met in 4,277 patients and 1,167 (27%) patients were transported to an ED. After risk adjustment, the best performing model included Rural Urban Commuting Area (p=0.083), time from arrest to highest level of EMS provider (p<0.001) and Area Deprivation Index (p=0.01). For patients transported to EDs, the best performing model included population per square mile (p=0.001), time from arrest to highest level of EMS provider (p=0.041) and census tract with high school education or greater (p<0.001). In both cohorts, the model with all three factors and other risk adjustment was superior to the baseline and models with any individual factor (likelihood ratio test p<0.05). Conclusions: The addition of rurality measures to cardiac arrest characteristics improves model association with good outcome and the addition of markers of EMS response and SDOH explained additional variability. These factors should be considered in combination when evaluating outcome disparities among patients with cardiac arrest.