LO70: Emergency department use and migration patterns of people experiencing homelessness

Abstract

Introduction: Understanding how homeless patients interact with healthcare systems can be challenging. The nature of the population is such that identifying and following these persons can be severely limited by data. Previous studies have used survey data which relies on self-reporting and selected samples such as those persons admitted to homeless shelters (Gray et al. 2011). Other studies have been able to leverage administrative data but only for selected local geographic areas (Somers et al. 2016, Tompkins et al 2003). It is possible that the current literature has not examined a large proportion of homeless persons and their healthcare use. This is concerning because this population can have higher associated medical costs and greater medical resource utilization especially with regards to psychiatric and emergency department (ED) resources (Tulloch et al. 2012, Forchuk et al, 2015). Methods: Administrative health data (2010 to 2017) is used to analyze ambulatory care records for homeless individuals in Ontario, Canada. Uniquely, we are able to use ED contacts as a way of identifying homeless migrations from region to region within Ontario. Using a network analysis we identify high impact ED nodes and discrete hospital networks where homeless patients congregate. We are also able to more fully characterize this population's demographics, health issues, and disposition from the ED. Results: We provide a more complete understanding of migration patterns for homeless individuals, across Ontario and their concomitant ED use and hospitalizations. The three most frequented regions in Ontario (n = 640,897) were Toronto Central (35.96%), Hamilton Niagara Halimand Brant (8.9%) and Champlain (7.84%). In subsequent visits, the majority of patients presented to different EDs, however a subgroup who always presented to the same site was present. Over the 7 year period, migration between visits occurred most often between urban areas, and increased as a whole. Conclusion: The results of the study allow for the enhancement care coordination for vulnerable populations and enhance the availability and delivery of services for sub-groups of homelessness whose care needs may differ based on migration patterns. Services can be coordinated between jurisdictions for homeless individuals, and appropriate referrals can be made across the health care system. Further evidence is provided for a novel method of mapping migration among the homeless and its associations and effects on ED use.