334 Does the Closure of a Homeless Shelter Increase the Prevalence of Homelessness in the Emergency Department?

Abstract

Emergency departments (EDs) are the primary providers of health care for the homeless, and also act as vital social safety-net institutions, providing vital resources (food, water, shelter) necessary for survival and subsistence. For example, one study found that nearly a quarter of homeless adults sought care in an ED when they first became homeless. We therefore hypothesized that the closure of our city’s largest emergency homeless shelter in August 2017 would be associated with a subsequent increase in the prevalence of homelessness among our ED patients. We conducted a cross-sectional before and after survey of a convenience sample of patients presenting to an urban, safety net ED in Atlanta, Georgia. Participants were surveyed over a 3-month period in 2016 and 1 year later following the closure of the homeless shelter. A team of trained research assistants administered a structured survey instrument. Patients were surveyed in all areas of the ED and at all times of the day. Patients were at least 18 years old, English-speaking, not incarcerated, and able to provide informed consent. 1,152 ED patients (55.5% [52.6 - 58.4] male; median age = 44 [IQR: 31 - 55]) completed the survey. The racial makeup of the sample was as follows: 70.8% [68.2 - 73.5] Black, 17.1% [14.9 - 17.6] White, 6.3% [4.9 - 7.7] Hispanic, 1.9% [0.8 - 2.9] Asian, and 3.9% [2.8 - 5.1] self-identified as “other.” Most patients had at least a High school education (40.51% [37.66 - 43.36] high school/GED; 23.75% [21.28 - 26.22] had at least some college, 17.24% [15.04 - 19.44] did not graduate High school and 18.55% [16.29 - 20.81] completed college. 47% [44.1 - 49.9] were unemployed, 15.5% [13.4 - 17.6] were employed part-time, and 37.5% [34.7 - 40.3] were employed full-time. With respect to housing insecurity, 30.04% [27.38 - 32.71] moved in with others to save money, 30.21% [27.54 - 32.88] lived with others but didn’t pay rent, 12.62% [10.69 - 14.56] skipped mortgage or rent payment, 8.87% [7.22 - 10.52] have been evicted, 20.45% [18.11 - 22.79] lived in a hotel or motel, 8.76% [7.12 - 10.39] lived in a place not meant for human habitation, 14.46% [12.43 - 16.50] slept in a shelter, and 13.0% [11.06 - 14.95] slept on the street. A generalized estimating equation revealed that exposure to homelessness among our patient population increased following the closing of the shelter (unadjusted OR = 1.25, 95% CI: 1.01 - 1.54, p = .04; adjusted OR = 1.38, 95% CI: 1.08 - 1.78, p = .01). We found a significant increase in the prevalence of homelessness at our safety net setting following the closure of Atlanta’s largest homeless shelter. Our study highlights that EDs are “first-stop” access sites for homeless persons. Given the well-documented association between homelessness, poor health, and repeat ED visits, our study suggests that EDs should expand and integrate the availability of homeless services to improve the health of these patients and to mitigate their reliance on ED services.