Background: Any given night, an estimated 553,000 individuals experience homelessness in the USA. Many turn to hospitals for healthcare, food, and shelter. These factors and lack of follow-up care upon discharge contribute to high rates of early readmission and frequent ED visits. Due to frequently missed homelessness status by healthcare providers, it is difficult to assess patient demographics and readmission rates among this population. The Homeless Management Information System (HMIS) collects information on individuals that have stayed at shelters or transitional housing in Washington, DC. We utilized the HMIS to identify patients experiencing homelessness at an urban academic hospital.
Methods: Full names and birthdates from adults in HMIS were searched in the Electronic Medical Record (EMR) to locate patient matches. Inclusion criteria were having had at least 1 ED visit (without admission) or hospital admission from 9/1/2019-2/29/2020. Descriptive statistics were performed for demographics (gender, race, age, family status) and encounter information from the EMR. We used Fischer’s exact test or unpaired t-test to compare the demographics of patient matches to those of HMIS-listed individuals who did not have hospital records.
Results: Out of 5,025 HMIS entries searched, 2,523 (50.2%) had hospital records, with 702 patients (14%) having had at least 1 ED visit or admission in our 6-month timeframe. From 702 patients, we reviewed 586 ED visits without admission and188 admissions (up to 1 ED visit and 1 admission per patient). 2507 unmatched persons in HMIS were used as non-patient comparisons. Most patients were single adults without families (93.5%). Most patients had Medicaid (39.59% of ED visits, 54.26% of admissions) or municipally-funded insurance for low-income residents (29.18% of ED visits, 30.85% of admissions). 23.04% of patients who visited the ED without admission and 7.54% of admitted patients had no insurance. The most common primary diagnosis was acute ingestion or drug abuse/overdose (15.8%) in ED visits, and psychiatric disease (19.5%) in admissions. Primary teams caring for admitted patients were Internal Medicine (37.0%), Psychiatry (20.0%), OB/GYN (15.4%), and Cardiology (14.5%), Surgery (11.8%), and Intensive Care (3.18%). The 30-day ED revisit rate including only ED visits without admission was 27.99%. 10.11% of admitted patients were readmitted within 30 days. Patients were, on average, older [45.23 years old (SD=14.7) versus 43.26 (SD=15.92);p=0.0034], more commonly African American (87.32% versus 77.22%; p< 0.0001) and less commonly Hispanic (4.42% versus 9.45%; p< 0.0001) when compared to non-patients. Fewer homeless patients than non-patients were male (59.26% versus 71.44%; p< 0001).
Conclusions: Homeless patients in this urban academic hospital were mostly middle-aged African American men. Compared to non-patients, there was a higher percentage of women and African Americans among patients, and a lower percent of Hispanics. More admitted patients had Medicaid and more ED visits without admission were among uninsured patients. More than half of admitted patients were cared for by Internal Medicine and Psychiatry teams. Readmission rates were low, however nearly one-third of ED visits had a subsequent ED visit within 30 days. This information may help guide the distribution of resources for interventions within hospitals.