READMISSION RATES AMONG PATIENTS AMONG PATIENTS EXPERIENCING HOMELESSNESS IN AN INTEGRATED HEALTHCARE SYSTEM

ABSTRACT

BACKGROUND: Homelessness is a significant public health concern in the United States and is an important risk factor for poor health outcomes. There is limited data regarding the hospital utilization by this vulnerable population, especially inmanaged care settings. Historically, it has been difficult to identify patients experiencing homelessness at the population health level. In 2019, due to the passage of state law SB1152, hospitals across California now need standardized documentation policies for patients experiencing homelessness. This study assessed hospital readmission rates among hospitalized patients experiencing homelessness as identified through documentation in the electronic health record (EHR) as compared to the general hospitalized population within a large integrated health system. METHODS: This was a retrospective cohort study following adult patients (age≥18 years) hospitalized in Kaiser Permanente Northern California (KPNC) Medical Centers between 1/1/2019 through 12/1/2020. Patients were identified as homeless or housing insecure if they had SB1152 documentation, a homeless diagnosis code, or address history indicating homelessness within the extensive integrated KPNC EHR. A control group was created using 12 propensity score matching using Elixhauser comorbidities and demographics. Sensitivity analyses were performed to compare patients with an index hospitalization occurring during the COVID-19 shelter-in-place period between March 2020 and December 2020 with those whose index hospitalization occurred before this period. The primary outcome was 30-day readmission rate to the hospital or ED, and secondary outcomes included length of index hospitalization, and time to inpatient (IP) readmission. RESULTS: A total of 12,909 patients were included with 4,303 patients in the homeless group. Patients experiencing homelessness had increased odds for any 30-day readmission (OR 1.59;95% CI 1.44-1.76), for inpatient readmission (OR 1.36;95% CI 1.17-1.57), for ED readmission (OR 1.63;95% CI 1.47-1.80), and had longer stays during their index hospitalization (IRR 1.12;95% CI 1.04-1.21). The COVID-19 shelter-in-place period was not associated with any changes in the primary or secondary outcomes studied. CONCLUSIONS: Patients experiencing homelessness are at an increased risk for readmissions and longer hospitalizations compared to the general hospitalized population. Documentation of housing status following SB1152 has improved the ability to study hospital utilization among patients experiencing homelessness. Understanding patterns of hospital utilization in this vulnerable group will help providers to identify timely points of intervention for further social and healthcare support.