Background: Individuals experiencing homelessness have increased disease burden and healthcare needs compared to the general population. Homelessness is correlated with poor health outcomes and higher early mortality rates. Despite increased health care needs, individuals experiencing homelessness have more barriers and therefore less exposure to ambulatory care with consequently higher rates of acute hospital care. Unhoused individuals have longer hospital stays with higher annual costs and readmission rates. Despite these challenges, various interventions can significantly impact the health of undomiciled individuals, including case management. Hospitalization of patients experiencing homelessness is a unique opportunity to connect them to care and improve health outcomes.
Methods: Stakeholder Meetings: Social work leadership, care management, nursing leadership, discharge coordinators and community partners were identified as key stakeholders. Process mapping: An admission flowsheet was developed through shadowing in the Emergency Department and stakeholder meetings to better understand patient flow through the emergency department and into the inpatient setting. Chart query: Chart query was performed by searching the keyword “homelessness” in all Epic flowsheets for admitted patients. This notably excludes text input in progress notes. Manual chart review: A random selection of patients identified as homeless in their chart were manually reviewed to determine whether housing status was documented within provider notes, and whether social work had been consulted and involved in their care.
Results: Stakeholder meetings and process mapping identified key processes through which all patients pass, including registration, nursing intake, and provider evaluation. None of these steps had a defined process for homelessness screening. A chart query for homelessness in EPIC flowsheets from September through November of 2020 identified 22 different locations in which housing status was captured. 61 of 87 (70%) encounters had homelessness documented in multiple locations and the most common location (care management evaluation) was documented in 49% of encounters. An address was provided on admission in 76% of cases. Manual chart review of 20 randomly selected individuals from this data indicated that housing status was documented in provider notes in 50% of cases. 15/87 encounters did not have a social work consult placed; however, social work still evaluated the patient in 11/15 cases. Data pulled from March- May and June-July of 2021 indicate that 71% and 83% of encounters had a social work consult and 36% and 50% of encounters had housing status filled in on the nursing admission navigator.
Conclusions: The process through which individuals experiencing homelessness are identified and documented within the medical record is highly variable and there is no centralized process for identifying housing status for inpatients. Areas identified for intervention include patient registration or nursing intake, although both these face barriers including increasing workload and the lack of notification to the inpatient team if a patient screens positive. Encouragingly, nearly all patients that are identified as homeless or housing insecure during admission are seen and evaluated by social work. Data from this project was used to support funding for the Homeless Transition Team that is now in place and follows patients longitudinally and facilitates complex discharge planning.