Background: Hospitals nationwide increasingly shoulder the cost of unaddressed social determinants of health (SDOH). Preventable readmissions and prolonged hospital stays among patients experiencing homelessness increase healthcare expenditures and highlight unmet social needs linked to poorer outcomes. Every county in the United States administers housing assistance – including Permanent Supportive Housing, rapid rehousing, and emergency shelter – through a Continuum of Care (CoC), which uses a Homelessness Management Information System (HMIS) and a standardized prioritization process to allocate resources. Although every CoC relies on an HMIS-linked prioritization process, hospitals lack access to these data. In the county in which this pilot was implemented, the Vulnerability Index – Service Prioritization Decision Assistance Tool (VI-SPDAT) informs housing prioritization, and unhoused individuals cannot access county housing assistance without a valid VI-SPDAT score recorded in HMIS. Because VI-SPDAT scores expire annually but can be renewed after hospitalization, inpatient stays represent an often-missed opportunity for reassessment.
Purpose: To pilot a data-sharing initiative linking hospital services to county HMIS, identify hospitalized patients eligible for county housing assistance, and develop a scalable model for integrating SDOH and housing eligibility data into inpatient workflows to improve housing access for patients experiencing homelessness.
Description: From June 2023 to June 2024, unhoused patients admitted to the inpatient medicine service were identified through EHR demographic data and cross-referenced with the county HMIS to determine the presence and validity of VI-SPDAT assessments. This collaboration, the first of its kind in the county, operated under a data-use agreement with the county’s homelessness services office, allowing HMIS-trained inpatient social workers to view county homelessness data. Among 272 unhoused patients (mean age 54 years; 72% male; 85% English-speaking), 38% had a VI-SPDAT ever recorded in HMIS, 16% had a valid score within the prior 12 months, and 5% had an updated score within three months of discharge. Of those with any VI-SPDAT in HMIS, 72% met eligibility criteria for Permanent Supportive Housing.
Conclusions: Integrating hospital and county data systems offers a practical strategy to address SDOH and strengthen transitions of care. Allowing inpatient social workers to view county data and complete VI-SPDAT assessments during hospitalization supports more effective discharge planning, improves linkage to housing resources, and may reduce avoidable utilization. The low rate of post-discharge VI-SPDAT assessment (5%) in this pilot highlights a significant system gap, suggesting that most unhoused inpatients could benefit from VI-SPDAT screening during hospitalization and from inpatient teams having access to county HMIS data. This pilot provides a replicable framework for hospitals seeking to bridge longstanding data gaps between healthcare and social service systems and improve outcomes for patients experiencing homelessness.