Purpose of Study:
To evaluate whether screening for homelessness during hospitalization and deploying outpatient care managers (CMs) to the inpatient setting to rapidly connect patients to community social services reduce hospital readmissions.
Primary Practice Setting:
A large safety-net hospital in Minnesota.
Methodology and Sample:
A quality improvement pilot proactively engaged hospitalized adults experiencing homelessness and enrolled in the Hennepin Health Accountable Health Model (HH AHM). Patients were screened for homelessness using a novel housing indicator tool, and eligible patients were visited in the hospital by HH AHM community CMs. If patients chose to pursue offered services, they were added to the CM's caseload with whom they met while hospitalized. Outcomes were compared between those patients who engaged in community case management initiated in the hospital (intervention group) and those who declined services or were discharged prior to an inpatient CM visit (control group). Chi-square, Fisher's exact, independent-samples t test, and Mann–Whitney U tests were conducted, as appropriate.
Seventy-two patients were included in the intervention group and 61 patients were included in the control group. Both groups were primarily English-speaking, African/African American and Native American men in their early to mid-40s. In total, 5.6% and 18% of intervention and control patients were readmitted to the hospital within 30 days. Of note, claims data confirmed that no readmissions occurred at outside hospitals in the first 9 months of the program. Twelve (16.7%) intervention patients were housed over the course of the study period, 5 occurring in the first 6 weeks.
Implications for Case Management Practice:
- Upstream patient identification, relationship-based care, and inpatient connection to social services drive readmission reduction.
- Simple workflow redesign to avoid reliance on referrals results in improved identification of intervention candidates and may be completed without additional funding.