Effects of Permanent Supportive Housing on Health Care Utilization and Spending Among New Jersey Medicaid Enrollees Experiencing Homelessness.

BACKGROUND: Permanent supportive housing (PSH) programs have the potential to improve health and reduce Medicaid expenditures for beneficiaries experiencing homelessness. However, most research on PSH has been limited to small samples of narrowly defined populations. OBJECTIVE: To evaluate the effects of PSH on Medicaid enrollees across New Jersey. RESEARCH DESIGN: Linked data from the Medicaid Management Information System and the Homeless Management Information System were used to compare PSH-placed Medicaid enrollees with a matched sample of other Medicaid enrollees experiencing homelessness. Comparisons of Medicaid-financed health care utilization and spending measures were made in a difference-in-differences framework 6 quarters before and after PSH placement. SUBJECTS: A total of 1442 Medicaid beneficiaries enrolled in PSH and 6064 Medicaid-enrolled homeless individuals not in PSH in 2013-2014. RESULTS: PSH placement is associated with a 14.3% reduction in emergency department visits (P<0.001) and a 25.2% reduction in associated spending (P<0.001). PSH also appears to reduce inpatient utilization and increase pharmacy spending with neutral effects on primary care visits and total costs of care (TCOC). CONCLUSIONS: Placement in PSH is associated with lower hospital utilization and spending. No relationship was found, however, between PSH placement and TCOC, likely due to increased pharmacy spending in the PSH group. Greater access to prescription drugs may have improved the health of PSH-placed individuals in a way that reduced hospital episodes with neutral effects on TCOC.

Medicaid Utilization and Spending among Homeless Adults in New Jersey: Implications for Medicaid-Funded Tenancy Support Services.

Policy Points Large numbers of homeless adults gained Medicaid coverage under the Affordable Care Act, increasing policymaker interest in strategies to improve care and reduce avoidable hospital costs for homeless populations. Compared with nonhomeless adult Medicaid beneficiaries, homeless adult beneficiaries have higher levels of health care needs, due in part to mental health issues and substance use disorders. Homeless adults are also more likely to visit the emergency department or require inpatient admissions. Emergency care and inpatient admissions may sometimes be avoided when individuals have high-quality community-based care and healthful living conditions. Offering tenancy support services that help homeless adults achieve stable housing may therefore be a cost-effective strategy for improving the health of this vulnerable population while reducing spending on avoidable health care interventions. Medicaid beneficiaries with disabling health conditions and more extensive histories of homelessness experience the most potentially avoidable health care interventions and spending, with the greatest opportunity to offset the cost of offering tenancy support benefits. CONTEXT: Following Medicaid expansion under the Affordable Care Act, the number of homeless adults enrolled in Medicaid has increased. This has spurred interest in developing Medicaid-funded tenancy support services (TSS) for homeless populations as a way to reduce Medicaid spending on health care for these individuals. An emerging body of evidence suggests that such TSS can reduce avoidable health care spending. METHODS: Drawing on linked Homeless Management Information System and Medicaid claims and encounter data, this study describes the characteristics of homeless adults who could be eligible for Medicaid TSS in New Jersey and compares their Medicaid utilization and spending patterns to matched nonhomeless beneficiaries. FINDINGS: More than 8,400 adults in New Jersey were estimated to be eligible for Medicaid TSS benefits in 2016, including approximately 4,000 living in permanent supportive housing, 800 formally designated as chronically homeless according to federal guidelines, 1,300 who were likely eligible for the chronically homeless designation, and over 2,000 who were at risk of becoming chronically homeless. Homeless adults in our study were disproportionately between the ages of 30 and 64 years, male, and non-Hispanic blacks. The homeless adults we studied also tended to have very high burdens of mental health and substance use disorders, including opioid-related conditions. Medicaid spending for a homeless beneficiary who was potentially eligible for TSS was 10% ($1,362) to 27% ($5,727) more than spending for a nonhomeless Medicaid beneficiary matched on demographic and clinical characteristics. Hospital inpatient and emergency department utilization accounted for at least three-fourths of "excess" Medicaid spending among the homeless groups. CONCLUSIONS: A large group of high-need Medicaid beneficiaries could benefit from TSS, and Medicaid funding for TSS could reduce avoidable Medicaid utilization and spending.