Medicaid Home and Community-Based Services Initiation and Acute Services Use.

Importance: For more than a decade, Medicaid has funded the majority of long-term services and supports through home and community-based services (HCBS). Whether access to Medicaid HCBS may affect dual-eligible beneficiaries' use of Medicare-covered medical services is not well understood. Objective: To determine whether Medicaid HCBS initiation is associated with changes in acute services use and medication fills. Design, Setting, and Participants: This cohort study used Southern Community Cohort Study data linked to Medicare and Medicaid claims to identify a cohort of older adults residing across 11 southeastern states, largely recruited from community health centers between 2002 and 2009. The sample included older adults with 12 months of continuous traditional Medicare or Medicare Advantage enrollment centered on first month of Medicaid HCBS initiation. All analyses were conducted from spring 2023 to fall 2025. Exposure: Initiation of Medicaid HCBS between 2006 and 2018. HCBS was identified by presence of a Medicaid personal care claim or 1915(c) waiver enrollment or claim. Main Outcomes and Measures: Inpatient discharges, emergency department (ED) use, and unique prescription drugs filled estimated using linear regressions with an event study structure. Event study models with person and year fixed-effects estimated changes in person-month probability of ED use and inpatient discharge and unique drugs filled in the 6 months before and after initiation of HCBS. Subanalyses estimated differences by pre-HCBS Medicaid enrollment status, Medicaid waiver vs state plan HCBS use, Alzheimer disease and related dementia, and diabetes diagnosis. Results: In a sample of 1218 new HCBS users (75% female, 77% Black; mean [SD] age, 70.5 [7.5] years), Medicaid HCBS initiation was associated with a decrease in probability of ED use (-2.70 percentage points; 95% CI, -4.18 to -1.22 percentage points) and a decrease in probability of inpatient discharge (-2.63 percentage points; 95% CI, -3.75 to -1.51 percentage points). These differences represent a 24% decrease in the within-person probability of any ED use and a 32% decrease in inpatient discharge down from adjusted pre-HCBS probabilities of 11.4% and 8.1%. Event-month trends demonstrated discontinuity at HCBS initiation and reduced the probability of acute services use that was maintained 6 months thereafter. The number of unique drugs filled steadily increased before and after HCBS, with no discontinuity observed at HCBS initiation. Conclusions and Relevance: In this study among a large cohort of older adults with low income across the southeastern US, Medicaid HCBS was associated with a persistent decrease in acute services use. Future research should explore additional outcomes to better inform policies that can improve HCBS outcomes and an understanding of its tradeoffs with acute services use.