JAMA Netw Open. 2026 May 1;9(5):e2612232. doi: 10.1001/jamanetworkopen.2026.12232.
ABSTRACT
IMPORTANCE: Medicare home health is a critical postacute care source for individuals with Alzheimer disease and related dementias (ADRD). The Home Health Value-Based Purchasing (HHVBP) program has substantially reshaped financial incentives in home health, but its implications for patients with ADRD are unclear.
OBJECTIVE: To assess the association between residence in a state with an HHVBP program and home health service volume within 30 days of home health initiation among Medicare beneficiaries with ADRD, and how this varies by patient race and/or ethnicity, dual eligibility, and home health agency (HHA) racial composition.
DESIGN, SETTING, AND PARTICIPANTS: This cohort study linked 2021 to 2022 Medicare claims data with Home Health Focus and Provider of Services files for Medicare fee-for-service beneficiaries with ADRD older than 65 years who initiated home health care within 14 days of hospital discharge in the 9 states in which an HHVBP program was piloted compared with non-HHVBP states. The analysis was conducted from November 2024 to June 2025.
EXPOSURE: Residence in an HHVBP pilot state vs a non-HHVBP state.
MAIN OUTCOMES AND MEASURES: Number of nursing and therapy visits received within 30 days (main analysis) and 14 days (sensitivity analysis) of home health initiation. Negative binomial regression models with HHA random effects were estimated, and average marginal effects were reported.
RESULTS: The analytic cohort included 264 601 Medicare beneficiaries with ADRD (median [IQR] age, 83 [77-89] years; 160 947 males [60.8%]), of whom 68 765 (26.0%) resided in HHVBP states. The overall racial and ethnic composition included Asian (3.1%; n = 8099), non-Hispanic Black (8.9%; n = 23 634), Hispanic (6.3%; n = 16 562), and non-Hispanic White (81.7%; n = 216 306) individuals, as determined by Research Triangle Institute race codes from the Master Beneficiary Summary File. Adjusted 30-day outcomes showed HHVBP state residence vs non-HHVBP state residence was associated with 0.46 fewer nursing visits (3.94 vs 4.40 visits; average marginal effect [AME] = -0.46; 95% CI, -0.52 to -0.41; P < .001) and 0.32 more therapy visits (6.37 vs 6.05 visits; AME = 0.32; 95% CI, 0.24-0.39; P < .001). In non-HHVBP states, dual-eligible beneficiaries and those receiving care from HHAs primarily serving a racial and ethnic minority population received more nursing but fewer therapy visits compared with non-dual-eligible beneficiaries and those receiving care from HHAs serving a predominantly White population. These differences became smaller in HHVBP states. For example, therapy visit gaps between patients receiving care from HHAs primarily serving a racial and ethnic minority population and those receiving care from HHAs serving a predominantly White patient population narrowed from -0.81 (95% CI, -0.90 to -0.73) to -0.27 (95% CI, -0.44 to -0.09) visits in non-HHVBP vs HHVBP states (P < .001). The findings of increasing use of therapy visits and smaller group differences were consistent within a 14-day window.
CONCLUSIONS AND RELEVANCE: In this cohort study of Medicare beneficiaries with ADRD, residence in an HHVBP state was associated with increased therapy use and smaller differences in home health service patterns between racial and/or ethnic minority and White patients, dual-eligible and non-dual-eligible beneficiaries, and patients receiving care from minority-serving HHAs and those receiving care from HHAs serving a predominantly White patient population compared with residence in a non-HHVBP state. These findings suggest that VBP models could reduce differences in postacute care.
PMID:42126880 | DOI:10.1001/jamanetworkopen.2026.12232
