Home Health Pathways Relevant to Patient-Driven Groupings Model Classification Among Medicare Beneficiaries With and Without Dementia.

OBJECTIVES

To characterize admission-source and episode timing pathways among Medicare Home Health (HH) beneficiaries with and without Alzheimer's disease and related dementias (ADRD) immediately prior to the Patient-Driven Groupings Model (PDGM) implementation, which incorporated these factors into HH payment.

DESIGN

Retrospective secondary analysis of national Medicare claims data from 2019, with propensity score matching of beneficiaries classified as ADRD highly likely and ADRD unlikely.

SETTING AND

PARTICIPANTS

A total of 129,429 ADRD-highly likely HH beneficiaries were matched 1:1 to 129,429 ADRD-unlikely HH beneficiaries.

METHODS

Outcomes included annual counts of community and institutional HH admissions and late HH episodes, as well as Health Insurance Prospective Payment System 1 classifications reflecting admission source and episode timing. Incidence rate ratios (IRRs) and odds ratios were estimated to compare HH utilization and classification patterns between matched groups.

RESULTS

The matched cohort included 129,429 beneficiary pairs. Compared with ADRD-unlikely beneficiaries, ADRD-highly likely beneficiaries had more community admissions (mean, 2.7 vs 2.4; IRR, 2.75; 95% CI, 2.73-2.76) and more late episodes (mean, 2.35 vs 2.05; IRR, 2.35; 95% CI, 2.34-2.37). ADRD status was not associated with the likelihood of any institutional admission (odds ratio, 1.00; 95% CI, 0.98-1.01). Among beneficiaries with at least 1 institutional admission, ADRD-highly likely beneficiaries had higher institutional admission rates (IRR, 1.12; 95% CI, 1.09-1.14). ADRD-highly likely beneficiaries also had higher odds of community early and community late classifications relative to institutional early episodes.

CONCLUSIONS AND IMPLICATIONS: Immediately before PDGM implementation, ADRD-highly likely beneficiaries were more likely to enter HH from the community and to experience later episodes of care. These findings establish a pre-PDGM benchmark for understanding how beneficiaries with ADRD were distributed across admission source and episode-timing pathways subsequently incorporated into HH payment. Future research should evaluate whether PDGM has differentially affected HH access, utilization, payment, and care delivery for beneficiaries with ADRD.