PDGM (Patient-Driven Groupings Model)

Every 30-day payment period is scored on five variables, and the combination determines the case-mix group:

• Admission source: community or institutional, based on whether the patient had a qualifying inpatient stay in the 14 days before admission
• Timing: early (the first 30-day period in a sequence) or late (all subsequent periods)
• Clinical grouping: one of 12 groups assigned from the principal diagnosis
• Functional impairment level: low, medium, or high, scored from OASIS functional items
• Comorbidity adjustment: none, low, or high, based on secondary diagnoses

That structure (2 x 2 x 12 x 3 x 3) produces 432 case-mix groups. Each group carries a case-mix weight that is multiplied against the national standardized 30-day base rate, then wage-index adjusted for geography.

Before 2020, the home health prospective payment system paid for 60-day episodes classified into 153 Home Health Resource Groups, and payment rose with the number of therapy visits delivered. CMS concluded that therapy thresholds created an incentive to add visits regardless of patient need. PDGM removed therapy volume from the payment calculation entirely and cut the payment unit from 60 days to 30. The 60-day certification period and comprehensive assessment schedule did not change, so each certification now contains two payment periods. Visit counts still matter at the low end: periods that fall below the Low Utilization Payment Adjustment (LUPA) threshold are paid per visit instead of at the full period rate.

The CY2026 home health final rule reduced aggregate Medicare payments by an estimated 1.3 percent, roughly 220 million dollars, versus CY2025. The net figure combines a 2.4 percent payment update, a 1.023 percent permanent behavioral adjustment cut, a 3.0 percent one-year temporary adjustment cut, and a small update to the outlier fixed-dollar loss amount. CMS also recalibrated the case-mix weights, functional impairment thresholds, comorbidity subgroups, and LUPA thresholds using CY2024 utilization data. The practical takeaway: the same patient mix can produce different revenue year over year, so agencies should re-model their book of business against the new weights each January.

PDGM rewards accuracy and utilization discipline rather than volume. Agencies that hold margin under the model tend to do a few things consistently: they code the principal and secondary diagnoses to the highest specificity the documentation supports, they complete OASIS functional items accurately so the functional level reflects the patient, they track visit counts against LUPA thresholds in real time, and they build visit plans around clinical need instead of habit. They also reconcile expected versus actual payment at the period level, because Medicare recodes admission source and timing from its own claims history, and surprises show up on remittances.