LEAD Benchmark Brief
Your benchmark is being set right now, and it will not be rebased until 2036.
Illustrative worked example — 100,000 attributed lives (VOL_III_91_CALIBRATED_INPUTS). Zero customer data. Illustrative
Download PDFThe 2026 Benchmark Note (one page)
§1 · What changed
- ACO REACH concludes 31 December 2026.CMS: ACO REACH Model
- LEAD Performance Year 1 begins 1 January 2027.CMS: LEAD (Long-term Enhanced ACO Design) Model
- LEAD runs 10 years with no benchmark rebasing (through 31 December 2036).CMS: LEAD Model Overview
- The LEAD baseline is built from the three preceding years of claims (CY 2024, CY 2025, and CY 2026).CMS: Centers for Medicare & Medicaid Services Long-term Enhanced ACO Design (LEAD) Model FAQ
§2 · Why that changes the maths
Annual avoided spend (gross, pre-arrangement retention): $1,050,000 Illustrative
| Arrangement | One-year retained | Cumulative (ten-year) | What CMS takes |
|---|---|---|---|
| LEAD Professional | $521,850 Illustrative | $5,218,500 Illustrative | $63,000 Illustrative |
| LEAD Global | $1,012,389 Illustrative | $10,123,890 Illustrative | $376,110 Illustrative |
| MSSP BASIC Level E (rebases) | $525,000 Illustrative | $2,625,000 Illustrative | $0 Illustrative |
§3 · The one thing that cannot be done later
Once ordering behaviour changes, nobody can reconstruct what would have happened under the old pattern. A locked counterfactual is how you know whether your programme moved the needle — and the LEAD Base Years close when Performance Year 1 opens. You cannot go back and measure the pre-period after the intervention has already entered the claims. Separately, and for the same reason, 2026 is not just another base year: for a new-to-accountable-care ACO it carries 60% of the three-year benchmark weight. Efficiency that lands in 2026 is absorbed into the ACO's own ten-year benchmark rather than retained as savings; efficiency that waits until PY1 compounds through 2036 without rebasing. That is why we measure now and intervene on 1 January 2027 — the counterfactual and the weighting are the same argument from two directions. Locking both is the one thing that cannot be done later.
§4 · The offer
Free baseline from your claims extract. Required fields are listed below; we handle hashing and suppression; sample report attached.
If you never buy anything, you keep the report.
Turnaround: ten business days from receipt.
Required fields: Member id · Claim id · Date of service · HCPCS / CPT · Ordering / referring NPI
Member identifiers are SHA-256 hashed at the ingestion boundary. Small cells (n < 11) are suppressed and stated, never silent.
§5 · Who we are
- We shipped the arm that didn't work first, like everybody else — then published the correction from the randomised trial.
- ARKA can auto-approve; it can never auto-deny.
- There are no FHIR scopes, because there is no integration.