340B Rebate Pilot for Manufacturers: The Data Work Behind the 1 January 2027 Launch

What 340B rebate pilot manufacturers have to be ready for

340B rebate pilot manufacturers start paying rebates instead of giving upfront 340B discounts on 1 January 2027, for 21 drugs from 10 companies, all through the Beacon platform. From that day each approved manufacturer must receive claims data from covered entities, pay or deny every complete claim within 10 calendar days, keep unjustified denials well under 5 percent, and report rebate, purchase and denial data to HRSA. That is a data operation, not just a pricing policy, and it touches trade, gross to net, market access and compliance teams at once.

The legal summaries on page one explain the rules well. What they do not cover is the work inside the manufacturer: matching rebate claims to WAC sales, deduplicating against the Medicare maximum fair price, watching denial rates daily, and accruing a liability that now moves with claims instead of chargebacks. This guide covers both: the rules as published, then the data layer, a readiness checklist and the reporting a commercial leader will ask for in the first week of January.

We are a data and analytics firm, not a law firm. Use this for operational planning and confirm legal positions with your counsel.

The 2027 pilot in numbers

How the revised 340B rebate model pilot program works

HRSA announced the revised 340B Rebate Model Pilot Program on 31 July 2026 and published the notice in the Federal Register on 3 August 2026. Manufacturers had until 24 August to submit plans. On 1 October 2026 HRSA posted approved plans from AbbVie, Amgen, Astellas, AstraZeneca, Bristol Myers Squibb, Boehringer Ingelheim, GSK, Merck, Pfizer and Teva, covering Imbruvica, Linzess, Vraylar, Enbrel, Otezla, Otezla XR, Xtandi, Farxiga, Eliquis, Pomalyst, Jardiance, Ofev, Tradjenta, Breo Ellipta, Trelegy Ellipta, Januvia, Janumet, Janumet XR, Ibrance, Austedo and Austedo XR, according to rxalmanac.com's summary of the approvals.

The mechanics, as summarised by Mintz and Epstein Becker Green:

• Covered entities buy at WAC. For drugs in the pilot, hospitals and clinics pay wholesale acquisition cost instead of the 340B price at purchase.
• They submit claims after dispensing. Claims go to the manufacturer through the rebate platform within 45 calendar days of the dispense.
• The manufacturer pays or denies within 10 calendar days of a complete submission, with documentation for any denial. If a submission is returned as incomplete, the clock restarts.
• The rebate is WAC minus the 340B ceiling price on the date of dispense, paid per prescription at the NDC-11 unit level, not bundled.
• Denials are restricted. Concerns about diversion or Medicaid duplicate discounts cannot be used to deny a rebate; they go to audit or administrative dispute resolution instead.
• The manufacturer pays for the platform. Covered entities are not charged fees, and must get 90 days' notice before a manufacturer's rebate model applies to them.

Scope is narrow on purpose. Only drugs selected for the Medicare Drug Price Negotiation Program for initial price applicability years 2026 and 2027 qualify, and only while their negotiated price is in effect. Covington describes the pilot as designed to help deduplicate the maximum fair price (MFP) and 340B discounts, so the same unit is not discounted twice.

Why this pilot exists, and what could still change

This is the second attempt. HRSA's first pilot, announced in 2025 and approved for nine drugs to start on 1 January 2026, was blocked by the US District Court for the District of Maine before it launched and was later vacated. On 21 July 2026 the D.C. Circuit ruled in Novartis v. Kennedy that manufacturers cannot switch to rebate models on their own: HRSA has to approve them first. The revised pilot is HRSA's approved route.

It is still contested. Pharmaceutical Commerce reported on 22 October 2026 that the American Hospital Association, the Maine Hospital Association and four safety-net hospitals have sued in federal court in Maine to block the revised pilot, and that a bipartisan House bill, the 340B Rebate Model Delay Act, would delay it. HRSA has committed to publish a claims dispute pathway by 30 January 2027 and an evaluation report is due by 30 April 2028.

For planning, that means two things. Build for a 1 January start, because the 10-day payment clock does not allow a slow ramp. And build so the same data layer still earns its keep if the pilot is paused: the claims, WAC sales and MFP data you join for the pilot are the same data your gross to net and trade teams need anyway.

340B rebate claims data: what arrives and what it is for

Pharmacy claims

Epstein Becker Green lists the approved pharmacy fields as date of service, date prescribed, Rx number, fill number, NDC-11, quantity dispensed, prescriber ID, service provider ID, 340B ID, Rx BIN and Rx PCN. Together they identify one dispense, at one site, for one product, and the payer route that BIN and PCN imply.

Medical claims

For drugs billed on medical claims the fields include date of service, claim number and claim line number, unit of measure, NDC-11, quantity, rendering physician ID, service provider ID, 340B ID, health plan name, health plan ID and its qualifier. Unit of measure is where errors start: a quantity in milligrams against a price per package will overstate or understate a rebate badly.

What you must not do

Covington's summary says manufacturers may not collect, aggregate, share or license covered entity data for purposes outside the pilot, with an exception for deduplication in Medicaid managed care. HRSA's guidance points to de-identification under 45 CFR 164.514, and notes that automated de-identification can happen before ingestion so the manufacturer or vendor never receives PHI. Design the warehouse so pilot data sits in its own governed zone and cannot drift into targeting or segmentation models.

The data layer 340B rebate pilot manufacturers need

Beacon collects and validates claims. It does not run your gross to net forecast, reconcile to your sales ledger or tell your VP of trade why rebates spiked in one region. That work sits with the manufacturer, and in most companies it spans four teams that have never shared a dataset.

1. A unified rebate claims table. Every claim from the platform lands with its status history: received, returned incomplete, resubmitted, paid, denied, disputed. Each status change carries a timestamp, because the 10-day clock is measured from the complete submission, not the first one.

2. A price table by NDC-11 and date. The rebate is WAC minus the 340B ceiling price on the date of dispense, and HRSA requires a quarterly price file for each NDC-11. Keep both prices with effective dates so any rebate can be recomputed exactly as it was paid. A WAC change mid-quarter is exactly where a rebate has to be recomputed and defended line by line.

3. WAC sales and chargeback history by 340B ID. Covered entities now buy at WAC, so the chargebacks your wholesalers used to process for 340B sales of these drugs should fall away. Joining purchases to rebate claims by 340B ID and NDC shows whether each site is claiming roughly what it bought, and flags sites still being charged back by mistake.

4. Medicare and MFP flags. Per HRSA's documentation as summarised by Epstein Becker Green, a rebate is only withheld when the unit went to a Medicare beneficiary at the maximum fair price and that price is lower than the 340B ceiling price. Your MFP effectuation data is the other side of that test, so the two programmes have to meet in one place.

5. Denial reasons as structured codes. Free-text denial notes cannot be counted. A short, fixed list of reason codes lets you report denials to HRSA, defend each one, and see early if the rate is heading towards 5 percent.

If you already run a pharma commercial data warehouse, this is a new subject area inside it, not a new system. If you do not, this is a sensible, contained place to start one.

A worked example: one month of rebate claims

The figures below are illustrative, chosen to show the arithmetic. They are not real prices for any drug in the pilot.

Suppose one product has a WAC of $600 per unit and a 340B ceiling price of $250 per unit. The rebate is $350 per unit. In January, covered entities submit 4,000 claims averaging 30 units each, so 120,000 units and a gross rebate exposure of $42 million for the month.

Two numbers matter to the leadership team. First, the denial rate: 80 other denials out of 4,000 is 2 percent, comfortably under 5 percent, but a single data error, such as a unit-of-measure mismatch on medical claims, could add hundreds of denials in a week. Second, the accrual: under the old model the 340B discount showed up as a chargeback near the sale; now the liability builds as units are dispensed and claims arrive up to 45 days later. Finance needs an estimate of units dispensed but not yet claimed at every month end, which is exactly the lag our gross to net guide warns about.

Readiness checklist for the next eight weeks

The 340B rebate analytics dashboard leaders will ask for

• Payment SLA

  Share of complete claims paid or denied within 10 calendar days, by day and by product, with the oldest open claims listed first.

  • Days open per claim
  • Clock restarts shown
  • Breaches named, not averaged

• Denial rate against 5%

  Denials by reason code, site and NDC, trended daily, so a rising rate is caught in days rather than at quarter end.

  • Coded reasons only
  • Split by pharmacy and medical
  • Disputes tracked to outcome

• Claims against purchases

  Units claimed versus units bought at WAC by 340B ID, to spot over-claiming, under-claiming and sites still receiving chargebacks.

  • By site and month
  • Lag-adjusted
  • Exceptions exported for audit

• MFP deduplication

  Medicare units at MFP matched against rebate claims, with the amount withheld and the evidence for each case.

  • Matched once per unit
  • Evidence stored
  • Feeds the dispute file

• Rebate liability and accrual

  Paid, pending and estimated unclaimed liability by product, ready for month-end close and the GTN forecast.

  • Paid vs accrued
  • True-up history
  • Ties to the ledger

• HRSA reporting pack

  Rebate, purchase and denial data in the format HRSA asks for, produced from the same tables the team uses daily.

  • One source of truth
  • Reproducible per period
  • Audit trail kept

Common mistakes, and when not to build anything new

The mistakes we expect are the ones we see in every new pharma data feed:

• Treating the platform as the system of record. Beacon validates and routes claims. Your finance, trade and compliance answers still have to come from data you control and can recompute.
• Pricing on the wrong date. The rebate uses prices on the date of dispense. Using the purchase date or the claim date creates small errors on every claim around a price change.
• Ignoring units of measure on medical claims. Physician-administered products billed in billing units need conversion to NDC units before the price is applied.
• Letting pilot data leak into commercial models. The data use limits are explicit. A shared table that a targeting model can read is a compliance problem waiting to surface.
• Reporting denials as a monthly average. A 2 percent month can hide a 9 percent week. The 5 percent line needs daily visibility.

When not to build: if you are not among the approved manufacturers and have no plan to apply in a later round, you do not need a rebate claims layer yet. Spend the time on your MFP and gross to net data instead, which pays off whether or not the pilot expands. And if your trade team already has a vendor that delivers reconciled claims, purchases and MFP data into your warehouse, add the dashboard rather than rebuild the pipes.

Greenwolf builds this kind of data layer for pharma commercial and market access teams: the pipelines, the matching logic, the warehouse model and the dashboards, on the cloud you already run. Our pharma commercial analytics service page explains how an engagement runs, and the wider specialty pharmacy analytics hub covers the patient services and access data that sit next to it. Related reading: payer hierarchy for pharma and Sunshine Act reporting, another federal reporting obligation built from the same governed data.

Questions about the 340B rebate pilot