You already know the shape of this problem. A claim adjudicates, it pays, and the number is wrong in the direction that costs you money. What is worth spending time on is the mechanism, because the three things that produce a below-cost claim have three different remedies, and treating them as one problem is why most pharmacies get nowhere with it.
In its January 2025 member survey, NCPA reported that 40.8 percent of responding independent pharmacists said they were paid below NADAC on more than 40 percent of the prescriptions they filled for Medicare Part D patients, and 29.2 percent said the same about half or more of their Part D volume. Whatever your own figure is, it is not an anomaly in the market.
Mechanism one: MAC decides the ingredient cost, and it is not tied to your invoice
For multi-source generics, most network contracts do not pay a stated discount off AWP. They pay a Maximum Allowable Cost, a per-unit ceiling the PBM sets for a group of pharmaceutically equivalent products.
Three properties do most of the damage.
MAC prices the molecule, strength and form, not the NDC on your shelf. Two pharmacies filling the same script from different labelers are paid the same and earn very different margins.
The PBM controls the update cadence. When an acquisition market moves quickly, your cost can double in a week while the MAC sits still for a month. That lag is not an error from the PBM’s side. It is part of how the aggregate guarantee described below gets funded.
The response tells you MAC was applied, if you keep the field. NCPDP Basis of Reimbursement Determination (522-FM) distinguishes an ingredient cost paid as submitted from one reduced to MAC, from usual and customary, and from contract or acquisition pricing. If you archive one field beyond the money, archive that one. It separates “this contract is thin” from “we capped ourselves with our own cash price”, and those need opposite responses.
Mechanism two: the guarantee is an aggregate, and the true-up is still retroactive
Most network agreements contain a generic effective rate (GER) and a brand effective rate (BER). These read like rates and behave like budgets. The PBM commits that generic reimbursement across a defined claim population over a defined period will average out to some percentage off AWP.
Nothing in that commitment applies to your claim. Individual claims are priced by MAC, and the aggregate is the residual. As Frier Levitt and others have documented, the reconciliation typically runs after the measurement period, often annually, and is assessed against the aggregate performance of a network rather than a single store. Where a PSAO signed on your behalf, the population being measured is the PSAO’s book, not yours. If the network came in above the guaranteed rate, the recoupment lands on pharmacies that had no visibility into the arithmetic and no ability to influence it.
This matters for how you read the 2024 change. CMS redefined the Part D negotiated price so that, from 1 January 2024, all pharmacy price concessions must be reflected at the point of sale, with the negotiated price set at the lowest amount a pharmacy could receive under its contract. The retroactive Part D clawback in its old form is largely gone. The concession did not disappear. It moved into the number on the screen, which is why the transition felt like a rate cut and a cash-flow event at the same time rather than a reprieve. GER and BER reconciliation in commercial and other non-Part D business was not touched by that rule and continues to operate retroactively.
Mechanism three: everything else that moves after adjudication
Per-claim transaction and network fees. Statement-level adjustments with no claim reference. Audit recoupments, which have become materially more aggressive. Frier Levitt’s 2025 guidance notes that PBMs increasingly frame reviews as investigations rather than audits, which side-steps the protections in state fair audit laws.
None of these show up in a gross margin report built from dispensing data alone.
What the arithmetic looks like
The figures below are illustrative. They show the structure, not a real contract.
| Line | Amount |
|---|---|
| Net acquisition cost, 30 count bottle | $89.55 |
| Ingredient cost paid, 522-FM indicates MAC | $32.85 |
| Dispensing fee paid | $1.25 |
| Patient pay amount | $0.00 |
| Total amount paid | $34.10 |
| Gross margin before labour | ($55.45) |
Bought six months earlier at $18.40, the same NDC against the same MAC produces a positive margin. Nothing in the contract changed. The acquisition market moved and the MAC did not follow.
That is the entire diagnosis. A below-cost claim is a purchasing problem, a contract problem, or the lag between the two, and you cannot tell which from the profit and loss statement.
What actually moves the number
Buy against the MAC, not against the invoice. Where the PBM prices by NDC rather than by molecule, the labeler you stock changes what you are paid. Package size matters too. A MAC expressed per unit interacts badly with a package that forces a billing quantity you did not intend.
Appeal, and appeal inside the window. MAC appeal deadlines vary by PBM and by state, and several are counted in business days from the initial claim rather than from the date the remittance arrives. Optum Rx, Express Scripts, Prime Therapeutics and WellDyne all publish their own submission routes and turnaround commitments, and some states, Washington among them, restrict a PBM’s ability to reject an appeal on the grounds of incomplete fields. The constraint is almost never the argument. It is finding the claim before the window closes.
Check your usual and customary price. Most contracts pay the lower of the contracted amount and U&C. A cash price last reviewed two years ago quietly caps a whole category.
Read the measurement period and the exclusions. If your contract carries a GER, find out which claims are excluded from the calculation and when the period closes. Those two facts explain most mid-period rate movements that otherwise look arbitrary.
Measure per claim, not per month. A monthly gross margin percentage hides the shape of the distribution. Twelve claims at minus $55 and four hundred at plus $2 average out to something that looks survivable and is not.
The common requirement across all five is the same: each claim sitting next to what that specific NDC actually cost you, close enough to the fill that the appeal window is still open and the next purchase order has not yet gone out. That comparison is what Attergo Margin does off the live prescription event stream, though the method is worth running whether or not you automate it. The pharmacies that recover this money are not the ones with better contracts. They are the ones who can see the claim in time to do something about it.