Every pharmacy has a short list of drugs everyone knows are bad. The list usually lives in someone’s head, gets applied inconsistently, and is rarely revisited when the market moves. That is an expensive way to run a decision that touches purchasing, patients and network obligations at the same time.
This is a way to structure it.
The question is not “is this drug profitable”. It is narrower and more useful: for this NDC, under the specific plans I actually see it under, after the real cost to serve, is the contribution negative, is the cause fixable, and what happens to the patient and the rest of their business if I stop.
Step one: get the claim-level truth, by plan
Pull ninety days for the NDC in question. Not the drug, the NDC, because MAC and acquisition both move at that level.
For each claim you want: date of service, BIN, PCN and Group, quantity dispensed, ingredient cost paid, dispensing fee paid, patient pay amount, total amount paid, Basis of Reimbursement Determination, and your true net acquisition cost for the pack you dispensed from.
Then group by plan, not by payer. This is the step most people skip and it is usually where the answer is. A drug that looks marginal in aggregate frequently turns out to be fine under four plans and badly underwater under one. That is a contract problem with a specific counterparty, not a stocking problem, and pulling the product punishes every patient to solve a problem with one plan.
Look at the distribution, not the mean. Ten fills at minus $40 and forty at plus $3 average out to something that looks tolerable and is not.
Step two: classify the cause before choosing a remedy
Four causes, four different responses.
Purchasing. Your net cost sits above the market. Test it against NADAC for the same NDC and unit. If you are meaningfully above the national average acquisition cost on a high-volume generic, the problem is upstream of the PBM: wrong labeler, wrong pack size, a generic compliance rate you are not hitting, or a secondary source you default to out of habit.
Self-inflicted pricing. Most contracts pay the lower of the contracted amount and your usual and customary price. If field 522-FM says usual and customary, you capped yourself. A cash price set three years ago can quietly govern a whole category.
Contract. MAC sits below any acquisition cost realistically available to you. This is the case that justifies an appeal, and appeal windows at several PBMs run in business days from the initial claim rather than from the remittance, which is why a monthly review misses most of them.
Product. Single-source generic, shortage, limited distribution, or a package size that forces a billing quantity you cannot buy economically. Nothing you do at the counter changes this one.
Only the fourth is genuinely a stocking decision. The other three have cheaper remedies that should be exhausted first.
Step three: price the cost to serve honestly
Gross margin per fill is not contribution per fill. The items that move the answer:
Staff time, including the fills that need a call to the prescriber. Prior authorization work, which is the most under-costed item in most pharmacies because the labour is invisible and unbilled. Cold chain handling. Delivery. Inventory carrying cost and expiry risk, which is disproportionate on slow-moving high-cost items. Rebilling and appeal handling for the claims that come back wrong.
You do not need precision here. You need order of magnitude. A drug at plus $4 gross margin that reliably consumes twenty minutes of technician time and a prior authorization is not a plus $4 drug, and treating it as one distorts every comparison you make.
This is also where automation changes the arithmetic rather than the price. Attergo Authorizations exists because prior authorization labour is a real cost of dispensing that never appears in a margin report, and moving it off a pharmacist’s afternoon can make a marginal product viable without anyone renegotiating anything.
Step four: the case that is currently hardest
Medicare’s negotiated maximum fair prices took effect for the first ten selected drugs on 1 January 2026, and the mechanism creates a stocking question that is genuinely new.
You buy at wholesale. The patient pays based on the MFP. You are made whole by a refund from the manufacturer, routed through the Medicare Transaction Facilitator. CMS requires manufacturers to transmit refunds within 14 calendar days of receiving claim-level data from the MTF, and states that refunds reach dispensing entity bank accounts within 21 days of a claim’s date of service on average.
An average of 21 days is a working capital position, not a margin problem. On high-cost products it can be a large one. NCPA’s January 2025 member survey found 32.8 percent of respondents had already decided not to stock one or more of the first ten selected drugs, with a further 60.4 percent considering it, and cash flow was the concern its leadership put front and centre.
The right analysis here is not gross margin. It is: what is my peak outstanding refund balance across all MFP claims at any point in the month, and can the business carry it. That is a treasury question and it deserves a number rather than a feeling.
Step five: the decision frame
| Situation | Reasonable response |
|---|---|
| Negative under every plan, no purchasing fix, low volume, ready alternatives | Stop stocking, arrange a transfer path, document the reason |
| Negative under one plan only | Keep stocking, escalate the contract, appeal the claims |
| Negative but anchors a patient who fills everything else with you | Evaluate contribution at patient level, not SKU level |
| Negative and you are the only local source | The decision is not purely financial. Decide deliberately, not by default |
| Negative because of working capital timing, not rate | A financing question, not a stocking question |
Two constraints sit over all of it. Your network agreements may contain obligations about dispensing and stocking, and several states have any willing provider and non-discrimination provisions that interact with them. Iowa’s SF 383, for example, pairs its reimbursement floor with strong any willing provider language. Read the contract and take advice before a decision hardens into a pattern.
Step six: make it a decision, not a habit
Write the outcome down. A do-not-order list with, for each item, the reason, the date, and a review date. Tell the team, so the answer is consistent whoever is at the counter. Then actually review it, because a decision made during a shortage is often wrong six months later and nothing prompts you to revisit it.
The re-check is the part that decays first, and it is the part that pays. Attergo Inventory watches the cost and reimbursement position on products you have already ruled out, so a drug that becomes viable again surfaces on its own instead of sitting on a list nobody has opened since the day it was written.