The “essentially a copy” rule in compounding, explained

What is the test?

Two questions about the approved product, and neither is about quality.

FDA's current statement asks whether a commercially available drug product contains the same active pharmaceutical ingredient as the compounded preparation, at a strength that is the same, similar or easily substitutable; and whether that commercially available product can be used by the route of administration prescribed for the compounded one. Where both hold, the compounded preparation is treated as essentially a copy.

Notice what is absent. The test says nothing about the compounder, the state, the patient's insurance or the practice's ordering volume. It compares two products. That is why a supplier's answer to "can you compound this" can be correct in one month and wrong in the next without anything about the supplier changing.

Is there a way through it?

One, and it belongs to the prescriber.

The restriction does not apply where a prescriber determines and documents that the compounded product contains a change producing a significant difference from the commercially available product for an identified individual patient. Three features of that sentence do the work: the determination is the prescriber's, it is recorded rather than assumed, and it is made for a named patient rather than adopted as a formulary position.

A pharmacy cannot make it for you, and a standing protocol is not a determination for an identified individual patient. If a supplier presents a formulation as pre-cleared under this route, the question to ask is who they say made the determination and where it is documented.

Does adding an ingredient avoid it?

No — and FDA chose a combination product as its own worked example of the problem.

The agency's published illustration describes a compounded product combining one active pharmaceutical ingredient with a second, where the amounts of both sit within ten percent of the strengths of the respective commercially available products, and states that FDA may consider such a product essentially a copy. The example exists to show that a second ingredient does not by itself take a preparation outside the rule.

There is a separate point about frequency that is often merged with this one and should not be. FDA has stated that it does not intend to take action against a compounder for compounding a product that is essentially a copy regularly or in inordinate amounts where the compounder fills four or fewer prescriptions of that compounded product in a calendar month. That is an enforcement position about volume. It does not reclassify the product.

What does shortage-list status change?

Different things for the two entity types, and keeping them apart is the whole skill here.

For a state-licensed 503A pharmacy, FDA's position is that a drug is not considered commercially available while it is on the shortage list, and the limitations on compounding essentially copies do not apply in those circumstances. The restriction switches off, and switches back on when the listing ends.

For a facility registered with FDA as an outsourcing facility under section 503B, shortage status does something else entirely. It is one of only two lawful gateways to compounding from a bulk drug substance — the other being the 503B bulk drug substances list — and the condition requires the drug to be on the shortage list at the time of compounding, distribution and dispensing. FDA has said it may take action where a facility continues to fill new orders after the approved drug is removed from the list, or continues filling existing orders more than sixty days after removal.

The two gateways are independent. A proposal to change the bulks list decides nothing about the shortage route, and a shortage ending decides nothing about the list.

What is the status of the substances this bites hardest on?

Dated, per substance, and not the same answer for all three.

Two further facts sit across all three. FDA proposed on 30 April 2026 to exclude all three from the 503B bulks list; the comment period closed on 30 July 2026 after an extension, and no final determination had been made at this page's review date. And the litigation over the shortage determinations produced a Fifth Circuit judgment affirming FDA in the tirzepatide case on 27 August 2026, with the mandate scheduled for 19 October 2026 — so not yet final — while the separate semaglutide appeal remained pending.

Because those statuses expire, the per-substance records with their check dates live on the compounding status tracker rather than in prose here.

What may not be said about any of this?

Four things, and FDA has enumerated them itself in guidance addressed to companies promoting compounded drugs.

A compounded preparation may not be described as a generic version of, or the same as, an approved drug; nor claimed to have been approved by FDA or evaluated by the agency for safety and effectiveness; nor asserted to be clinically equivalent to the approved product; nor described as coming from a facility FDA has approved or licensed, since the agency states it does neither. The reason a copy analysis exists at all is that the two products are not interchangeable in the regulatory sense, so language implying they are contradicts the framework being relied on. That point is developed at why a compounded drug is never FDA-approved.

Where does this leave a prescriber?

With a question that has a date attached to it.

The eligibility answer for any preparation is the product of the copy analysis above, the bulk-substance route available to whichever entity type will compound it, and — separately — whether your own state lets your practice hold stock at all. The first two are federal and move; the third is state law and moves more slowly.

The condition underneath all of it is set out at the patient-specific prescription requirement, the two entity types are compared at 503A pharmacies and 503B outsourcing facilities, and the state layer is in the office-use guide. The test itself, stated in FDA's own words in a single passage, is at essentially a copy.