503A vs 503B: what the FDA 2026 compounding proposal actually changes

Reported GLP-1 medication errors rose from about 2,000 in 2020 to more than 25,000 in 2025, according to a KFF Health News analysis of FDA adverse-event data. That single number is why the FDA's 2026 compounding proposal is suddenly on every health desk's radar, and why your inbox may already have a worried message from a friend asking whether the medication they get is about to change. If you have ever felt confused about where your compounded medication comes from, you are not alone, and this is the week that confusion finally has a clear answer.

Short answer: this is a proposal, not a final rule. The comment period closes on June 29, 2026. And the part most people quietly assume is at risk, the patient-specific compounding your own prescriber orders for you, is not the subject of this proposal at all.

Here is what the 503A vs 503B distinction actually means, in plain language, and who each side affects.

The 2026 compounding proposal is the FDA's published plan to update parts of how outsourcing facilities operate, including how the agency manages its bulk-ingredient lists. You can read the agency's own compounding overview on fda.gov. It is open for public comment until June 29, 2026.

A proposal is a draft. It is the stage where the FDA invites pharmacists, clinicians, patient groups, and the public to weigh in before anything becomes binding. Reporters tend to pre-write these stories ahead of the comment close, which is exactly why this peg is moving now.

• A draft, open for public comment through June 29, 2026.
• About 503B. Focused on outsourcing-facility (503B) compounding and the bulk-ingredient lists those facilities rely on.
• Normal rulemaking. A normal part of how federal rulemaking works, with a comment-and-revise step built in.

• A final rule. Nothing here is settled yet.
• A change to 503A patient-specific compounding. Compounding tailored to an individual patient under a prescription is not the subject of this proposal.
• A statement about efficacy. This proposal does not speak to whether any compounded medication works better or worse than a branded product.

The two numbers come from sections of federal law, and the difference is mostly about scale and oversight.

503A pharmacies make a medication for a specific patient, based on a specific prescription. Think of the local compounding pharmacy preparing a tailored formulation because a patient cannot use the commercial version. The batch is small and patient-specific.

503B outsourcing facilities make larger batches, often without a patient-specific prescription, and register with the FDA to do so under tighter manufacturing standards. They supply clinics and providers at a different scale.

This is where the headlines and the reality tend to drift apart, so it helps to be precise about who feels what.

• People who receive patient-specific compounded medication (503A): this proposal does not target your prescription. If you have questions, the right move is to ask your own prescriber and pharmacy directly.
• Clinics and telehealth providers that source from outsourcing facilities (503B): this is the group most directly touched, because the proposal concerns how those facilities and their ingredient lists operate.
• Pharmacists and clinicians who want a say: the comment period is the formal channel, and it is open until June 29.
• Patients watching the GLP-1 supply story: the rising medication-error figure (the KFF analysis of FDA data) is the backdrop, a reminder of why oversight of how these medicines are prepared and dispensed is getting attention. It is KFF's measured figure, not a claim about any single provider.

The headline figure (reported GLP-1 medication errors rising from about 2,000 in 2020 to more than 25,000 in 2025) is a KFF Health News analysis of the FDA Adverse Event Reporting System (FAERS), reported in 2025. It counts reported medication errors, not confirmed harm, and it is KFF's figure, attributed to KFF, not original sipra data. The regulatory facts in this piece (the 503A and 503B definitions, the proposal status, and the June 29, 2026 comment close) are drawn from the FDA's published compounding pages. No efficacy or equivalence claim is made about compounded versus branded products. Sources verified 2026-06-06.