The standards list never names a device

A device file cites a consensus standard by number and edition — ISO 10993-1:2018, say — and that citation is only as good as FDA’s current recognition of it. Recognition is not permanent, and nothing about a product-code watch, a CFR-part watch, or a firm watch was ever built to notice when it lapses.

Recognition runs on a notice, not a rule

Section 514(c) of the FD&C Act — 21 U.S.C. 360d(c)(1)(A) — requires FDA to recognize a standard, in whole or in part, by publication in the Federal Register before a manufacturer can rely on a declaration of conformity to it. The same mechanism runs in reverse: recognition is withdrawn the same way it was granted, by a further notice, when a newer edition replaces an older one or FDA decides a standard no longer supports the requirement it was recognized for. There is no separate rulemaking, no CFR section devoted to the mechanics. The Federal Register notice is the whole act, both directions.

One list, many standards, indexed by number

FDA does not issue one notice per standard. It revises the list roughly twice a year under a running Recognition List Number — list 065 posted in February 2026, list 064 the summer before it — and each notice carries a table: standard reference numbers, titles, editions, and what changed — added, corrected, or withdrawn — for every standard FDA touched that cycle. The table is organized the way a standards body organizes a catalog, by designation and edition. It is not organized by device type, product code, or CFR part, because the list is not about devices. It is about which documents a submission is currently allowed to cite.

The notice knows exactly which standard it withdrew. It has no way of knowing, and does not try to say, which submissions on file anywhere still cite it.

Why no device-level watch catches this

A portfolio watch — this product’s or anyone else’s — joins on identifiers a device actually carries: a product code, a regulation part, a firm name, a clearance number. A standards-recognition notice carries none of those. It names a standard, not a device, and a match keyed to your product has nothing in that notice to join against — not because the ingestion missed it, but because the document genuinely contains no device-shaped key for a matcher to find. FDA Radar’s own sources page already says plainly that standards content isn’t part of what this product tracks yet; a notice like this is exactly the shape of document that gap describes.

The practice

The check has to run the other direction: keep your own list of every standard and edition your files actually cite — test reports, declarations of conformity, design history files — and diff it against FDA’s Recognized Consensus Standards Database yourself, on a cadence that assumes a list revision roughly every six months rather than waiting for one to announce itself. FDA’s own guidance says as much directly: verify the current recognition status of a standard before submitting, every time, because the database reflects today’s list and a saved PDF from your last submission does not.

Whether a specific withdrawal changes what a specific submission needs is a question for the regulatory professional holding that file, on those facts — this is a description of how recognition changes and why no device-level feed surfaces it, never a determination about any one standard or submission.