A market withdrawal isn't a recall
FDA's own definitions split a firm pulling product back into three distinct actions — recall, market withdrawal, stock recovery — by whether a violation exists at all, not by severity. Only the first is built to reach the enforcement database FDA Radar reads.
By Connor Griggs — Regulatory & Quality Strategist
A firm pulls a lot back from a distributor’s warehouse before it ever ships. A firm sends a correction letter to every hospital that bought a device with a labeling error. A firm stops shipping after a complaint pattern emerges and tells FDA about it. All three get called a “recall” in a press release, an investor filing, or a hallway conversation. FDA’s own rules say only one of them necessarily is.
Three actions, one test
21 CFR 7.3 defines all three terms, and the test that separates them isn’t severity — it is whether a legal violation exists and whether the product ever left the firm’s hands. A recall is a correction or removal of a marketed product FDA considers to violate the law and would act on, e.g. by seizure. A market withdrawal is a correction or removal involving either a minor violation FDA wouldn’t pursue, or no violation at all — the regulation’s own examples are normal stock rotation and routine equipment repairs. A stock recovery is narrower still: product that was never released for sale or use, still under the firm’s own control when it’s pulled. Three different facts about the same act of taking product back, and only the first one is a recall in FDA’s own vocabulary.
What that costs the other two
The device reporting rule at 21 CFR 806.10 requires a report to FDA within ten working days when a correction or removal is initiated to reduce a risk to health arising from a violation. A market withdrawal, by its own definition, usually has no violation to reduce a risk from — so the trigger that puts a correction on FDA’s desk in the first place was never built to catch it. A stock recovery never reached the market at all, which is most of why 806.10 has nothing to say about it either. Neither category carries FDA’s Class I, Class II, or Class III designation, because that classification exists to rank the health hazard of a recall specifically — there is no fourth value for “withdrawal” or “recovery” in the field. openFDA’s device enforcement endpoint, the one FDA Radar’s own recall pipeline reads, is keyed to recall_number and populated from FDA’s Recall Enterprise System — built to hold recalls, under FDA’s own definition of the word, not the two categories the same regulation carves out from it.
“We recalled it” is a sentence a firm can say accurately about an action that FDA’s own enforcement database was never going to record.
The practice
Finding nothing in FDA’s device enforcement data for a firm you’re watching is not the same fact as that firm having pulled nothing. A market withdrawal handled quietly — unshipped lots recovered, a labeling correction with no violation behind it — can be real, can be disclosed in a firm’s own communications, and can still never generate a recall_number. Reading a competitor’s or a supplier’s own press releases and investor filings alongside FDA’s recall data, not instead of it, is the only way to catch the gap between the word a company chooses and the category FDA’s rule actually assigns. This is regulatory intelligence about how the term is defined and where the record does and doesn’t reach, never a determination about what any specific firm’s own action was.
Primary sources
- eCFR — 21 CFR 7.3, Definitions
- eCFR — 21 CFR 806.10, Reports of corrections and removals
- FDA — Recalls, Corrections and Removals (Devices)
- openFDA — Device Enforcement field reference
- FDA Radar — what we monitor and how often
Regulatory intelligence, not regulatory advice. This post describes method and published FDA records as of its date; decisions about a specific device belong with your regulatory professional.