“The quality unit approved the release” — how FDA measures quality unit authority
483 response filed 15 January 2026 · Citations 21 CFR 211.192, 211.42(c)(10), 211.113(b), 211.63
A viable air monitoring plate in an ISO 5 aseptic filling area came back too numerous to count, with confluent bacterial growth and insect larvae. The same day, personnel monitoring plates for an operator in the adjacent ISO 7 room did the same. The firm investigated, concluded the contamination had entered the plates after sampling, rated the risk low, and released the lot.
Issues 01 through 03 were about how a firm reasons its way to a conclusion. This one is about who was supposed to stop it. A quality unit’s authority is not established by the org chart or the quality manual. FDA reads it out of the dispositions the unit was willing to refuse.
Nothing here says the unit lacked a charter or a signature block. It says the unit approved a release it should have refused.
What FDA found
The firm manufactures APIs and sterile injectables. During manufacture of epinephrine injection, lot EA038A5, a viable air plate beside the filling equipment in the ISO 5 area returned a too-numerous-to-count result with confluent bacterial growth and numerous insect larvae. Personnel plates from an operator in the adjacent ISO 7 room, the same day, did likewise.
The investigation attributed both to what FDA called “an unsubstantiated hypothesis involving insect infiltration of sealed monitoring plates.” The firm’s own records cut against it: procedures require plates to be inspected before use and handled strictly after sampling, the media batch passed negative control testing, and the deviation report showed no defect or sampling error.
Risk was nonetheless rated low — in FDA’s words, “based on overreliance on passing sterility test results” — and the quality unit approved release.
- 9–19 Dec 2025
- FDA inspects the South El Monte facility.
- 15 Jan 2026
- Firm files its 483 response: risk low, no corrective actions needed.
- 26 Mar 2026
- FDA recommends recall of lot EA038A5 by teleconference.
- 1 Apr 2026
- Firm recalls the lot for lack of sterility assurance.
- 2 Jul 2026
- Warning letter issues, citing quality unit authority.
The letter sets that decision against a longer record: three field alert reports for contamination in sterile products since 2021, and over 90 complaints since December 2023 — many closed without a root cause.
What the firm said
The response was not evasive. It restated a position the firm had reached honestly, on evidence quality units are trained to weigh: monitoring data elsewhere that day was mostly acceptable, making a facility-wide event look unlikely, and the sterility test passed.
“post-exposure adventitious contamination external to the filling suite” Firm’s 483 response, as quoted by FDA
From that the firm concluded the two events were isolated incidents, the risk was low, and — the sentence that decided the letter — that no corrective actions were needed.
Why FDA rejected it
A hypothesis your own procedures contradict is not a root cause
Plate infiltration required the plates to have been compromised despite pre-use inspection, specified handling, and passing negative controls — with no evidence of a defect. FDA’s verdict: the outcome “remains poorly supported by data.” A theory that survives only because nothing was done to test it is not a finding.
Two events, same day, adjacent rooms
The response treated the findings as isolated incidents. FDA read them as one signal. Coincidence is defensible only after correlation has been ruled out, and co-occurrence in adjacent classified rooms on a single day pointed the other way.
“The occurrence of two severe contamination events on the same day in adjacent rooms indicates a significant breakdown in contamination control.” FDA, Warning Letter 320-26-99
“No corrective actions were needed” is not a conclusion an investigation can reach
FDA held that the absent CAPA “reflects a critical deficiency in your investigation.” Note the direction of that sentence: the missing CAPA is evidence about the investigation, not a separate finding. A deviation grave enough to investigate that yields nothing to correct usually means the investigation stopped at the first explanation permitting release.
A passing sterility test is not evidence the process was in control
The risk rating rested on the final test. FDA is categorical: a sterility test “cannot be solely relied upon as justification to release each drug product batch,” because “contamination is typically episodic and not uniformly distributed.” A passing test confirms control already established; it cannot create control where direct observation contradicts it.
The pattern worth carrying forward
Quality unit authority is rarely lost by decision. No one removes the veto. It simply never gets exercised, because each case arrives with a plausible reason not to use it — a technical explanation, a passing result, a schedule. The authority goes ceremonial while the org chart shows it intact.
So FDA does not audit authority by reading the quality manual. It reads the dispositions. Here the unit released a lot filled in a room where a plate came back covered in growth and larvae, on a hypothesis the firm’s own procedures argued against. Ten weeks later the firm recalled that lot for lack of sterility assurance. The recall is what the release decision should have been.
- When did your quality unit last reject a lot? If every disposition this year has been release, the authority has never been tested. An unexercised veto is indistinguishable, on inspection, from an absent one.
- Does your risk assessment rest on a test result or on the conditions of manufacture? A passing test confirms a process you already know to be in control. When observed conditions contradict the test, the conditions are the finding — and more testing will not resolve it.
- Did the investigation conclude that no corrective action was required? Then say plainly why the deviation cannot recur. If that sentence can’t be written, the root cause was never established, and FDA will read the empty CAPA field as the proof.
A defensible response would have started from the conditions rather than the hypothesis: treat two same-day TNTC events in adjacent classified rooms as a contamination control failure until shown otherwise, test the infiltration theory against the firm’s own handling records, and reject the lot pending resolution. It would also have reopened every complaint closed without a root cause — because a unit that could not reject this lot has probably not been rejecting others.
Authority you never use is not authority. It is a signature block.
Veritas Quality Consultants works with pharmaceutical, medical device, and biologics manufacturers on Form 483 responses, root cause analysis, and CAPA development — built to FDA expectations inside the 15-day window.
Request a consultation