Two facilities. Same contract manufacturer. Same product line — oral solid dosage, film-coated tablets, commercial batch. Same atypical result: a viscosity reading during the coating process that fell within specification but deviated from the historical process mean by a margin that triggered the site's exception review protocol.
In both cases, the batch record flagged the exception automatically. In both cases, the Qualified Person was notified. In both cases, the QP reviewed the exception, evaluated the data, determined that the batch met all release criteria, and authorized release.
Both QPs signed the batch record. Both batches shipped to distribution. Both had compliant batches by every measurable quality standard.
Eighteen months later, both facilities were inspected by FDA.
The first QP had done everything the procedure required. The exception was captured in the batch record. The QP signature was present. The batch record review section was complete. Every checkpoint was satisfied.
From a documentation standpoint, the file was closed. The batch record was a record of the batch — and the QP's signature was a record that the QP had reviewed it. That's the document. That's what exists.
Most QA Directors in this situation would say: the file is complete. The exception is documented. The signature demonstrates review and authorization. There is nothing missing.
That belief is the exposure.
The investigator at the first facility didn't question the batch specification. He didn't challenge whether the atypical result was a genuine deviation. He didn't ask to see the QP's credentials.
"Who authorized the release of this batch, and what specific evidence did they review in making that authorization decision?"
The QP was in the room. The QP pointed to the signature. The investigator noted the signature. Then he asked the question again, differently:
"What document captures the authorization decision itself — the evidence reviewed, the regulatory standard applied, the alternatives considered, and the rationale for concluding this batch was suitable for release?"
There was no such document.
The signature proved the QP reviewed something. The batch record proved an exception existed. Neither document — alone or together — constituted an authorization record.
- The specific evidence the QP reviewed in evaluating the exception
- The regulatory standard that governed the release decision (not the specification — the decision standard)
- The alternative interpretations of the atypical result that were considered and rejected
- The documented rationale for concluding the batch met release criteria despite the deviation from process mean
- The identity of the authorization — not just who signed, but who made the decision and under what authority
The QP remembered making all of these evaluations. The reasoning was sound. The conclusion was correct. None of it existed in the file at the time the investigator asked.
A signature is not an authorization record. It is evidence that someone reviewed a document. The authorization record captures the decision itself — frozen at the moment it was made.
"There is no written record of the investigation sufficient to justify the conclusion that the atypical result was not caused by a production error, nor sufficient to justify that the batch was suitable for distribution."
The second facility received no observations related to this batch. Not because their QP made a better decision. Not because their process was different. Not because their specification was stricter.
The second QP had created a separate document at the time of the release decision — not a replacement for the batch record, but a companion to it. A document that captured the authorization: what evidence was reviewed, which regulatory standard governed the decision, what interpretation alternatives existed, and what rationale supported the conclusion.
When the investigator asked the question, the second QP handed him the document. The investigator reviewed it. No observation was issued.
The decision was the same. The outcome was the same. The authorization record made the difference.
The investigator's question at the first facility was not a trap. It was not an ambush. It was the standard question that any FDA investigator will ask when an exception exists in a batch record: show me the authorization decision, not just the authorization signature.
Every batch release with an exception is a decision moment. Every decision moment requires an authorization record. Not a summary written after the inspection is announced — a record produced at the moment of decision, when the evidence is present and the rationale is live.
The authorization record doesn't replace your existing documentation. It exists alongside it. It answers the one question that batch records are not designed to answer: who decided, what did they review, and why was this the right call.
That record is producible in advance. It takes minutes to create at the time of decision. And when the investigator asks — and they will ask — it is the difference between a 483 observation and a clean inspection.
Discussion opens when you join. Login with LinkedIn to participate.
Join to Comment →