Available evidence versus objective evidence.

Every consequential quality decision sits on top of records that already exist. For a batch release, that typically means the batch manufacturing record, in-process and laboratory results, the certificate of analysis, any deviation report and the investigation behind it.

21 CFR 211.188(b) requires batch records to document that each significant step was accomplished, including in-process and laboratory control results, the identity of the people who performed and checked each significant step, and “any investigation made according to § 211.192” (eCFR, 21 CFR 211.188). That is a requirement to record what happened. It does not, by itself, show which of those records a decision owner weighed, or how.

That gap is what an investigator probes. The question is rarely “Do you have a batch record?” It is “What did the person who released this batch look at, and what did it tell them?”

What the evidence established, and what it didn’t.

The most useful discipline in a decision record is a plain split between what the evidence established and what the decision owner inferred from it. Here is that split from the evidence section of a worked batch-release record. The batch, people and dates are fictional, taken from the ComplianceWorxs Batch Release Authorization Case File.

The situation. Batch TAB-0626-118 was ready for release with deviation DEV-0626-044 open: an operator’s qualification had expired six days before he performed the blending step.

  • Qualification expired six days before execution
  • Blend time was 42 minutes, inside the validated range of 25 to 45
  • All in-process and final QC results met specification
  • No documented processing failure in the batch record
  • The lapse did not affect how the operator performed
  • No adverse effect on identity, strength, quality or purity
  • No pattern across prior batches; the lookback was still open

The evidence relied on was listed by name: the executed batch manufacturing record, the deviation report, product impact assessment PIA-0626-044 (signed by QA review the day before release), in-process and final QC results, and the operator’s training history showing 11 prior batches on the same operation in the preceding 12 months without a deviation.

Notice what the split does. It does not make the decision look stronger than it was. It shows a reviewer exactly where the judgment began. The release still rested partly on inference, and the record says so, which is what makes the rest of it credible.

Contemporaneous records versus retrospective documentation.

FDA’s Data Integrity and Compliance With Drug CGMP: Questions and Answers (December 2018) says complete, consistent and accurate data “should be attributable, legible, contemporaneously recorded, original or a true copy, and accurate (ALCOA).” Industry commonly extends this as ALCOA+, adding attributes such as complete, consistent, enduring and available. The guidance also states that data must be documented “at the time of performance.”

Evidence has to reach beyond one batch.

21 CFR 211.192 requires that an investigation of an unexplained discrepancy “extend to other batches of the same drug product and other drug products that may have been associated with the specific failure or discrepancy,” and that the written record include “the conclusions and followup” (eCFR, 21 CFR 211.192).

For the evidence lens, that means a decision record has to state the scope of what was examined. If the check on other batches was complete, the record should say what it found. If it was still open, as in the example above, the record should say that too, with an owner, a due date and the finding that would reopen the decision.

An inspection readiness checklist for evidence.

Pick one decision your team closed in the last six months: a batch release with a deviation, a CAPA effectiveness determination or an OOS investigation. Using only the record, check five things:

  1. Named evidence. Does the record list the specific documents the decision owner reviewed, or only reference the system they live in?
  2. Established versus inferred. Can a reviewer tell which conclusions the evidence proved and which it only supported?
  3. Timing. Is the evidence section dated with the decision, or was it assembled later?
  4. Scope. Does the record say whether other batches or products were examined, and what was found?
  5. Gaps. Does it state what was unknown at the time and who owned resolving it?

Where ComplianceWorxs fits.

ComplianceWorxs does not replace your QMS, your batch record system or your deviation and CAPA tools. Those systems hold the available evidence. The Inspection Response Record preserves the evidence one consequential decision actually relied on, alongside the judgment, the risk accepted and the authorization, in one reviewer-ready record. ComplianceWorxs does not make, approve or recommend regulated decisions. Your organization and its authorized people remain accountable.

For the documentation side of the same problem, read Good Documentation Practices Prove What Happened. Can Your Records Prove Why?

Next step: Check one completed decision in about three minutes. No documents, email or sales call required. Or see a completed record first.

Sources: eCFR, 21 CFR 211.188, Batch production and control records; eCFR, 21 CFR 211.192, Production record review; FDA, Data Integrity and Compliance With Drug CGMP: Questions and Answers, December 2018.