SOP: Batch Record Review

This is a ready-to-use SOP for the formal quality assurance review of an executed batch record before disposition. Replace every <<FILL: ...>> placeholder with your own specifics, set your document numbers and dates, and route it through your normal document control, review, and approval. A worked filled specimen follows the template so you can see how a completed review reads. Verify each cited regulation against the current source before you rely on it. This is general guidance to adapt to your own quality system, not legal or regulatory advice.

Document control header

1. Purpose

This procedure defines how <<FILL: COMPANY NAME>> performs the formal quality assurance review of an executed batch record before a batch proceeds to disposition, so that every batch is reconstructed from a complete, attributable record before it is released or distributed. It sets the intake gate that keeps an incomplete record out of the review queue, the sequence a reviewer follows, and the escalation path when a finding is significant enough to hold the batch.

2. Scope

This procedure applies to the executed batch production and control record, including packaging and labeling records, for every batch manufactured or packaged at the sites in the header, whether the record is paper, hybrid, or a full electronic batch record. It covers the review activity itself. It does not define the deviation process (<<FILL: SOP-ID for deviations>>), the out-of-specification investigation process (<<FILL: SOP-ID for OOS>>), review by exception design (<<FILL: SOP-ID for review by exception>>), or the disposition decision itself (<<FILL: SOP-ID for batch disposition>>); this procedure feeds those and consumes their outputs.

3. Responsibilities

4. Definitions

• Executed batch record (EBR): the record documenting what actually happened for one specific batch, whether paper, hybrid, or electronic; not to be confused with the master batch record, which defines what should happen.
• Unreviewable record: an executed batch record that cannot be meaningfully reviewed in its current state because it fails one or more of the criteria in section 5.1, and must be returned for correction before the review clock starts.
• Discrepancy: any difference between the executed record and the master record, an established written tolerance, or another record it should match. Every discrepancy resolves to either a documented non-issue within a written tolerance, a match to an existing approved deviation, or a new deviation; there is no fourth outcome.
• Critical step: a step the master record designates as requiring a second-person check because an error there could affect product quality, patient safety, or the reliability of the record.

5. Procedure

5.1 Screen for reviewability before the review clock starts

Before beginning the substantive review, confirm the record does not meet any of the following unreviewable criteria. If any is present, return the record for correction and log the return; do not begin the timed review.

1. Missing attribution on a critical step (no operator name or ID).
2. Missing second check on a critical step that requires one.
3. A correction made incorrectly (white-out, obliteration, illegible original, no reason where required).
4. A deviation referenced in the record with no deviation record on file.
5. An unexplained blank in a required field.
6. A value outside specification with no deviation or explanation.

Use the batch record completeness and review readiness checklist to apply this screen consistently, and log every return with the batch record return-for-correction log. A record returned twice for the same defect is escalated to QA management as a process signal, not just corrected again.

5.2 Confirm the record and the master version

Confirm the batch number, product, strength, and any holds or open events flagged at intake. Confirm the master record version referenced is the approved current version in effect at the time of manufacture, not a superseded one.

5.3 Review in sequence

Perform the review in the following order, because deviations and material accountability are the most likely to stop the batch and should be resolved before spending time on the rest of the record:

1. Deviations. Confirm every deviation referenced in the record is formally captured, root-caused, impact-assessed, and closed, or the batch is explicitly held pending closure. Question a zero-deviation result on a complex batch rather than accepting it at face value.
2. Material accountability. Trace every active ingredient and excipient lot from dispensing through use, confirming lot numbers match across the dispensing record, the materials issue log, and the batch record.
3. Yield and reconciliation. Confirm yield is reconciled at each defined stage against the validated range, using the yield reconciliation worksheet. Investigate unexpected loss, consistent loss at the same step, and yield that is suspiciously perfect.
4. In-process results in sequence. Confirm values are within specification and make physical sense in time order; for automated data capture, compare the electronic process data to the manual entries and require an explanation for any mismatch.
5. Corrections and audit trail. For paper entries, confirm every correction follows good documentation practice. For electronic entries, review the audit trail for changes after batch completion, changes by uninvolved users, changes to critical parameters, and bulk changes.
6. Attribution and the performer-verifier rule. Confirm every critical step carrying a second-check requirement has two different, qualified signatures.
7. Release testing package. Confirm every required test was performed, results pass against the current approved specification, and the Certificate of Analysis references match the batch number, strength, and specification.

5.4 Resolve every discrepancy to a defined outcome

For every discrepancy found during the review, apply the decision in <<FILL: reference: the discrepancy-to-deviation decision logic in the parent article or an internal job aid>>: confirm it is within a written tolerance, confirm it matches an existing approved deviation exactly, or open a new deviation now. Do not proceed on an assumption that an unresolved discrepancy will turn out fine.

5.5 Escalate significant findings

Escalate to QA management the same working day when a finding involves: a discrepancy with no plausible authorized explanation, a suspected data integrity issue (an unexplained audit trail gap, a correction that appears to conceal rather than correct, values that are suspiciously perfect), or a record returned for the same unreviewable defect a second time. Do not proceed to disposition recommendation until the escalation is resolved.

5.6 Document the review

Complete the review record (section 8) for every batch reviewed, including batches with no findings. The record states what was checked, what was found, and how each finding was resolved, so it stands alone for a disposition authority or an inspector reading it later without asking the reviewer.

5.7 Time standard and aging

Target a review cycle time of <<FILL: e.g. 3 business days>> from the record being marked reviewable to review completion, measured as median and percent within target, not mean alone. A record aging beyond <<FILL: e.g. 5 business days>> without review completion is escalated to QA management with the reason for the delay.

6. Acceptance criteria

A batch record review is acceptable when all of the following are true:

• The record passed the unreviewable-criteria screen, or was returned and corrected, before the timed review began.
• Every required field is present, legible, and attributable to a single named individual.
• Every discrepancy resolved to a documented non-issue, a matched approved deviation, or a new deviation; none was left unresolved.
• Yield and material reconciliation fall within the validated range, or a variance is explained and investigated.
• The audit trail (electronic) or corrections (paper) show no unexplained post-completion change, no change by an uninvolved user, and no improper correction.
• Release testing is complete and in specification, and the Certificate of Analysis matches the record.
• The review itself is documented: who reviewed, when, what was found, and how each finding closed.
• Any significant finding was escalated per section 5.5 before the review was signed complete.

7. Records generated

Batch record review record (section 8); returns for correction logged per the return-for-correction log; escalations and their resolution; cycle-time data feeding the metrics program.

8. Record generated: batch record review record

9. References

21 CFR 211.188 (batch production and control records) and 211.186 (master production and control records). 21 CFR 211.192 (production record review by the quality control unit before release, and investigation of unexplained discrepancies). 21 CFR 211.103 (yield calculation). 21 CFR 211.22 (responsibilities of the quality control unit). EU GMP Volume 4, Chapter 4 (documentation) and Annex 16 (certification by a Qualified Person and batch release). ICH Q7, Sections 6.5 and 6.7 (production record review for active pharmaceutical ingredients). ICH Q9 (quality risk management, for the risk-based time standards and escalation criteria).

Confirm the current version and clause numbers of each reference before issue.

10. Revision history

11. Approvals

Filled specimen

The following shows the review record completed for an illustrative solid-dose batch. The company, batch, and numbers are illustrative; replace them with your own.

The reviewer confirmed the one open item, the hold-time deviation, had a specific and adequate impact assessment before moving on, rather than treating “a deviation exists” as sufficient by itself. That is what a real review looks like on a clean batch: not the absence of anything to check, but confirmation that everything that needed checking was checked.

Common inspection findings this SOP prevents

• A batch record review signed within minutes with no evidence of meaningful examination.
• A record accepted into the review queue despite missing attribution, a missing second check, or an unresolved discrepancy, none of which the unreviewable-criteria screen was ever applied to catch.
• A discrepancy noted in the record with no documented resolution, neither matched to an approved deviation nor opened as a new one.
• Review cycle time reported only as a mean, hiding a tail of aging records.
• A significant finding, such as a suspected data integrity issue, handled by the reviewer alone with no escalation to QA management.

How to adapt this SOP

1. Set your document number, owner, and effective date in the header.
2. Point the cross-references in sections 2, 5.1, 5.3, 5.4, and 5.5 to your real deviation, review by exception, disposition, and calibration procedures.
3. Set your own cycle-time target and aging-escalation trigger in section 5.7, based on your product complexity and current baseline.
4. If your site does not use a manufacturing-supervisor first-pass check, remove that row from section 3 and state the review starts directly with QA.
5. Confirm every regulation in section 9 against the current published version before issue.