21 CFR 514.8: Build a Postapproval Change Record

A useful 21 CFR 514.8 change record answers a specific question: what supports distributing this product after this change? The answer should connect the approved application, the before-and-after conditions, the effects assessment, the reporting route, and the applicable FDA receipt or response. A completed engineering project or revised master record cannot answer that question alone.

This guide addresses selected US postapproval provisions dated October 20, 2026, for approved animal-drug applications. It is not an initial-approval manual, a worldwide clearance framework, or a substitute for product-specific regulatory judgment. Labeling-change pathways are outside this manufacturing-change workflow. Other industries use the generic concept of clearance, but their legal frameworks are outside this scope.

Under 514.8(b)(1)(i), the applicant must notify FDA of each change to an approved-application condition beyond variations already provided for, fully describing the change through the applicable supplement or annual-report route. Under (b)(1)(ii), the application holder must assess the effects before distributing drug made with a manufacturing change.

Start with the condition established in the approved application—not merely the latest purchasing specification or an engineer's description of current equipment. Preserve the approval reference, any applicable supplement, and the variation already allowed. Then compare the proposal with that baseline. If the team cannot locate the baseline, record an unresolved question rather than declaring the proposal equivalent.

Use a bounded statement: change one identified condition from a specified current state to a specified proposed state for identified products and sites. List connected changes separately. A new supplier, revised material specification, different manufacturing instruction, and new test method may be related, but a single project title should not conceal their individual effects.

A specification is more than its numerical limits. Section 514.8(a)(2)(iv) defines it as the approved quality standard comprising tests, analytical procedures, and acceptance criteria. Accordingly, compare all three elements. An unchanged acceptance range does not establish that the testing standard is unchanged when the method or required test has changed.

For the internal review, distinguish three dates: the proposed change date, the actual implementation date, and the first proposed distribution date for product made using the change. Keep them separate from submission and FDA receipt. This makes it possible to explain what occurred operationally without implying that implementation itself settled the distribution question.

Section 514.8(a)(2)(i) defines assessing effects as evaluating a manufacturing change's effects on identity, strength, quality, purity, and potency as those factors may relate to safety or effectiveness. Structure the assessment around those attributes rather than around departmental completion percentages.

As a practice, write the mechanism of potential effect before selecting supporting evidence. For example, a packaging-material change might prompt questions about product interaction, protection during storage, or the performance of the closure. These are assessment questions, not a universal testing panel or a predetermined reporting category. Select methods, studies, and acceptance criteria for the actual product and change.

• State the question each study answers and the affected product attribute.
• Identify the protocol or method, its revision, the comparison baseline, and the acceptance criteria used.
• Link the study report to the underlying data and identify the materials, equipment, or product represented.
• Record deviations, excluded data, conflicting observations, and limitations that affect interpretation.
• Separate completed evidence from planned work, ongoing stability commitments, and unresolved questions.
• Write a conclusion for each affected attribute, then explain how those conclusions support the overall effects assessment.

The assessment should distinguish absence of an observed adverse effect from absence of adverse-effect potential. The reporting categories in 514.8(b) use substantial, moderate, or minimal potential. A passing result is evidence for the assessment; it does not, by itself, choose the category. Conversely, a broad risk label without traceable supporting evidence is not a useful assessment conclusion.

Build a question-to-evidence matrix before drafting the submission narrative. Give each open question an owner and a consequence: does it block the assessment, require a document revision, or affect the proposed reporting route? When an answer changes, update the conclusion and identify which downstream decisions need reconsideration. Avoid leaving an early draft conclusion attached to a later, materially different data package.

Classify the actual change, not its project cost, vendor description, or local change-control label. The following map summarizes 514.8(b); it is not automatic route advice for a particular equipment, material, or process substitution. Record the applicable paragraph and the product-specific rationale.

A useful classification memorandum addresses alternatives. Explain why the proposed route fits, why a more burdensome route is not indicated by the evaluated facts, and whether a specific exception or category applies. Identify uncertainties that would change the conclusion. The memorandum should be reviewed against the final change scope—not an earlier, narrower proposal.

Treat Comparability Protocols as Change-Specific Evidence

Under 514.8(b)(2)(v), a comparability protocol describes specific tests, studies, and acceptance criteria for specified manufacturing changes. A protocol not already in the approved application, or a change to an approved protocol, requires a supplement approved before distribution of drug produced with the manufacturing change. An approved protocol supplement may justify a reduced reporting category for the particular change; it is not a blanket waiver.

Section 514.8(b)(1)(iii) directs use of an applicable regulation or guidance providing less burdensome notification. Record the particular authority and why it applies. A local procedure, a prior project with different facts, or the existence of a comparability protocol does not by itself establish that authority.

Create the index while evidence is being assembled. For PAS submissions other than the comparability-protocol exception, 514.8(b)(2)(iii)(A)–(J) identifies the required content. CBE30 and designated-CBE supplements incorporate that same information through (b)(3)(iv) and (vi). The index below maps those requirements to a practical retrieval structure; the folder names are recommendations.

Section 514.8(b)(1)(v) requires a list of all changes in the supplement or annual report; for supplements, it belongs in the cover letter. Under (b)(1)(iv), each major/moderate supplement and amendment needs a statement certifying that a field copy was provided to the appropriate FDA district office. No field copy is required for a supplement concerning a change to a drug manufactured outside the United States.

Preserve the exact submitted index separately from the live working index. An amended report may be the current internal version without having been included in the original submission. Record which version went to FDA, which amendment supplied a replacement, and which conclusions changed. A reviewer should not have to infer submission contents from today's folder.

The change record should connect proposed documentation to authorized operational use. Use a revision crosswalk showing the prior approved reference, proposed revision, final reviewed revision, effective date, and affected operations. Include unchanged documents when their continued suitability is important to the assessment, but explain that choice rather than generating unnecessary revisions.

The selected Part 211 provisions establish separate responsibilities. Section 211.22(c) assigns the quality control unit approval or rejection of procedures and specifications affecting identity, strength, quality, and purity. Section 211.100(a) requires appropriate organizational review and approval of production/process-control procedures and changes, including quality-unit review and approval. Under (b), execution is documented when performed, and deviations are recorded and justified.

Section 211.186(a) specifies master-record preparation, dating, and a full handwritten signature by one person, followed by independent checking, dating, and signing by a second. Section 211.192 separately requires quality-unit review and approval of production and control records before batch release or distribution, and investigation of unexplained discrepancies and specification failures. These are not replaced by a regulatory change-route memorandum.

Keep equipment readiness as another linked decision. As a practical distinction, validation concerns demonstrated capability under defined conditions; event verification concerns evidence for a particular cleaning or clearance event. Neither a study report nor a completed project checklist proves that a particular equipment-use event was acceptable. Link the equipment identity, applicable procedure, preceding use, event evidence, exceptions, and responsible review without prescribing a universal cleaning protocol or checker count.

For this workflow, missing or exception event evidence remains unresolved until responsible human reviewers approve the assessment and any necessary corrective work before next use. That internal readiness decision is distinct from product-batch release. Use the Part 211 readiness guide for deeper operational controls rather than duplicating them inside the application-change record.

The route-specific event matters. For a PAS, 514.8(b)(2)(iii) requires FDA approval before distribution and the label “Prior Approval Supplement.” For CBE30, (b)(3)(iii) requires the basis for the change, its date, and the label “Supplement-Changes Being Effected in 30 Days.” Under (iv), the minimum interval runs from FDA receipt—not the applicant's sending date.

Under 514.8(b)(3)(v), do not distribute drug made using the change if, within 30 days after receipt, FDA informs the applicant that PAS approval is required or that required information is missing. For missing information, the restriction continues until the supplement is amended to supply it. An elapsed calendar interval does not erase either notification.

For an agency-designated category under (b)(3)(vi), distribution may commence upon agency receipt of the qualifying supplement, labeled “Supplement-Changes Being Effected.” The full basis and change date are still required by (b)(3)(iii). Do not relabel an ordinary moderate change as designated CBE to remove the waiting interval.

If FDA later disapproves the supplemental application, (b)(3)(vii) allows the agency to order the manufacturer to cease distribution of drug made with the change. Keep correspondence review active after first distribution. As a practice, route new communications to the same decision owners and preserve any revised restrictions alongside the earlier authorization.

Section 514.8(a)(2)(iii) defines the Minor Changes and Stability Report as an annual report submitted once each year within 60 days before or after the application's original approval anniversary, or on a mutually agreed date. It includes minor manufacturing/control changes, or a statement that none were made, plus stability data from commercial or production batches generated according to an approved protocol or commitment.

For minor changes, 514.8(b)(4)(iii) calls for Form FDA 356V, the holder's effects-assessment statement, detailed descriptions, sites or areas, implementation dates, study/test data, applicable validation protocols or SOPs, appropriate revised and prior documentation, updated stability data, and other FDA-directed information. The report must be labeled “Minor Changes and Stability Report.”

• Maintain a running minor-change register with the application, change identifier, classification rationale, and actual implementation date.
• Give each reported change a document-index entry rather than relying on a year-end narrative alone.
• Track stability evidence separately so a year with no minor changes does not become a year with no stability-data review.
• Reconcile the final report against the register and retain the submitted version and submission evidence.

The MCSR is distinct from the periodic drug experience report in 514.80(b)(4). Keep separate report identifiers and calendars, even when the same team prepares both. Also keep the original approval anniversary separate from later supplement dates. The worked schedule below uses the original-anniversary anchor and ordinary calendar-date arithmetic.

Fictional Rowan Animal Health holds the assumed approved application ANADA-DEMO-R for Product R, a nonsterile oral solution identified for cats. The alias is not an application number. Change CH-R-026 replaces closure assembly CA-1 with CA-2 for one presentation at Site R. Formulation, labeled conditions, and the manufacturing process are outside the proposed change. No labeling change is proposed.

The fictional assessment records favorable reviewed conclusions for the evidence needed at this stage and identifies continuing stability work under the applicable commitment. It does not invent universal sample counts, migration limits, or test panels. The training question is whether the documented evidence and limitations support the stipulated decision—not whether the reader can derive a CBE30 classification from the words “closure replacement.”

During readiness review, manufacturing discovers that the first-use equipment verification attachment is missing from the controlled packet. A supervisor's message says the activity occurred, but the event evidence is not retrievable. The change record therefore shows two different states: regulatory interval pending or satisfied, and operational readiness unresolved. Neither state overwrites the other.

The team preserves the missing-evidence issue, determines what can be established from existing records, and documents the necessary corrective work and human assessment before next use. It does not backfill a passing event. The revised master and specification remain linked to this decision. Any product-batch disposition remains a separate quality-unit decision, not an inference from the change project's completion.

The exercise deliverable is a signed assessment package with a navigable index: baseline, differences, effects questions, study conclusions, document revisions, classification rationale, submission contents, receipt, correspondence review, readiness resolution, and distribution authorization. Another reviewer should be able to follow those links without asking the project author to explain what a filename means.

These are fictional calendar dates, including planned annual-report dates. The CBE30 calculation uses FDA receipt as the anchor and adds 30 calendar days; it is not a working-day reporting exercise. The main branch assumes no FDA notification under 514.8(b)(3)(v) during the 30-day interval. Alternative branches below change that assumption explicitly.

Test the Notification Branches

Alternative A: FDA notifies the applicant on September 22 that required study data are missing. October 2 does not permit distribution while that omission remains. The fictional amendment supplies the information on October 5. Record the amendment and reassess the applicable conditions; do not treat the calendar cell as automatic authorization or invent an automatic clock-reset rule.

Alternative B: the September 22 notification says the change requires PAS approval. The team cannot use October 2 as its distribution date. It needs the applicable approved supplement before distributing product made with the change. No fictional approval date is supplied or inferred.

For the annual exercise, CH-R-026 remains indexed to its CBE30 supplement. A separately assessed minor change, CH-R-009, enters the MCSR register with its own evidence and implementation date. Include the applicable stability data; do not recategorize the CBE30 change as minor merely because an annual filing is approaching.

Run a retrieval exercise from a decision, not from a known folder. Give a reviewer unfamiliar with CH-R-026 this prompt: “Explain why Product R made using CA-2 was internally authorized for distribution on October 7, and why October 1 and October 2 were not authorized dates in the main scenario.” The exercise tests evidence connections, not memory or presentation quality.

1. Find the application/product assignment and retrieve the assumed approved closure baseline. Identify the exact condition changed.
2. Open the before-and-after comparison. Confirm that the final submitted scope matches the assessed scope.
3. Trace each material effects conclusion to its study report, supporting data location, limitations, and review record.
4. Retrieve the classification memorandum and explain why the exercise uses CBE30 rather than designated CBE or a protocol-based reduction.
5. Open the submitted index and receipt evidence. Recalculate September 2 plus 30 days without using the sending date.
6. Locate the agency-correspondence review and distinguish the main scenario from the two alternative notification branches.
7. Retrieve the missing-event-evidence issue, its assessment, corrective work, and the readiness decision before next use.
8. Open the October 7 authorization and identify its scope, decision-makers, supporting quality decisions, and continuing conditions.

Pass the exercise only when an independent reviewer can retrieve the underlying records, reproduce the date arithmetic, and explain the remaining conditions. Record broken links and ambiguous decisions as findings with owners. A polished summary is not a substitute for accessible evidence.

Keep historical decisions intact when later evidence arrives. Add the new assessment and its consequences instead of editing the old authorization to appear more informed than it was. This preserves the distinction between what was known at first distribution and what the organization learned afterward.

This change record ends with a documented regulatory and operational decision, not with a duplicate lot packet. Hand the effective master/specification references, change identifier, assessment conclusions, and distribution restrictions to the veterinary manufacturing lot-packet workflow. Leave lot quantities, execution accounting, and shipment reconciliation there.

Keep the separate experience-reporting handoff open through the 514.80 guide. Its periodic drug experience report is not the MCSR. When a complaint, defect, or new study raises a question about the change, send the original information and affected change references to the responsible quality and regulatory reviewers. Do not wait for a change record to be administratively closed before making that internal handoff, and do not import working-day alert calculations into the receipt-based calendar exercise above.