QA Disposition Step: Release, Reject or Hold Decisions

QA Disposition Step is the governed point in a material or batch workflow where authorised quality personnel record a decision to approve, reject or retain identified material on hold or in quarantine. This is operational terminology, not a universal regulatory definition or prescribed set of status labels.

The decision object may be a component lot, specified containers within a lot, in-process material or a product batch. A useful disposition record identifies both the material scope and the intended next use. “Approved” alone is incomplete operationally: approval of a component for manufacture answers a different question from approval of a finished product for distribution. The scope should make clear which material the decision covers, without silently extending it to related inventory.

The following comparison is an operational aid, not a universal legal taxonomy. Disposition can occur at several material stages; finished-batch release concerns the final release process for a finished batch.

An exception record documents the issue; a disposition record documents the resulting material decision. The distinction between nonconformance versus deviation helps classify the exception, but neither label establishes whether affected material is acceptable. Closing an exception should not be treated as automatic approval of that inventory.

Likewise, completing a test or collecting a review signature is an input or activity, not necessarily the disposition itself. The gate is identifiable when an authorised person records what decision applies to which material and for what use.

Map the disposition gate to the product, jurisdiction and decision being made before assigning approvers. A useful site mapping separates regulatory authority from workflow configuration: who has decision authority is a different question from how a review task reaches that person. The following sources address different scopes; they are not interchangeable templates.

US drug products: authority and acceptance criteria

Under 21 CFR 211.22(a), the quality control unit has responsibility and authority to approve or reject components, drug product containers, closures, in-process materials, packaging material, labeling and drug products. The provision therefore reaches beyond a finished-batch decision. See the broader 21 CFR 211 framework for the drug CGMP context.

21 CFR 211.165(d) requires sampling and testing acceptance criteria adequate to assure that drug-product batches meet each appropriate specification and appropriate statistical quality-control criteria as a condition of approval and release. Those statistical criteria must include appropriate acceptance levels and/or rejection levels. This is not a universal test panel or a single numerical acceptance limit.

US supplements: product-specific disposition controls

Treat supplement material review and disposition within the dietary supplement CGMP framework, rather than importing the drug-product approval model. For packaged and labeled dietary supplements rejected for distribution, 21 CFR 111.425 requires clear identification, holding and control under a quarantine system for appropriate disposition. This particular provision addresses that rejected-product condition, not every supplement review decision.

The central supplement gate is the specification-failure rule. Under 21 CFR 111.77(a), for the listed specifications that are not met, quality control personnel “must reject the component, dietary supplement, package or label unless such personnel approve a treatment, an in-process adjustment, or reprocessing that will ensure the quality of the finished dietary supplement and that the dietary supplement is packaged and labeled as specified in the master manufacturing record.” The same paragraph adds: “No finished batch of dietary supplements may be released for distribution unless it complies with § 111.123(b).” Paragraphs (b) and (c) are narrower and offer no alternative: a component failing a §111.70(b)(1) specification must be rejected and “must not be used in manufacturing”, and received product failing a §111.70(f) specification “may not be packaged or labeled for distribution as a dietary supplement.” A disposition record for a supplement site should therefore name which paragraph applies, because that determines whether a treatment, adjustment or reprocessing option exists at all.

Returns connect directly to the same gate. 21 CFR 111.510 states: “You must identify and quarantine returned dietary supplements until quality control personnel conduct a material review and make a disposition decision.” Section 111.515 then requires destruction or other suitable disposal unless that review results in quality control personnel approving salvage for redistribution or approving reprocessing. In workflow terms, the returned-goods status should block use until a recorded material review and disposition exist, and the only non-disposal outcomes the record should offer are those two approvals.

Corrective work does not end the decision. 21 CFR 111.90(c) states that any batch “that is reprocessed, that contains components that you have treated, or to which you have made in-process adjustments to make them suitable for use in the manufacture of the dietary supplement must be approved by quality control personnel and comply with § 111.123(b) before releasing for distribution.” A supplement disposition workflow therefore needs two separate recorded approvals when reprocessing is chosen: one authorising the treatment, adjustment or reprocessing under §111.77(a), and a later one approving the resulting batch for distribution. Recording only the first approval and letting inventory release follow automatically would collapse those two decisions into one, which is exactly the shortcut the regulation’s wording rules out.

EU certification and the boundary of this mapping

Annex 16, General principles, batch-release process point ii, is EU GMP guidance describing certification of the finished-product batch by a Qualified Person (QP), signifying compliance with GMP and the requirements of its marketing authorisation. QP certification is a specific role within finished-batch release, not a synonym for every QA disposition decision. Keep that distinction explicit when naming approval steps.

As a site-designed practice, document routing, escalation and status labels separately from the regulatory mapping. A local workflow name should describe the decision without implying a different authority or broader scope.

Check the current text of 21 CFR 211.89 for rejected-material controls and EU GMP Chapter 1 for the applicable quality-system context. For food, devices and chemicals, use the applicable product and jurisdiction framework; do not extend the provisions above by analogy.

A practical, site-designed workflow starts with the decision scope, not a presumed outcome. Illustrative triggers include receipt, a production checkpoint, packaging completion, returned goods or an exception signal. Each changes the evidence checklist: a goods receipt note may support receipt review; production records support checkpoint review; packaging reconciliation supports completion review; return history supports returned-goods assessment; and out-of-specification handling informs an analytical exception. The trigger alone does not determine approval or rejection.

1. Identify scope: name the material, lot, quantity and intended use covered by this decision.
2. Collect evidence: apply the trigger-specific checklist and identify contributors responsible for each item.
3. Resolve questions: assign an investigation owner for disposition-relevant uncertainties. Return missing evidence to its owner and record the specific decision blocker. Track longer-term corrective actions separately, subject to applicable procedures.
4. Assign an authorised reviewer: confirm that the disposition authority covers this material and decision type.
5. Record the decision: have the authorised person document the outcome, rationale and scope.
6. Confirm execution: have the inventory execution owner check that downstream restrictions or permissions match the recorded decision and report unresolved execution failures.

This responsibility model distinguishes evidence contributors, the investigation owner, disposition authority and inventory execution owner; it does not assume four different people. Assignments and any separation of duties should follow applicable procedures and product context.

Define entry criteria such as identifiable scope and named owners, and exit criteria such as a recorded decision and confirmed execution. Urgency flags, reminders and reviewer escalation can prioritise work or locate an authorised alternate; they should never confer approval authority. Actual approval or rejection remains with authorised people.

Use four separate data dimensions: decision outcome records the authorised conclusion; reason code classifies why; inventory restriction defines blocked or permitted actions; physical location identifies where material sits. This proposed site model is an information-design recommendation, not a regulatory status list.

Here, “required restriction” means a control required by this example model. Define quarantine status separately as a control state where appropriate. Hold and quarantine are not universally interchangeable workflow labels: the site model must state their meanings without overriding product-specific rules. A location named “quarantine” cannot alone identify the decision, its scope or its authority; moving material does not record those facts.

Illustrative reason-code categories include incomplete evidence, specification failure, damaged packaging and corrected laboratory data. Pair a controlled disposition code with a case-specific rationale explaining what happened and why the outcome follows. A code is not the entire decision, and corrected data is not itself approval.

Scope every transition to the identified lot or containers, specified use and decision version. With container status tracking, an authorised change can apply to selected containers without silently changing the remainder. Keep the finished-batch boundary established above intact.

For drug products subject to Part 211, 21 CFR 211.192, first sentence requires quality control unit review and approval of all production and control records, including packaging and labeling records, for compliance with established, approved written procedures before release or distribution. The batch record therefore belongs in the review, not merely the laboratory result.

Under the provision’s second sentence, a batch or component specification failure must be thoroughly investigated even if the batch has already been distributed. Connect out-of-spec handling to the disposition evidence rather than treating a status change as the investigation itself.

The third sentence extends the investigation to other batches of the same drug product and other drug products potentially associated with the failure or discrepancy. The fourth sentence requires a written investigation record containing conclusions and followup.

A practical, versioned decision model

The following is a proposed traceability design, not a list of individually prescribed regulatory fields. Use it within an electronic release record or another controlled record:

• Scope: material identity, affected lots or sublots, quantity, unit and intended use.
• Basis: evidence references and versions, applicable criteria, related exceptions and investigation conclusions.
• Decision: outcome, rationale, authorised decision maker and decision time.
• Execution: references to downstream transactions and their acknowledgements.

Data-model diagram: Evidence A v1 + Evidence B v3 → Decision D1 ← Exception X + Exception Y; Decision D1 → Execution references.

One decision can reference several evidence records and exceptions; each reference should identify the version actually reviewed. Record why the evidence supports the stated material scope, including how associated-batch findings affect that scope.

If Evidence A is corrected to v2, preserve D1’s link to v1. Open a controlled reassessment that references v2, explains the correction’s impact and records the authorised conclusion. Link any subsequent decision to D1 rather than silently replacing its basis. This preserves both what was known originally and why the later conclusion changed—or remained unchanged.

Use a documented integration contract as a recommended design pattern, not a universal regulatory workflow. Laboratory systems provide identified results and their versions through laboratory integration; the decision record captures the authorised human outcome and evidence references. ERP and warehouse systems apply the corresponding inventory restrictions only to the specified material scope.

Keep the quality decision separate from its execution state. A recorded approval can have propagation pending, acknowledged or failed. An unavailable interface should not rewrite that approval as a rejection; nor should approval alone imply that inventory is available. Track execution separately for each target system.

For ERP status updates, include decision_id, decision_version, material_scope, expected_previous_status, intended_status and event_id. Define material_scope using lot and container identifiers, quantities and units. Use event_id for idempotency: recognise a repeated message without applying its inventory effect twice. Check decision_version and expected_previous_status before applying an update so delayed messages cannot silently overwrite newer decisions.

1. Preserve the authoritative decision and record the execution failure against it.
2. Identify the failed target, affected containers and unapplied quantity.
3. Maintain the relevant inventory restriction while execution remains unresolved.
4. Retry safely using the same event identity for the same intended effect; route conflicting updates for review.
5. Reconcile acknowledgements against the entire decision scope, distinguishing applied, outstanding and disputed quantities.

In warehouse integration, treat partial quantities and split containers explicitly: preserve parent–child identity and account for each quantity without duplication. Define whether partial application is permitted and how remaining stock stays restricted. Competing updates need conflict handling, not last-message-wins behaviour. Retain message and acknowledgement history for recovery; detailed challenge tests belong in the validation section.

All three cases are illustrative. Quantities, outcomes and recovery steps are declared assumptions—not regulatory thresholds, customer results or permissions applicable to every site.

Case 1 — Illustrative container-level rejection

Assume 100 incoming containers contain 10 kg each; seven have assessed damage. The applicable procedure permits separately identified container-level disposition after quality review. Total quantity is 100 × 10 = 1,000 kg; affected quantity is 7 × 10 = 70 kg; the other 93 containers contain 930 kg.

Subtracting 70 kg establishes a quantity balance, not the acceptability of 930 kg. Assume receipt photographs, container identifiers, handling history and documented inspections establish the damage boundary; the assessment addresses possible shared exposure, and applicable sampling and test results support acceptance of the other containers. Without that basis, the proposed split remains unresolved. Use the evidence model described above.

Closing record: scope—seven identified containers versus 93 separately identified containers; evidence—approved damage assessment and acceptance results; authorised outcome—reject 70 kg and approve 930 kg only after review supports both decisions; execution—record separate rejected and approved container-status changes, preserving their parent-lot relationship through location-level genealogy.

Case 2 — Illustrative supplement-batch rejection

Assume a packaged dietary-supplement batch comprises eight pallets of 1,500 bottles each: 8 × 1,500 = 12,000 bottles. A labeling issue prompts review, and authorised quality personnel reject the identified batch. Apply the supplement control described in the regulatory mapping section.

Assume the pallet register and physical reconciliation account for all eight pallets, with no bottles transferred or missing. Record each pallet’s identity, quantity, location and quarantine status against the rejection decision. A proposed relabeling action opens a separate assessment; it neither overwrites the rejection nor establishes permission to proceed.

Closing record: scope—all 12,000 bottles; evidence—labeling assessment, batch record and eight-pallet reconciliation; authorised outcome—batch rejected, with any relabeling proposal assessed separately; execution—all eight pallets accounted for under the rejection control, with the original decision retained.

Case 3 — Illustrative recovery after partial execution

Assume an authorised approval covers 6,000 packs. Downstream acknowledgements account for 4,000 packs, leaving 6,000 − 4,000 = 2,000 awaiting confirmed execution. The decision remains approved, while its execution state remains unresolved. Assume the site procedure blocks shipment of all 6,000 packs during reconciliation. Follow the integration controls described above.

Reconcile transaction identifiers before retrying the outstanding update. Assume recovery produces acknowledgements for the remaining 2,000 packs without duplicating the first 4,000. Retrying delivery executes the existing decision; it is not a new quality approval.

Closing record: scope—6,000 packs; evidence—original approval and transaction reconciliation; authorised outcome—unchanged approval; execution—acknowledgements cover 4,000 + 2,000 = 6,000 packs, with recovery history retained and the temporary restriction removed under the assumed procedure.

For US drug products, 21 CFR 211.204 requires returned products to be identified as returned and held. In the exception workflow, keep receipt of a return separate from any decision about its subsequent use.

For dietary supplements governed by 21 CFR 111, the trigger extends beyond the returned batch: if the reason for return implicates other batches, 21 CFR 111.530 requires investigation of manufacturing processes and each implicated batch to determine compliance with specifications.

As a workflow practice, treat rework, reprocessing and relabeling as proposed operations for assessment under the applicable product rules—not as synonyms for an approved disposition. Record the proposed operation, affected material and assessment owner before routing it for review. None of these labels should be treated as general permission to proceed.

Replace an ambiguous conditional-release label with four explicit entries. For example, a proposed return-assessment record could specify:

• Permitted action: inspect the returned units.
• Restricted action: no distribution.
• Unresolved condition: assessment of the reported storage excursion.
• Responsible reviewer: the named authorised quality reviewer.

This recommended structure makes the scope of the decision readable without implying that permission to inspect also authorises reprocessing or release.

Use risk-oriented workflow challenges as a practical test approach, not a complete validation standard. Test both permitted and prohibited transitions, linking each case to its intended control through requirements traceability. Include scope selection, reviewer permissions, evidence-version binding and preservation of superseded decisions.

For each case, record expected and observed quantities, units and statuses, including unaffected containers. An interface displaying “approved” is not an end-to-end result: inspect whether the intended material scope has the correct restrictions in every relevant system. Investigate mismatches using a CSV deviation report. These challenges alone do not establish compliance with Part 11, Annex 11 or any validation framework.

For a proposed dashboard, use elapsed calendar hours and timestamps from the process event log. Track three intervals:

• Review queue: evidence marked ready to first recorded review start.
• Decision review: review start to authorised decision recorded.
• Execution: recorded decision to acknowledgement by every required downstream system.

For each interval’s mean, divide total completed hours by cases reaching that interval’s end event during the reporting period; count each case once per interval. Report unfinished-interval ageing separately, from its start event to the reporting cutoff—not within completed-time averages.

Example delay categories include missing laboratory evidence, investigation dependency, reviewer availability and interface recovery; these are not mandatory codes. Use trends as prompts for management review, not pressure to change scientific conclusions. Neither approval percentage nor a low hold count should be a standalone quality target.

Use this practical review to check workflow coherence, not as a universal inspection checklist or a promise of compliance. Refer to the earlier sections for the underlying controls.

• Confirm product and jurisdiction scope, decision objects and authorised roles.
• Check entry criteria, outcome definitions and exception routing.
• Identify evidence versions and reconcile quantities against the relevant batch execution history.
• Assign recovery ownership and define observable test acceptance criteria.

Walk through three records

Select one approved, one rejected and one unresolved record. Ask a reviewer to reconstruct what was decided, for which material and quantity, on what evidence and by whom. Then compare the intended inventory effect with the actual result. Use the audit trail review workflow to follow changes. For the unresolved record, identify the outstanding question and its owner.

What should a contract-manufacturer handoff contain?

Agree a handoff record identifying material, decision scope, evidence references, unresolved issues and receiving responsibility. This is recommended practice, not a universal document package.

Can EU medicinal products move before QP certification?

EU GMP guidance in Annex 16, paragraph 4.1 distinguishes sale or supply from quarantine shipment. Batches should only be released for sale or supply after QP certification. Until certification, they should remain at the manufacturing site or be shipped under quarantine to another site approved for that purpose by the relevant Competent Authority.