Do not close a bulk tea supplier corrective action when the supplier sends a revised procedure or a passing retest. Close it only when objective evidence shows that the defined cause was addressed, the action was implemented across the affected scope, the relevant control is operating as intended, and recurrence has not appeared during a risk-based verification period. Product disposition and supplier-system effectiveness are related decisions, but they are not the same decision.
Separate four decisions that buyers often combine
A response labelled “CAPA completed” can conceal four decisions. Correction fixes an observed condition. Containment and disposition control the affected tea and decide its fate. Corrective action addresses the cause to reduce recurrence. Effectiveness verification checks later evidence to determine whether that action worked.
The buyer should keep the affected lot under the applicable nonconforming-lot disposition plan. Closing the supplier action does not retrospectively release that lot, and releasing or returning the lot does not prove that the supplier's process has improved.
Build one corrective-action review sheet
| Control field | Buyer should require | Closure question |
|---|---|---|
| Trigger and requirement | Complaint, audit finding, test, inspection or document gap linked to the exact legal, specification or contract requirement | Is the failure stated without guessing its cause? |
| Affected scope | Products, lots, dates, lines, suppliers, packing formats, shipments and customers considered | Does evidence support the boundary? |
| Immediate control | Hold, stop, correction, notification and disposition records with quantity reconciliation | Was exposure controlled while the cause was investigated? |
| Cause evidence | Data, interviews, records, measurements and tests supporting the direct, contributing and system causes | Would removing the stated cause prevent the same failure path? |
| Action plan | Action, owner, due date, affected documents/equipment/people and change-control needs | Does every material cause have a matching action? |
| Implementation proof | Approved revisions, training competence, maintenance, calibration, supplier controls and dated operating records | Was the action used in practice, not merely issued? |
| Effectiveness test | Measure, sample or record set, period, frequency, acceptance rule, reviewer and escalation path | Could this test detect recurrence? |
| Buyer decision | Open, conditionally approved, closed, reopened or supplier reapproval required, with authority and date | Is the decision proportionate to residual risk? |
Use an eight-step effectiveness check
1. Anchor the case to evidence and a requirement
Quote the failed limit, approved sample attribute, procedure, label text or contract clause. Record the observed result, method, unit, date and lot. “Quality problem” is not a usable problem statement. Review the COA evidence boundary before treating one report as proof about a whole shipment.
2. Challenge the affected scope
Ask what changed before the event and what shared the same input, equipment, operator, method, packaging, artwork, storage zone or period. Compare quantities and records in both directions. If scope depends on a lot boundary, run the traceability and quantity-reconciliation check. A narrow scope needs positive evidence.
3. Verify containment separately
Confirm that identified stock was blocked physically and electronically, customers or authorities were notified where required, and every kilogram was accounted for. A supplier may need prompt containment before root-cause work is complete. Do not accept a root-cause discussion as a substitute for controlling potentially affected tea.
4. Test the cause statement against the facts
“Operator error” merely renames the event unless the supplier explains why the step was missed and why the system failed to prevent or detect it. Look for evidence of direct and enabling causes: unclear limits, wrong instrument range, uncontrolled documents, maintenance failure, supplier change, weak sampling or missing review.
5. Map each action to a cause
Retraining can be suitable when knowledge or competence is actually causal. It is weak when the real cause is an impossible instruction, unsuitable equipment, a missing approval gate or an uncontrolled input. Require an owner and completion date for each action, and identify consequential changes to specifications, hazard analysis, sampling, labels, calibration, maintenance, purchasing or records. Use the supplier change-control plan when the action changes an approved process, material or site.
6. Define effectiveness before reviewing the result
Prewrite the evidence, number of lots or period, locations and reopening trigger. The window should reflect frequency and risk: five consecutive packing lots may test a seal-control issue better than a month with no production. One passing sample cannot demonstrate control of an intermittent failure.
7. Verify implementation with operating evidence
A signed procedure shows authorisation, not use. Sample later records, observe the activity where risk warrants, interview staff and verify equipment or software settings. For analytical controls, confirm matrix, method, laboratory scope, sampling frame and decision rule using the guide to selecting a laboratory for bulk tea.
8. Make and record the buyer decision
Close only when actions and effectiveness evidence meet the prewritten rule. Conditional approval can pair limited orders with enhanced inspection or testing when lawful and justified. Reopen the case when recurrence, missed due dates, unsupported scope or changed evidence invalidates the conclusion. Serious or repeated failures may require supplier reapproval, an onsite audit, different verification frequency, or suspension under the buyer's programme.
Match verification to the failure mode
| Example failure | Correction is not enough | Useful effectiveness evidence |
|---|---|---|
| Wrong lot code on bulk packs | Relabelling the identified bags | Approved master-data change, line-clearance records, barcode/text checks and no mismatch across defined subsequent lots |
| Moisture result above contract limit | Drying or replacing the tested tea | Cause-linked process or storage control, representative results using the agreed method, and trend review across the risk-based period |
| Recurring liner seal damage | Replacing visible damaged liners | Material/equipment/operator cause evidence, controlled settings, pack-performance checks and inspection results across locations and shifts |
| Residue-panel omission | Ordering one additional test | Controlled destination list, laboratory-panel mapping, change notification and verified review on later purchase and release records |
| Foreign-material finding | Sorting the complaint sample | Source investigation, affected-scope evidence, repaired prevention/detection controls and verification capable of finding the same failure path |
Understand what the sources require—and what they do not
Codex CXC 1-1969 is a voluntary international food-hygiene reference unless incorporated into law or contract. It recommends identifying and correcting the source of a deviation to minimise recurrence, documenting cause and disposition, periodically reviewing corrective actions for effectiveness, and using ongoing verification such as observation, audit, calibration, sampling/testing and records review.
For covered U.S. operations, 21 CFR 117.150 requires written corrective-action procedures as appropriate, correction of preventive-control implementation problems, action to reduce recurrence when necessary, safety evaluation of affected food and prevention from commerce when safety cannot be assured. Section 117.410 separately requires risk-based supplier verification and prompt documented action when a receiving facility determines that a supplier is not controlling an identified supply-chain hazard. Those provisions are legal requirements for their defined operations; they are not a universal tea-industry CAPA form.
FDA's human-food supply-chain guidance is a draft containing nonbinding recommendations. It describes audits, sampling/testing, record review and other risk-based verification, with records covering both the material and the supplier problem. Buyers may adopt stricter commercial controls but should label them accurately.
- Legal requirement: binding duties depend on jurisdiction, operator role, product and activity.
- Voluntary standard: Codex or a management-system standard supplies a control framework but is not automatically law.
- Trade reference: “CAPA,” root cause and effectiveness have varying company usage; define the required fields and decisions.
- Buyer specification: sets evidence, acceptance rules, timing and enhanced verification for the supply programme.
- Contract requirement: allocates notification, access, audit, costs, suspension and remedies, without overriding food law.
Common buyer mistakes
- Closing on a supplier's completion date instead of an effectiveness result.
- Accepting “human error” without examining system conditions.
- Using retraining as the default action for equipment, method or workload failures.
- Approving a new procedure without sampling records created under it.
- Testing the corrected lot and calling that proof of non-recurrence.
- Letting the supplier select only favourable later records.
- Checking one line or shift when the change was supposed to cover the whole site.
- Setting an effectiveness window with no production or exposure opportunity.
- Closing the system action while affected product remains unreconciled.
- Failing to reopen the case after the same symptom returns.
Practical conclusion
A defensible closure file connects requirement - failure - scope - containment - cause - action - implementation - effectiveness - buyer decision. If one link is missing, keep the action open or record a controlled conditional approval. For a new supply programme, use a written bulk green tea specification and send Yunjing Tea the destination, product, evidence and change-notification requirements before defining supplier approval terms.
Sources checked 13 September 2026: Codex CXC 1-1969, General Principles of Food Hygiene, 2022 revision / 2023 edition; 21 CFR 117.150 and 21 CFR 117.410, eCFR Title 21 current through 10 September 2026; FDA, FSMA Final Rule for Preventive Controls for Human Food; and FDA Draft Guidance for Industry, Hazard Analysis and Risk-Based Preventive Controls for Human Food, Chapter 15: Supply-Chain Program. Recheck the destination rule, certification scheme and contract for the actual operation.
