A failed Proposition 65 screening result is not yet a final verdict on the lot; it is a signal to stop, segregate, and investigate before anyone authorizes shipment, rework, or a hasty retest. The most expensive mistake buyers make at this stage is retesting immediately without containment, because units continue to move, mix, and ship while the laboratory repeats the analysis on samples that no longer represent anything. This fifth part of the retest workflow series covers the first half of a disciplined response: freezing the failed lot, triaging the failure notice into a working hypothesis, and writing a retest scope that connects samples, methods, and decisions. Part six then covers supervising the correction itself.
Key takeaways
- Freeze the failed lot on the day the notice arrives: segregate, label, and reconcile counts before any diagnosis begins.
- Triage the failure into isolated excursion, systematic formulation issue, sampling or laboratory question, or documentation gap before choosing a response.
- Control the nonconforming lot under documented quality management discipline so it cannot ship, mix, or disappear. [1]
- Draw retest samples that represent the contained lot and send them to a competent laboratory within the international accreditation framework. [2]
- Write the retest scope down — samples, methods, thresholds, decision rule — before the laboratory starts work.
Most retest failures are not laboratory errors; they are workflow errors. The buyer receives a failure notice, asks the factory what happened, accepts a plausible explanation over the phone, and orders a retest of new samples supplied by the same factory. When the second result passes, nobody can say what it proves, because the lot moved, the samples changed hands without records, and the scope of the repeat was never written down.
A disciplined workflow reverses that order. Containment comes first and takes hours, not weeks: stop shipments, mark the stock, reconcile quantities, and preserve the failing samples and records. Triage comes second and takes a day or two: read the notice carefully, classify the failure mode, and decide whether the next step is a focused retest, a broader investigation, or a correction followed by verification.
This article walks through those two stages plus the bridge between them, the written retest scope. It is written for importers and QA staff who must turn a one-page failure notice into an auditable file that supports the next decision, whether that decision is release, rework, or rejection.
Stop the lot first: containment before diagnosis
Containment means the failed lot cannot move, mix, or ship until someone with authority releases it in writing. In practice that requires three actions on the day the notice arrives: tell the factory and the forwarder in writing to hold the named lot numbers, visit or video-verify that the stock is physically segregated with hold labels, and reconcile the counted hold quantity against production and packing records. Any gap between the expected and the counted quantity is itself a finding, because unaccounted cartons may already be on their way to the port.
The discipline behind this step is the same one described in quality management foundations: nonconforming outputs must be identified, segregated, and controlled so they are not used or delivered unintentionally, with the concession or release decision recorded. [1] The standard does not prescribe hold-label colors or warehouse layouts, but its logic is exactly what containment implements — the lot keeps its identity, its status is visible to everyone on the floor, and no informal instruction can override the hold. Buyers who skip this step often discover later that corrected and uncorrected cartons were stacked together, which invalidates both the investigation and the retest.
Preserve the evidence while the lot is frozen. Keep the failing laboratory report with its sample identifiers, retain the remaining failing samples under seal, pull the production records for the lot including material batches and process settings, and file the packing list that links cartons to the tested units. Photograph the segregated stock with legible lot markings and timestamps. This small archive costs almost nothing and becomes the reference point for every later argument about what the retest actually covered.
Reading the failure notice like a triage nurse
A failure notice answers fewer questions than buyers wish it did. It states which characteristic was measured, on which samples, by which procedure, and against which named requirement — but it rarely states whether the problem affects one production run or the whole formulation, one component or the finished article. Triage is the work of narrowing that uncertainty before spending money on laboratory work. Read the notice for scope first: sample description, lot linkage, component breakdown, measured values with units, detection limits, and any remarks on deviations or subcontracted steps.
Sort the situation into one of four working hypotheses. An isolated excursion fits when one component or one batch drifts while sibling lots pass, and it points toward focused verification of the suspect portion. A systematic formulation issue fits when the same characteristic fails across batches or the bill of materials names a material family known for the regulated substance, and it points toward substitution rather than repeated testing. A sampling or laboratory question fits when identifiers do not match, units look wrong, or the result contradicts a strong history of passes, and it points toward verification of the test itself before any correction. A documentation gap fits when the product may be compliant but the file cannot show it, and it points toward records rather than chemistry. At this point many buyers ask California Proposition 65 testing to confirm whether the reported result describes the whole shipment or only the sampled portion. California Proposition 65 testing
Write the triage conclusion in one paragraph with its reasons and its proposed next step, then get the factory's written agreement or written objection. This document prevents the most common drift in failure response, where each phone call quietly redefines the problem until the retest is answering a different question than the failure raised. A triage note that says isolated excursion in component X, supported by batch records and sibling-lot history, next step focused retest of component X from sealed stock, is worth more than a week of verbal reassurance.
Triage outcomes and what each one requires
| Working hypothesis | Typical evidence | Retest implication |
|---|---|---|
| Isolated excursion | One batch or component fails while sibling lots pass; batch records show a material or setting change | Focused retest of the suspect portion from sealed stock plus verification of segregation |
| Systematic formulation issue | Repeated failures across batches; bill of materials points to a high-risk material family | Do not repeat the same test; correct the formulation first, then verify the corrected lot |
| Sampling or laboratory question | Identifier mismatches, unit anomalies, or contradiction with strong pass history | Verify sample identity and method applicability before any new sampling; consider a second laboratory |
| Documentation gap | Product plausibly compliant but traceability between samples, lot, and report is broken | Rebuild the chain with sealed sampling and records; retest only once traceability is restored |
| Mixed or unclear picture | Conflicting records, mixed stock, or missing production data | Freeze everything, complete the count reconciliation, then re-triage; no testing until the lot is defined |
Writing a retest scope the laboratory can execute
- Define the lot in writingName the lot numbers, carton range, component breakdown, and the hold-quantity reconciliation the retest covers, so the laboratory report can reference the same definition and no other stock can borrow the verdict.
- Specify the sampling planState who draws the samples, from sealed and segregated stock, how many units per component and carton, how they are sealed and labeled, and what chain-of-custody record accompanies them to the laboratory.
- Name the procedures and thresholdsWrite down the analytical procedures, the components tested, the units of measurement, and the named requirement each result will be compared against, so the laboratory executes rather than interprets the scope.
- Set the decision rule in advanceRecord what combinations of results mean release, conditional correction, or rejection, including how marginal or split results will be handled and who has authority to decide, before any numbers arrive.
- Close the loop on recordsRequire the report to carry sample identifiers that match the sampling record, file the report with the triage note and the sampling record, and note any deviation from the scope so the file shows what was planned, what was done, and why they match.
Keeping the second verdict credible
A retest is only as credible as the laboratory behind it and the sampling in front of it. On the laboratory side, buyers should look for demonstrated competence under recognized accreditation, the kind of system coordinated internationally through mutual recognition arrangements among accreditation bodies. [2] Accreditation does not make a result correct by magic, but it means the laboratory's competence, impartiality, and procedures have been independently assessed, and that the report carries traceable identifiers a buyer can actually file and defend.
On the sampling side, the rule is simple: the retest samples must come from the contained lot through a witnessed or verifiable draw, never as loose units handed over by an interested party without records. Independent drawing, sealed bags, photographed carton marks, and a signed sampling record cost a fraction of a second laboratory round and eliminate the most common objection to a passing retest. When the sampling record, the laboratory identifiers, and the hold reconciliation all name the same lot, the verdict belongs to that lot; when any link is missing, the verdict belongs to nobody.
Resist the temptation to shop for a pass by testing repeatedly until a favorable number appears. Each additional unscoped round dilutes the file rather than strengthening it, because a future reviewer will ask why the earlier failures were set aside. One contained lot, one written scope, one competent laboratory, one recorded decision — that sequence is slower than a quick repeat, and it is the only sequence that survives scrutiny.
Containment first, triage in writing, scope before sampling: the retest earns its authority from the workflow around it, not from the laboratory alone.
Sources
- Wikipedia: ISO 9000 —— Overview of the ISO 9000 family of quality management standards.
- ILAC: International Laboratory Accreditation Cooperation —— ILAC homepage describing the international cooperation on laboratory and inspection body accreditation.
Frequently asked questions
How fast must containment happen after a failure notice?
The written hold instruction should go out the same day the notice arrives, with physical segregation verified within one to two days. Every shipment, transfer, or sorting step that happens in between creates stock whose status cannot be reconstructed. Speed here matters more than precision: a broad hold that is later narrowed is far cheaper than a narrow assumption that lets affected cartons ship.
Should we retest with the same laboratory or a different one?
Stay with the same competent laboratory when the question is whether the lot complies, because continuity of procedure makes the two results comparable. Bring in a second laboratory when the question is whether the first result can be trusted — identifier mismatches, method doubts, or results that contradict strong history. In either case the retest scope should be written before the laboratory is engaged, not negotiated afterward.
What if the factory says the failure is a laboratory mistake?
Treat that claim as a hypothesis to be tested, not a conclusion to be accepted. Ask for the specific alleged error — wrong component, wrong procedure, wrong unit, sample mix-up — and check it against the report identifiers and the sampling record. If the claim survives that check, the next step is verification of the test itself, such as review of the procedure's applicability or analysis by a second competent laboratory, before any correction or resampling.
Can part of the lot ship while the rest is under investigation?
Only if the shippable portion was segregated before the failure with records proving it never mixed with the affected stock, and the triage note explicitly releases exactly that portion with its own identifiers. Partial release on verbal assurance, or from mixed stacks sorted after the fact without supervision, contaminates the file and usually the stock as well. When segregation history is unclear, hold everything until the reconciliation is complete.