Skip to content
Home / Field Note
Est. 2009 · Portland, OR · Peer-reviewed

Retest workflow after ca proposition 65 compliance failure

By admin ·Long-form field note
The short answer

A Proposition 65 failure calls for containment first, root cause analysis second, and retesting last. The retest only means something when samples are drawn randomly from a corrected lot, in numbers that represent it, after the specific material or process cause has been fixed and documented. This guide details each stage of that workflow, from quarantine through verified corrective action to paired retest and inspection records.

Key takeawaysContain the affected lot before investigating so no suspect goods ship during the analysis. · Trace the failure to a specific material, process step, or control gap rather than retesting blindly. · Draw retest samples randomly from the corrected lot in numbers that represent it. · Acceptance sampling logic governs how many units to pull and what the retest result can claim about the lot. [1] · Embed the correction in the quality system so the same failure cannot recur silently. [2]

A failed Proposition 65 assessment stops the shipment, but it should not stop the thinking. The failure is information: something in materials, process, or controls allowed a listed chemical into the product above the applicable threshold, or the warning decision built on wrong assumptions. A disciplined retest workflow converts that information into a verified correction.

The expensive mistakes happen in the rush to retest. Teams retest the same uncorrected lot hoping for a better number, accept factory-picked samples, or change suppliers without documenting what changed. Each shortcut produces a passing paper that collapses under the next check, because the underlying cause is still in production.

This article lays out the retest sequence that actually closes failures: containment, root cause analysis, corrective action, independent resampling, and paired documentation that ties the retest to the corrected lot.

Contain the suspect lot immediately

The first hours after a failure decide whether the problem stays bounded. Identify every lot that shares the suspect material, process, or production window, place it on documented hold, and notify the warehouse, the forwarder, and any distributor who may already hold goods. Holds must be physical as well as administrative: segregated stacking, hold tags, and system blocks that prevent accidental release.

Preserve the evidence the investigation will need. Retain the failed samples with their seals and labels, keep the factory's material issue records for the production dates, and freeze the bill of materials version that was actually used. Factories sometimes update paperwork after a failure in ways that obscure what really went into the product; dated, contemporaneous records prevent that drift.

Resist pressure to ship the apparently unaffected portion before the analysis is done. Until the cause is bounded, nobody knows which portion is unaffected. A premature partial shipment that later proves contaminated doubles the commercial damage and destroys credibility with the customer. Patience at this stage is cheaper than recovery later.

Find the root cause before retesting

Structure the analysis around three candidate families: wrong material, wrong process, or wrong assumption. Wrong material means a listed chemical entered through a substituted resin, pigment, plating bath, adhesive, or recycled content. Wrong process means curing, washing, or coating steps that should have removed or sealed the substance did not run as specified. Wrong assumption means the exposure model understated contact in foreseeable use.

Illustration: Find the root cause before Decorative illustration for the section "Find the root cause before"; visual only, carries no data.

Work backwards from the failed unit through production records. Check incoming material certificates against the actual lots issued to the line, verify process parameters in shift logs, and interview operators about substitutions made under time pressure. In multi-supplier products, test components separately to isolate which part carries the chemical rather than retesting the finished assembly and learning nothing.

Document the cause in one falsifiable sentence: which input or step, during which dates, introduced what substance at what stage. Vague conclusions such as supplier quality issue explain nothing and therefore prevent nothing. A quality management approach insists that corrective action address the identified cause rather than the symptom, which is why systemic framing matters from the start. [2]

Correct, then verify on the line

Corrective action must match the diagnosed cause. Material causes require approved-source changes with new incoming verification, not promises from the same unapproved source. Process causes require parameter control with recorded evidence, such as curing temperature logs or wash records, not verbal confirmation that the step was done. Assumption causes require a revised exposure assessment with documented use scenarios, not a lower number asserted without basis.

Verify the correction where it happens, on the production line, before drawing retest samples. An on-site check confirms that the new material is actually issued, the corrected parameters are actually running, and operators understand the change. Independent verification of this kind is exactly what UTS Inspection teams perform during production monitoring and pre-shipment visits.

Only when line verification shows the correction in place should resampling proceed. The sampling plan must be written before selection: how many units, from which cartons, drawn by whom, and sealed how. Statistical sampling principles determine what a given sample size can and cannot claim about the lot, so agree the plan with the laboratory and the buyer in advance. [1] Random selection by the inspector, not factory handover, is non-negotiable.

Retest sequence with decision gates

  1. Hold and segregateQuarantine every lot sharing the suspect material or window, tag physically, and block system release.
  2. Preserve evidenceRetain failed samples, material issue records, and the as-built bill of materials for the investigation.
  3. Diagnose the causeIsolate the cause to material, process, or assumption with component-level testing where needed.
  4. Apply corrective actionChange the source, control the parameter, or revise the assessment, with dated evidence of each change.
  5. Verify on the lineConfirm the correction is running on the line through independent observation before sampling.
  6. Resample and retestDraw fresh random samples from the corrected lot under the agreed plan and submit sealed units.
Illustration: Retest sequence with decision Decorative illustration for the section "Retest sequence with decision"; visual only, carries no data.

Close the file so it stays closed

A closed failure file contains five elements: the original failure with its scope, the root cause statement, the corrective action with dated evidence, the retest report tied to the corrected lot by seals and sample records, and the pre-shipment inspection confirming the shipped goods match. Missing any one of these leaves the next auditor or enforcement inquiry an open question. Completeness is the point.

Update the standing controls so the cause cannot recur quietly. Add the restricted substance to incoming inspection, qualify the replacement source in the approved supplier list, and put the corrected process parameter on the routine monitoring sheet. Management-system thinking treats each failure as input for preventive improvement rather than a one-off event. [2]

Communicate the closure factually to customers: what failed, what was bounded, what changed, and what the verification shows. Customers forgive documented failures handled transparently far more readily than they forgive surprises. Keep the language technical and evidence-based, describing results and scope rather than offering absolute assurances about the future.

Sources

  1. Wikipedia: Acceptance sampling —— Overview of acceptance sampling including AQL concepts used in product inspection.
  2. Wikipedia: ISO 9000 —— Overview of the ISO 9000 family of quality management standards.

Frequently asked questions

What is the first thing to do after a failed result?

Quarantine first, investigate second, retest third. Shipping or reworking before the cause is known risks repeating the failure across the corrected lot. Segregate the affected quantity, preserve retained samples and material records, then run the root cause analysis. Only after corrective action is verified on the line should fresh samples be drawn for retesting.

Can one passing retest sample clear the whole lot?

No. A single passing unit proves that one unit passed, not that the systemic cause was fixed. Retest samples must be drawn from the corrected lot under documented random selection, in numbers sufficient to represent the lot under the agreed sampling approach. Factory-selected samples cannot validate factory corrections.

What records prove the failure was properly closed?

Track three things: which lots were affected and how they were dispositioned, what changed in materials or process with dated evidence, and the paired retest and inspection reports showing the corrected lot conforms. Include before-and-after photographs and updated supplier declarations. A future notice or customer audit will ask exactly these questions.

Is partial shipment ever acceptable after a failure?

Yes, when the evidence supports it. If resampling under documented selection shows the bulk of the lot conforms and the defect traces to a bounded cause such as one material batch or one workstation, partial shipment of the conforming portion with the suspect portion scrapped or reworked is defensible. The decision and its evidence must be written down before anything ships.

Filed under
Free membership · No paywall

Track your own sightings alongside 42,617 verified records.

Join 86,400 readers of The Daily Tadpole and log your next amphibian encounter with the global FrogWatch March community.

Join the Pond — Free →
— FrogWorld · Vol. XVII · Issue 03 —