Recurring Label Error and an Unverified CAPA
At a contract manufacturer, three lots of a powered surgical instrument shipped with the wrong sterilization indicator label: the 50-cycle label instead of the 25-cycle label. The initial corrective action was to retrain the two operators on the work instruction, add a signature on the training record, and revise the work instruction to bold the label part number. The CAPA was closed 30 days later after the QA supervisor checked five units and found correct labels. Four months later, the same error occurred on two lots during a weekend shift with a temporary operator. An internal audit now finds the CAPA file has no root cause statement, no containment record, and no defined effectiveness check. The customer requires CAPA closure only after objective evidence that the cause has been eliminated, and the site is preparing for an ISO 13485 surveillance audit.
1. Critique the original CAPA against the elements of containment, root cause, correction, corrective action, and effectiveness verification.
Read the worked answer
The original CAPA was a correction dressed as a corrective action.
- Containment: missing. There is no record of checking WIP, finished goods, warehouse stock, or labels already issued to the line, nor of notifying customers about the three shipped lots.
- Root cause: missing. Bold text in a work instruction is an attempt to make an error less likely; it does not explain why a look-alike label could be selected and applied without detection.
- Correction: retraining and document revision address the symptom for two named operators.
- Corrective action: none systemic - no change to label storage, pick/issue controls, scanning, or line clearance, so the same failure mode remained available to any operator, including temporaries.
- Effectiveness verification: five units checked by a supervisor is a spot check, not objective evidence; 30 days is too short to show the error does not recur.
The audit findings are accurate and would be a nonconformity under a CAPA clause requiring determination of causes and review of effectiveness.
2. Design the root cause investigation. What data and failure modes would you examine?
Read the worked answer
Investigate the system that allowed the wrong label, not just the person who applied it.
Data: pick and issue records for labels; label inventory locations and whether 25- and 50-cycle labels are stored adjacent or share similar part numbers and colors; line clearance records at changeover; staffing records for the weekend shift; whether any scan or reconciliation step exists.
Failure modes to test:
- Look-alike labels - similar part numbers or artwork, enabling selection error.
- No error detection - absence of barcode verification or second-person check at label application.
- Line clearance - previous lot's labels left at the station after changeover.
- Training dependency - a process that relies on memorized part numbers fails with temporary staff; this is a design weakness, not a training gap.
- Human factors - label placement, lighting, shift fatigue.
Use 5 Whys or a fishbone to drive from 'temporary operator picked the wrong label' to the control that permitted it. Verify each hypothesized cause with records, not interviews alone.
3. Define effectiveness verification and closure criteria, and state what must be done about the two lots already shipped.
Read the worked answer
Effectiveness must be measurable, time-bound, and independent of the person who made the change.
Closure criteria, defined before implementation:
- A mistake-proofed control (barcode scan that rejects the wrong label, or physically separated/differently keyed label stock reconciled per lot).
- Zero label errors over a defined period and volume - for example, 20 consecutive lots including at least two weekend shifts, verified by label reconciliation records, not by spot checks.
- Audit of the scan/reconciliation records to confirm the control is actually used.
- A named owner, review date, and a defined response if the control fails.
Shipped lots: immediately contain and trace the two lots; assess patient-safety impact from the wrong sterilization indicator (the user may follow the wrong cycle parameters); notify the customer and, with them, decide on field action, replacement labeling, or recall per the risk assessment. Document all of it in the CAPA record. Closure before this evidence is not defensible at the surveillance audit.