Holiday Surge, Undisclosed Subcontracting and a Report Bearing the Accreditation Mark
NordFood is a 15-person food testing laboratory accredited for 24 methods. On the Friday before a public holiday, a new client sent 120 samples. A junior customer service representative, who had been in post for five weeks, performed the contract review by telephone and accepted the work for all requested tests. Eighteen aflatoxin M1 samples were subcontracted to a university laboratory that is not accredited for that method; the client was never told. The samples arrived at 26 °C with the ice packs fully melted; they were logged in and stored at 4 °C for 30 hours before analysis. One analyst manually transcribed peak areas from the instrument printout into the laboratory information system without a second check or audit trail. The reports were issued on letterhead carrying the accreditation symbol and the phrase accredited testing for all results. The client complained three weeks later.
1. List the nonconformities in this sequence, classify each as a data-integrity issue or a system weakness, and give the clause basis.
Read the worked answer
1. Contract review by an unauthorized, untrained representative, verbal only. System weakness and record failure. Basis 7.1.1 (review of requests) and 6.2 (competence and authorization). The review must establish capability, resources, method and acceptance criteria, and be recorded.
2. Subcontracting without informing the client and without documented evaluation of the provider. Data-integrity risk, because the client believes the results are the laboratory's accredited results. Basis 6.6 (externally provided products and services), 7.1.1 and 7.8 (identification of externally provided results).
3. Receipt and storage of samples at 26 °C with melted ice packs for 30 hours before analysis. Data-integrity issue: the results may not be valid. Basis 7.4.1. The condition should have been recorded, the client consulted and the samples held pending instructions.
4. Manual transcription with no verification and no audit trail. Data-integrity issue. Basis 7.5 (technical records) and 7.11 (control of data and information management).
5. Reports carrying the accreditation symbol for tests outside the scope and for a non-accredited subcontractor. The most serious item: a misrepresentation, both a data-integrity matter and a breach of the accreditation agreement.
6. Capacity planning for a holiday surge. System weakness and a contributing cause under 7.1.1, because the review must consider resources.
Classification matters because item 5 can threaten the accreditation, whereas items 1 and 6 are corrected through process design.
2. Define the corrective action for the contract review and subcontracting failures: root cause, containment for the samples already reported, and what must be done with the reports.
Read the worked answer
Containment first. Identify every report affected: all 18 aflatoxin M1 results, any result from samples held 30 hours, and any report claiming accreditation beyond scope. Notify the client in writing, quarantine the records in the laboratory information system, and determine whether samples still exist for re-testing.
Reports. Issue amended reports under 7.8, clearly identifying the amendment, the reason and what has been superseded. Where the original report is invalid, the amendment must say so rather than quietly replace the file. For the subcontracted results, identify the external provider explicitly.
Root cause - go beyond the individual. Asking why the representative was careless is not a root cause. Ask instead: why did a five-week employee have authority to accept non-routine work; why was there no technical sign-off gate; why was the subcontracting route not part of the contract review form; why did the system allow a report to be released with a scope mismatch.
Corrective actions. A competence and authorization matrix for customer service staff; mandatory technical sign-off for non-routine, rush or subcontracted requests; an approved external provider list with evaluation records; a hard stop in the information system preventing release when accreditation status is not confirmed; and a customer notification clause in subcontracting.
Closure. Record under 8.7 and verify effectiveness (8.7.3) by auditing reports issued after the change, not by confirming that a procedure was written.
3. Design three controls that would have prevented the transcription error and the sample-storage failure, and explain how you would verify each control is effective rather than merely documented.
Read the worked answer
Control 1 - Direct data transfer with an audit trail. Replace manual transcription with an instrument-to-system import, or require a second-person verification of any manual entry, with the reviewer identified and the raw data retained. Verify by sampling 30 reports and tracing each result back to raw data, and by deliberately entering an incorrect value to confirm the system flags it.
Control 2 - Condition-on-receipt hold. Log temperature and condition for every consignment, with an automatic hold and a customer-consultation step when acceptance criteria are breached. Verify by auditing receipt records against storage records for a month and checking whether any held consignment was released without a documented decision.
Control 3 - Release gate tied to accreditation status. The information system checks that every method on a report is within the granted scope and that subcontractor results are flagged before release is possible. Verify by attempting to release a report for a method outside scope and confirming it is blocked.
For each, define the check, the person responsible and the frequency, and report the results at management review. A control that has only been written into a procedure has not been verified; 8.7.3 requires evidence that the corrective action prevented recurrence, which means looking at outputs after the change.