Ready-meal plant: batch trace, unapproved supplier and a paper recall test
During stage 2 of a ready-meal manufacturer producing 22,000 chilled meals a day, you trace chicken curry LOT 24-118 (1,860 packs, produced 14 March). Forward tracing shows 1,600 packs to retail and 180 to a staff canteen; 80 packs are unaccounted for. Backward tracing shows the chicken came from Valley Poultry, an approved supplier whose certificate on the approved supplier list expired five months ago; no re-approval is recorded. Two days before the audit the plant received an email customer notice from Valley Poultry flagging Listeria in a different lot. The last withdrawal/recall test was 11 months ago and was a desk exercise: the quality manager emailed the customer list to herself and recorded 'recall completed in two hours'. The hazard analysis sets cooking at 75 °C for 2 minutes core as an OPRP; the 14 March temperature records show a six-minute gap. The operations manager says the gap is 'just the probe being changed' and offers to explain it verbally.
1. Analyse the traceability evidence. What findings do you raise, and how do you distinguish a traceability failure from a potentially unsafe product situation?
Read the worked answer
Reconstruct the mass balance before drawing conclusions. 1,600 + 180 = 1,780 packs dispatched; 1,860 produced leaves 80. Establish where those 80 went: canteen consumption not recorded, samples, rework, waste, or unrecorded dispatch. A traceability system must be able to identify and trace lots and to reconcile quantities; an unreconciled balance in a live test is evidence, and the size and pattern matter. If the 80 packs are genuinely untraceable, that is a nonconformity against clause 8.3 (traceability system) and, if the product could have left the site unrecorded, a serious one.
Separately assess product safety. The supplier notice concerns a different lot, so first check whether any material from the flagged lot is in your store - that is a potentially unsafe product question under 8.9.4/8.9.5, not a traceability question. The expired supplier certificate is a nonconformity against the supplier-approval PRP (clause 8.2), and the hazard analysis (8.5.1.2) should state which raw-material hazards depend on supplier controls; if approval is the control for Listeria in raw chicken, the lapse undermines the hazard analysis.
Do not merge findings. Each nonconformity needs its own clause, its own objective evidence and its own statement of the requirement not met. State the reconciliation figures in the finding so the reader can see the arithmetic.
2. The temperature record shows a six-minute gap and the operations manager offers a verbal explanation. How do you proceed?
Read the worked answer
Take the explanation into account by verifying it, not by accepting it. Ask for the probe-change procedure, the calibration records for the probe fitted that day, the maintenance log for the 14 March changeover, and the records either side of the gap. If the probe change is planned and the procedure requires the product to be held or the process restarted, check whether that happened. Interview the operative who made the entry.
Then test whether this is an isolated event or systemic. Sample temperature records across other days and operators: how often do gaps occur, how are they closed out, and are corrections recorded? A single unexplained gap in an OPRP record with a plausible, verifiable explanation may be a minor or an observation. A pattern of unexplained gaps with no corrections is significant - it means the OPRP is not being monitored as designed and the organisation cannot show it acted when action criteria were not met (clause 8.5.3/8.9, corrections).
Weigh it against the hazard: cooking is the main Listeria control for this product, so loss of monitoring evidence at that step carries real risk and should be presented with that context. Record the verbal explanation in your notes, state what you verified, and state clearly where the explanation could not be evidenced - 'no record available to confirm the verbal explanation' is a factual, defensible statement.
3. How do you evaluate the effectiveness of the previous withdrawal/recall test, and what would you test live before leaving site?
Read the worked answer
The 11-month-old test fails as evidence. Emailing a customer list to oneself tests nothing: no product moved, no customer was contacted, no quantity was reconciled, and 'two hours' measures an email, not a withdrawal. Under clause 8.9.5 the organisation must be able to withdraw or recall, and the effectiveness of that capability must be verified - a desk exercise can be acceptable only if it genuinely exercises the plan (contacts, quantities, mock reconciliation, decision-making, communication) and is followed by review.
Run a live test during the audit: select a current lot, ask for the customers and quantities, and time how long it takes to produce a complete dispatch list including the canteen and any intercompany transfers. Check the contact list is current - ring one customer's recall contact on speaker with permission, or at least verify the named person still works there. Check that the 80 unaccounted packs could be found by this process.
Then assess the system that should have triggered action: the Valley Poultry notice arrived two days ago. What did the organisation do with it? Was there a supplier notification procedure, an assessment of affected lots, a decision recorded? If the notice was filed without action, the nonconformity is not about the old recall test but about the current inability to evaluate a supplier food safety alert - a much more serious finding.