Continual Improvement of an FSMS Based on ISO 22000
This domain tests whether you can distinguish genuine improvement from paperwork closure, and whether you can drive the FSMS to update itself when something changes.
What drives improvement
Improvement in ISO 22000 is data-led: trends in monitoring and verification results, repeat nonconformities, complaints, withdrawals, supplier failures, environmental monitoring clusters, near misses, internal audit findings and management review outputs. The standard expects the FSMS to be updated when hazards, processes, legislation, technology or the organisation's context change.
Correction, corrective action, improvement
- Correction fixes the immediate nonconformity (hold, rework, divert, reject).
- Corrective action addresses the cause, following root cause analysis - 5 Whys, fishbone, or fault tree - and includes an effectiveness check.
- Improvement changes the system so the failure mode is less likely or less severe, and is verified before closure.
Worked example
A bakery records the third foreign-matter complaint in six months, all metal fragments on one line. Correction: customer response and product withdrawal assessment. Root cause analysis points to a worn sieve mesh not covered by the maintenance PRP's inspection interval. Action: change the inspection frequency and add mesh integrity checks to the PRP, retrain maintenance, then verify with three months of sieve records and zero complaints. Only then is the nonconformity closed. The reasoning: without the effectiveness check and PRP update, the same failure recurs.
Traps
Recording a correction and calling it a corrective action; closing actions on the due date without evidence of effectiveness; treating "no complaints this month" as improvement; failing to update hazard analysis after a new ingredient, supplier or process; ignoring small recurring trends because each event was handled.