ISO 22000 verification: check that PRPs and hazard controls work

What clause 8.8 requires in practice

Clause 8.8 asks the organisation to check that its prerequisite programmes and hazard control plan actually work. It has two parts. Clause 8.8.1 requires planned verification activities with a defined purpose, methods, frequency and responsibilities, confirming that PRPs and the plan are implemented and effective, that hazard levels are within the acceptable levels, that inputs to the hazard analysis are kept up to date, and that other actions the organisation decided on are carried out. The monitor of an activity does not verify it, results are kept and communicated, and adverse samples trigger handling of affected lots as potentially unsafe. Clause 8.8.2 asks the food safety team to analyse the results and feed that analysis into performance evaluation of the system.

Define what each verification activity will establish

Plan the purpose, method, frequency and responsibilities. Begin with the question the activity needs to answer. A review of records, observation of work and suitable testing can provide different evidence. Choose methods capable of supporting the intended conclusion rather than selecting checks simply because they are easy to complete.

The coverage includes PRP and hazard-control implementation and effectiveness, acceptable hazard levels, updated hazard-analysis inputs and other actions determined by the organisation. It should reflect the real products, processes and conditions. Repeatedly checking one convenient area leaves uncertainty elsewhere.

Distinguish validation, monitoring and verification

Validation establishes that a control or combination is capable of achieving the intended result. Monitoring provides information during operation to support timely action. Verification checks whether the specified arrangements are fulfilled and effective. Their evidence can interact, but their purposes remain distinct.

For example, checking that a monitoring record contains an entry can contribute to verification of implementation. It does not by itself establish that the chosen control is capable of achieving the required hazard reduction. Similarly, validation evidence does not show whether staff applied the control correctly during a particular run.

Assign appropriate responsibility

The person responsible for monitoring an activity must not carry out verification of that same activity. Organise responsibilities so that the verification provides a separate check. This requirement concerns the activity; it does not automatically demand that every verification task be outsourced.

Those conducting and evaluating verification need competence appropriate to the method and evidence. A general supervisor may be able to review some implementation evidence but require specialist support for a technical interpretation. Define how uncertain or adverse findings reach authorised decision-makers.

Retain, communicate and act on results

Keep documented verification results and communicate them. Where verification uses end-product or direct-process samples and results show a hazard level outside the identified acceptable level, the standard requires affected lots to be handled as potentially unsafe and corrective action to be applied. A result should not remain in a laboratory inbox while product decisions continue on old assumptions.

The food safety team analyses verification results under 8.8.2 and uses that analysis as an input to performance evaluation. Trends can reveal weaknesses that isolated checks obscure. Consider whether findings affect the hazard analysis, PRPs, control plan or wider resources and responsibilities.

Why clause 8.8 matters to the business

Verification is the system checking itself. Without it, an organisation relies on the same people who operate controls to tell it that the controls are fine. A genuine second look finds weak methods, missing coverage and drifting practice before they become incidents, and it supplies the evidence customers and certifiers want when they ask how you know the system works. The analysis step turns scattered checks into insight for management. Done well, verification also reduces the expense of finding problems in the market, where they cost far more than they do in your own plant.

How to implement clause 8.8

Approach verification as a planned programme of questions, each with a method and an owner. Remember that the programme itself needs review, so ask each year whether the checks still address the weaknesses that complaints, audits and incidents have exposed.

1. List what needs to be verified: each PRP, each CCP and OPRP, the hazard levels, the currency of analysis inputs and any other determined actions.
2. For each, state the question, choose a method such as observation, record review or testing, and set a frequency that fits the risk.
3. Assign competent people who are not responsible for monitoring the same activity.
4. Define how results are communicated and how an adverse sample result leads to the handling of affected lots and to corrective action.
5. Carry out the activities, recording what was examined and what was found.
6. Analyse the results across time, products and sites, and report the conclusions to performance evaluation and to the people who can act.
7. Adjust the programme when it fails to reveal known weaknesses.

Illustrative manufacturing example: records completed, method inconsistent

Illustrative catering example: a PRP across several locations

A hypothetical caterer verifies a cleaning programme at workplace service points. It checks implementation at different locations and times, then compares findings with the programme's intended conditions. One location repeatedly reports an access problem that prevents the method being followed.

The food safety team analyses the pattern and escalates the resource or layout issue. Repeating the instruction would not resolve the physical obstacle. After suitable changes, further verification examines whether the programme now operates effectively at that location.

What an auditor may examine

Explore each numbered part

Common nonconformities with clause 8.8

Verification findings in food businesses tend to cluster around a few recurring weaknesses.

• The same person monitors and verifies an activity, so the check provides no independent view.
• Verification looks only at whether records are present and signed, and never at whether the control works in practice.
• Sample testing is carried out, but nobody has decided in advance how adverse results will be linked to product decisions.
• Results sit with the laboratory or quality department and are never communicated to the people who run the process.
• The programme never changes, even after complaints or incidents reveal problems it did not detect.