ISO 22000 clause 8.9.2.3: evaluate an OPRP action-criterion failure
ISO 22000 clause 8.9.2.3 requires the organisation to determine the food safety consequences and causes when an OPRP action criterion is not met. It must identify affected product, handle it under clause 8.9.4 and retain the evaluation results. The category OPRP does not make a failed criterion a minor administrative matter.
What clause 8.9.2.3 requires in practice
When an OPRP action criterion is missed, clause 8.9.2.3 asks for three things to be done. First, work out what the failure means for food safety. Second, find out why it happened. Third, identify the products affected and deal with them under 8.9.4. The organisation keeps the results of the evaluation as documented information.
Note what is not said. The clause does not declare OPRP products unsafe by default, and it does not declare them acceptable. It requires a reasoned evaluation. That evaluation then feeds clause 8.9.4.2, where release as safe is allowed only if other evidence shows the controls worked, the combined controls achieved the intended result, or sampling and verification show acceptable hazard levels. The standard also expects the correction details to be recorded under 8.9.2.4 and the corrective action to follow under 8.9.3. A common design is one form that carries the evaluation, but the standard does not insist on a form.
Determine what failed and what it means
Start with the defined action criterion, the actual observation or measurement and the relevant operating conditions. Clarify whether the result shows a failed criterion or an uncertainty requiring further evaluation. Avoid rewriting the criterion after the event to make the result appear acceptable.
Evaluate the consequence for the significant hazard the measure addresses. Consider the actual product and the role of the measure within any control combination. A general statement that other controls exist is insufficient without examining their relevance and performance.
Determine the cause of failure. Immediate facts may support product evaluation, while a fuller investigation can be needed for recurrence prevention. Keep uncertainty visible rather than delaying all action until every cause is known.
Identify and control affected product
Use process, monitoring and traceability information to define the affected product. The boundary needs a reasoned basis, particularly where a criterion is checked intermittently or an observation relates to several lots.
Handle product under the potentially unsafe-product requirements while the relevant evaluation and disposition are determined. Product should not remain freely available merely because the control is an OPRP rather than a CCP.
For release as safe after an OPRP failure, the specific evidence conditions in clause 8.9.4.2 apply. They concern evidence of effective control, the combined control outcome or appropriate verification evidence. A convenient assumption is not one of those conditions.
Record a decision others can follow
Retain the evaluation results, including the relevant facts, food safety consequences, cause determination and product handling. The connected correction and disposition requirements add further records and authority information.
Make clear who evaluated the evidence and within what competence and authority. Where specialist interpretation is needed, obtain it before relying on a conclusion. A signed release box cannot compensate for missing technical reasoning.
Why OPRP failures deserve proper evaluation
OPRPs often govern the quieter controls such as allergen changeover, handling at a handover point or a cleaning criterion, where a failure is easy to dismiss as minor. Yet these are the controls selected because a significant hazard depends on them. Teams that treat a failed OPRP as paperwork eventually meet the customer complaint or the allergen incident that follows. Evaluating consequence and cause promptly keeps good product moving where evidence supports it and stops weak product from leaving on assumption. It also produces the knowledge that management needs to decide whether an OPRP is realistic, monitored well and resourced, which is cheaper to learn from one evaluation than from a recall.
How to implement clause 8.9.2.3
Give your team a workable evaluation routine for OPRP failures, one that is quick but cannot be skipped.
- List each OPRP with its action criterion and the hazard it protects against, so the evaluator starts with the purpose.
- Provide a short question set: what was observed, which product was exposed, what does the hazard analysis say about consequence.
- Name competent evaluators for each area and state what evidence they should look for before judging.
- Set a rule that product stays held until the evaluation is complete and the decision under 8.9.4 is recorded.
- Capture the cause as soon as it is known and mark any part still uncertain, rather than leaving the field blank.
- Send the finished evaluation to whoever owns corrective action so that causes are not lost.
- Sample past evaluations during internal audit to see whether the reasoning is consistent between evaluators, and whether the cause entries were later confirmed or corrected.
Illustrative allergen-changeover example
A manufacturer has a justified OPRP within an allergen-control arrangement. An observation fails the defined criterion. The team identifies affected production, examines the food safety consequence and cause, and evaluates the actual evidence about the control combination.
It does not conclude that the product is safe simply because another cleaning step appears in the programme. The relevant performance and evidence must support the decision under the applicable release or disposition route.
Illustrative catering example
A caterer records a failed OPRP criterion at a handover stage. Competent personnel identify the meals and service route involved, examine the relevant conditions and retain the evaluation. Product remains controlled while the decision is made.
The team later improves the instruction after finding that personnel interpreted the criterion inconsistently. That corrective work does not retrospectively establish the acceptability of the affected meals.
Evidence and common mistakes
Useful evidence connects the failed criterion, product boundary, consequence assessment, cause, decision and action. Keep it linked with the correction record and any corrective action.
Common mistakes include treating OPRP failure as automatically acceptable, applying an unsupported product test as a universal release rule and recording only the process adjustment. Another is assuming that the absence of a consumer complaint establishes product safety.
Distinguish this route from the stricter release restriction after a CCP critical-limit failure. The standard deliberately gives different instructions.
Frequently asked questions
Does an OPRP failure automatically require destruction?
No. Evaluate consequences and affected product under the applicable requirements. Any release or alternative disposition needs the specified evidence and authority.
Can other controls justify release?
Only where relevant evidence demonstrates the required control outcome under clause 8.9.4.2. Their mere existence is insufficient.
Must the evaluation be recorded?
Yes. The clause explicitly requires retained documented information of the evaluation results.
Make an OPRP evaluation technically defensible
Bring Sterling an OPRP failure scenario for a scoped process review. Support can connect evidence, competence and suitable Effivity ISO & GRC automation software arrangements. Explore HACCP Awareness and the ISO 22000 hub.