ISO 22000 clause 8.5.4.3: detect CCP and OPRP failures in time to act

What clause 8.5.4.3 requires in practice

Monitoring is how the plan detects trouble. At each CCP, a monitoring system is set up for each measure or combination of measures to detect any failure to stay within the critical limits, covering all scheduled measurements against the limit. For each OPRP, monitoring shows when the action criterion has not been achieved. The monitoring system is written down. The documentation sets out the measurements or observations, and shows that results arrive soon enough to act on, the methods or devices, how calibration or equivalent verification of reliable readings is handled, the frequency, the results, and who is responsible and has authority for monitoring and for evaluating the results. At a CCP, method and frequency must allow timely isolation and evaluation of product. For an OPRP, they are proportionate to likelihood and severity, with instructions where observation is subjective.

Define the complete monitoring arrangement

Document the measurements or observations, method or device, frequency and results. Include applicable calibration methods or, for OPRPs, equivalent methods that establish reliable measurements or observations. Define responsibility and authority both for monitoring and for evaluating the results.

Check that the method measures or observes the condition specified in the plan. A convenient reading from a nearby location may not represent the parameter on which the control depends. The technical basis must fit the actual process.

Make the relationship between a result and the relevant product or production period clear. This supports the later isolation and evaluation of affected product when a failure occurs.

Apply the CCP timing requirement

At each CCP, the method and frequency must detect failure to remain within the critical limits in time to allow timely isolation and evaluation of product. Consider how much product can be affected between checks and how it remains identifiable and controllable.

Do not choose frequency solely because it is convenient for an operator's other tasks. Establish a reasoned arrangement consistent with the process and potential failure. The standard does not provide one universal monitoring interval.

A recorded value also needs evaluation. Define what happens when a reading is missing, uncertain or outside the limit, including the authority to initiate the relevant response.

Make OPRP monitoring proportionate and clear

For OPRPs, the method and frequency must be proportionate to the likelihood of failure and severity of its consequences. Where monitoring relies on subjective observation, supporting instructions or specifications are required.

An observable criterion should be clear enough that competent people interpret it consistently. Use suitable examples or descriptions where helpful, while preserving the actual technical meaning. A tick labelled satisfactory is weak if nobody can explain what was observed.

Connect monitoring with measurement control, competence and failure response.

Why clause 8.5.4.3 matters to the business

Monitoring that finds a problem too late, or measures the wrong thing, gives false comfort and leaves product at risk. The longer the gap between a failure and its detection, the more food has to be held, tested or discarded, so well-judged monitoring directly limits cost as well as risk. It also provides the data that proves control to customers and auditors. Weak monitoring, on the other hand, is a common root cause behind recalls, because records looked fine until the day somebody examined how the results were obtained.

How to implement clause 8.5.4.3

Design monitoring around the decision it supports, not around what is easiest to record. Pay particular attention to the link between a reading and the product it represents, since that link decides how much material has to be held when a result is bad.

1. For each CCP and OPRP, state what exactly is measured or observed and why that detects loss of control.
2. Choose the method or device and confirm it can read the parameter in the real process, then link it to measurement control under clause 8.7.
3. Set the frequency by asking how much product could be affected between checks and whether it can still be isolated.
4. For OPRPs relying on observation, write instructions or specifications so trained people reach the same conclusion.
5. Define who records, who evaluates, and who may act on a result, with cover for each shift.
6. Make records show the time and the product or production period they represent.
7. Trial the arrangement, correct what is confusing, and train the people who will use it.

Illustrative food-processing example

Illustrative allergen-control example

A manufacturer uses a justified OPRP with an observation during a specified changeover. Two trained operators interpret the criterion differently. The team clarifies the instruction, checks the observation method and reassesses competence before relying on the monitoring result.

The observation is one defined part of the control arrangement. It is not presented as universal proof that all allergen residue is absent.

Evidence and common failures