ISO 22000 clause 8.5.2: turn hazard information into justified control decisions
ISO 22000 clause 8.5.2 requires a food safety team analysis that determines which hazards need control and the degree of control necessary. It moves from identifying hazards to assessing significance and selecting suitable measures managed as OPRPs or at CCPs. The useful output is a defensible control decision, not merely a completed risk matrix.
What clause 8.5.2 requires in practice
Clause 8.5.2 is the heart of hazard control. It asks the food safety team to turn the preliminary information into decisions about which hazards need control and how much. There are four connected parts. First, a team analysis decides which hazards need controlling, using combinations of measures where that suits (8.5.2.1). Second, hazards are identified, located in the process, and given acceptable levels in the end product with a documented justification (8.5.2.2). Third, each hazard is assessed on likelihood and severity to find the significant ones (8.5.2.3). Fourth, control measures are selected and sorted into those managed as OPRPs and those managed at CCPs, using a systematic approach (8.5.2.4). Each stage leaves documented information explaining the reasoning. Throughout, the standard's emphasis is on reasoning that a competent reviewer can follow, not on a particular worksheet.
Identify the hazard with enough precision
Hazard identification draws on preliminary information, experience, relevant internal and external data, food-chain information and applicable requirements. Consider the product, process and environment, including where a hazard may be present, introduced, increase or persist.
A broad label such as biological contamination can be too vague to support an appropriate control decision. Identify hazards in enough detail for assessment, using competent expertise and reliable information. The analysis should include relevant stages before and after the organisation's own activities.
Determine the acceptable level in the end product for each identified hazard whenever possible and maintain its justification. Applicable requirements, intended use and other relevant information matter. An acceptable level is not the same thing as a process monitoring limit.
Assess whether prevention or reduction is essential
Hazard assessment considers likelihood in the end product before application of control measures and severity of adverse health effects in relation to intended use. It identifies significant food safety hazards.
Describe the methodology and maintain the results. A numerical matrix is one possible method, but the standard does not prescribe a universal formula. Whatever method is used, the reasoning needs to be consistent and technically supportable.
Avoid reducing the likelihood score simply because the measure being evaluated is expected to work. That can make the analysis circular and conceal why the measure is necessary.
Select and categorise suitable measures
Selection and categorisation addresses measures capable of preventing or reducing significant hazards to the defined acceptable levels. A combination may be needed. Use a systematic approach to decide which selected measures are managed as OPRPs and which at CCPs.
Consider failure likelihood, the consequences of failure, position relative to other measures, the measure's specific purpose and whether it operates alone or in a combination. Also assess the feasibility of meaningful limits or criteria, monitoring that detects failure and timely correction.
Keep the decision process, results and relevant external requirements documented. Follow with validation and a workable hazard-control plan. Categorising a measure does not establish that it has the necessary capability.
Why clause 8.5.2 matters to the business
This is where a food business shows that its controls are chosen for reasons rather than habit. Copied hazard lists and borrowed critical control points are among the most common weaknesses in food safety systems, and they fail precisely when a product or process differs from the one that was copied. A reasoned analysis helps spend effort where the risk is, avoids over-controlling harmless steps, and gives a clear answer when a customer or inspector asks why a control exists. It also provides the foundation for validation, monitoring and recall decisions later, so mistakes made here travel through the whole system.
How to implement clause 8.5.2
Run the analysis as a staged piece of technical work with a clear record at each stage. Allow enough time for each stage, because rushing identification or assessment tends to push unresolved questions into the control decisions where they are far harder to correct.
- Confirm the preliminary information is current and complete enough to analyse, and note any gaps with an owner for each.
- Assemble a team with product, process, equipment and food safety knowledge, and bring in outside expertise where the team lacks it.
- Identify hazards by step, using several sources, and state the acceptable level in the end product with its justification.
- Assess each hazard for likelihood before controls and severity for the intended use, using a method that is described and applied consistently.
- Select measures capable of reaching the acceptable level, and categorise them using a systematic approach rather than habit.
- Keep the reasoning, methodology and results as documented information a reviewer could follow without having attended the meeting.
- Reopen the analysis whenever the preliminary information changes.
Illustrative food-manufacturing example
A manufacturer reviews a process where a relevant hazard can be introduced before a later control stage. The team identifies the hazard and stage, establishes the acceptable end-product basis, assesses significance and examines the proposed control arrangement.
It then considers what happens if the selected measure fails and whether monitoring permits timely product control. The conclusion is based on the actual process and evidence, rather than declaring a familiar machine a CCP because another site uses that label.
Illustrative catering example
A catering team maps preparation, transport and service for an off-site contract. It identifies relevant hazards across the complete route and considers which controls depend on the receiving operation. The team checks that the intended-use and communication assumptions are supported.
Where several measures contribute, the analysis explains the combination and each measure's role. It does not assume that a general hygiene programme answers every significant hazard.
Evidence and common analytical failures
Keep a clear connection from source information to hazard, acceptable level, assessment and control decision. The analysis should be understandable to a competent reviewer who was not present at the original meeting.
Common failures include copied hazard lists, unexplained scores, predetermined CCP counts and treating every PRP as an OPRP. Another is completing the analysis once and failing to reconsider it after a relevant change. The general analysis guide explains the team's coordinating responsibility.
Explore each numbered part
8.5.2.1 · conduct a food safety team hazard analysis
The food safety team must turn preliminary information into an analysis that determines necessary control. This guide explains the coordinating purpose of clause 8.5.2.1, including the use of combined measures. Food-manufacturing and distribution examples show how to organise the discussion, resolve information gaps and avoid treating an inherited HACCP table as a substitute for analysis.
8.5.2.2 · identify hazards and justify acceptable levels
Hazard identification needs enough detail to support later assessment and control selection. This guide covers information sources, process locations and acceptable end-product levels, including their justification. Ingredient-production and catering examples show why generic hazard labels and unexamined limits can weaken an analysis, and how intended use and applicable requirements shape the decision.
8.5.2.3 · assess which food safety hazards are significant
Hazard assessment determines whether preventing or reducing an identified hazard to an acceptable level is essential. This guide explains likelihood, severity, intended use and documented methodology. Manufacturing and catering examples show how to challenge inconsistent scoring and avoid using the assumed success of a proposed control to hide the significance of the hazard it is intended to address.
8.5.2.4 · select controls and distinguish CCPs from OPRPs
CCP and OPRP decisions should follow a systematic examination of selected control measures. This guide explains capability, failure likelihood and consequences, control combinations, monitoring feasibility and timely correction. Manufacturing and catering examples show why a decision tree cannot substitute for technical reasoning, and why the category should follow the operation rather than a preferred audit outcome.
Frequently asked questions
Is a numerical risk matrix mandatory?
No. Describe and apply a suitable methodology that addresses the required considerations and supports the conclusions.
Does every hazard need a CCP?
No. Assess significance and select appropriate measures systematically. The number and category of controls follow the analysis.
Does ISO 22000 hazard analysis equal FSSC certification?
No. ISO 22000 requirements and a certification scheme's additional requirements are distinct. Apply the requirements relevant to the intended outcome.
Review the reasoning behind one control decision
Discuss a hazard-analysis workshop with Sterling, focused on assumptions, evidence and control selection. Support can include team learning, documentation and suitable Effivity ISO & GRC automation software arrangements. Explore HACCP Awareness, and connect the work through the hazard-control overview and hub.