Insights from our Industry Webinar on Strengthening Allergen Management in Food Manufacturing.

In this article, we recap key insights from the first webinar in our three-part series with Campden BRI, exploring why allergen management systems can still fail despite the procedures, training, checks, and controls already in place. The session featured key insights from:

  • Alec Kyriakides, Independent Food Safety Consultant.
  • Wendy Duncan, Safety, Quality and Allergens Manager, Campden BRI.
  • Alex Carlyon, Key Account Director, Klipspringer

If you are short on time, you will find a bulleted list of key takeaways at the end of each section.

Watch the complete webinar recording

Legal Obligations

Alec opened the webinar with a closer look at the legal requirements that apply to UK food and drink manufacturers, which include:

  • Assimilated Regulation (EU) No. 1169/2011 (for England and Wales)
  • Regulation (EU) No. 1169/2011 (for Northern Ireland)
  • Plus, enforcement regulations (Food Information Regulations 2014) and subsequent amendments, e.g. FIR 2019 (Natasha’s Law)

In terms of the specific areas of food production, Alec outlined the following requirements:

  • Prepacked food is expected to carry a full ingredient list, with any of the 14 allergens emphasised within it
  • Prepacked food for direct sale (PPDS) is also expected to carry a full ingredient list, with any of the 14 allergens emphasised within it. This category also relates to Natasha’s Law
  • Non-prepacked food requires a declaration of any allergens used as ingredients, which can be provided in electronic, written or oral form
  • Distance selling (e.g. e-commerce) requires allergens to be declared before purchase and again when the food is delivered

Precautionary Allergen Labelling

Precautionary allergen labelling (PAL) is not mandatory. However, failing to provide it where there is a genuine risk may expose consumers to undue risk and the product could be deemed unsafe.

It is essential that PAL does not mislead or confuse the consumer, and a big part of this is keeping it separate from the ingredients list. PAL can be used alongside “vegan” and “gluten-free” claims, but not alongside other “free-from” claims.

What’s Coming Next?

  • Benedict’s Law now applies to state schools, which are required to conduct training and hold adrenaline auto-injectors, among other measures
  • Owen’s Law is particularly focused on making allergen information mandatory in hospitality environments. It is supported by the FSA and is recommended guidance, but is not yet law
  • Codex has recently published allergen thresholds that are being used to support risk-assessed precautionary allergen labelling (PAL)

Key Takeaways

  • Prepacked and PPDS foods must carry full ingredient lists with allergens emphasised
  • Non-prepacked and online sales have their own allergen declaration rules
  • Precautionary allergen labelling isn’t mandatory, but omitting it where a real risk exists could make a product unsafe
  • PAL must be kept separate from the ingredients list and used alongside claims with care
  • Benedict’s Law, Owen’s Law and new Codex thresholds signal further change ahead

What Do Allergen Recalls Tell Us?

Looking at recall data can provide a useful starting point for understanding where allergen management is going wrong. Alec compared 2025 recall data from Australia, the UK and the USA. Despite the differences between these countries, the patterns were remarkably similar.

Milk was the most common allergen driving recalls, while multi-component products were among the most common food groups involved. There are several reasons why these products can be challenging to manage. Multiple ingredients make labelling more complex, while the same complexity can create additional challenges within the factory.

But recall data only tells us where something has gone wrong. It does not necessarily tell us why. Looking at UK Food Standards Agency allergen recalls from January to August, Alec highlighted several common causes:

  • 43% involved an allergen that was not listed in the ingredients list
  • Around a third involved mis-packs, where the wrong product or packaging reached the line
  • 10% involved cross-contamination
  • 10% involved an allergen being listed but not emphasised correctly

The figures provide a useful sense check for manufacturers. If these are the types of issues leading to recalls across the industry, could the same weaknesses exist within your own site?

What About Audit Non-Conformities?

Recall data is only one part of the picture. Audit non-conformities can provide another useful source of information. In hospitality and catering, allergen-related non-conformities commonly involve missing or inaccurate allergen information, or customers not being clearly signposted to where they can find it. Allergen contamination control accounts for much of the remaining issues.

BRCGS data from food manufacturing shows a slightly different picture. Cross-contamination management is a major driver of non-conformities, with weaknesses in procedures and risk assessments accounting for a significant proportion. On-pack labelling makes up much of the remainder. The important point is that these findings should not simply be viewed as audit statistics. They can be used as a prompt to ask: could the same weaknesses exist in our business?

Key Takeaways

  • Undeclared allergens and mis-packs are common causes of recalls
  • Labelling verification remains a significant area of risk
  • Cross-contamination continues to be a major challenge
  • Audit findings can highlight areas worth challenging within your own systems
  • Recall and audit data show symptoms, but not necessarily the root cause

Finding the Root Cause

An allergen incident rarely starts with the final mistake that is visible to the consumer. Alec highlighted the importance of analysing why these incidents happen, rather than only what went wrong at the point of failure.

Drawing on the FSA’s 2025 incident prevention report, Alec looked at the key drivers behind the majority of recalls, as explored through root cause analysis. The most common specific issues within the main categories were:

  • Process: labelling verification checks (30%)
  • Material: ingredient cross-contamination (11%)
  • Material: ingredient specification (8%)
  • Method: labelling declaration (14%)

Key Takeaways

  • Allergen incidents rarely start with the final, visible mistake
  • Understanding why incidents happen matters more than what went wrong at the point of failure
  • Root cause analysis from the FSA’s 2025 incident prevention report reveals the key drivers behind most recalls

The Swiss Cheese Model: Why Controls Fail

Wendy introduced James Reason’s Swiss cheese model as a useful way of thinking about allergen management. The model illustrates how food safety incidents rarely result from an isolated mistake.

Businesses have multiple controls and barriers in place, including supplier approval, segregation, cleaning, label verification and product release. Each of these controls can help to prevent an allergen from reaching the consumer. However, an incident can occur when weaknesses or failures align across several of these barriers, allowing the issue to pass through the system.

This is why focusing solely on the final error can miss the bigger picture. The objective is not simply to identify the last failed check, but to understand why the earlier controls did not work effectively.

Key Takeaways

  • James Reason’s Swiss cheese model shows that food safety incidents rarely stem from a single mistake
  • Incidents occur when weaknesses across several controls align
  • Focusing only on the final error can miss the bigger picture
  • The goal is to understand why earlier controls failed, not just identify the last failed check

Prevent, Eliminate or Reduce?

Businesses need to be very clear about the purpose of the allergen controls they have in place and the outcomes they are intended to achieve. Are you trying to prevent the hazard from being introduced, eliminate the hazard, or reduce the risk to an acceptable level? The intended outcome should then determine how the control is designed, validated, monitored and verified.

Examples of allergen controls include:

  • Supplier controls
  • Segregation
  • Formulation controls
  • Production controls
  • Packaging and labelling controls
  • Physical separation
  • Controlled handling
  • Production sequencing
  • Cleaning

Cleaning is an important example. Completing a cleaning procedure does not automatically mean that all allergenic material has been removed. Its effectiveness will depend on a range of factors, including the allergen, the product, the equipment and the cleaning method used. As Wendy highlighted, cleaning will be explored in more detail in Part 2 of the series.

Key Takeaways

  • Businesses need to be clear on whether each allergen control aims to prevent, eliminate or reduce the hazard
  • The intended outcome should shape how a control is designed, validated, monitored and verified
  • Allergen controls span the whole operation, from supplier and formulation controls to segregation, production sequencing and labelling

Where Are Your System Vulnerabilities?

One of the biggest areas of vulnerability in allergen management can be at changeover points. Think about the handovers between:

  • Supplier and manufacturer
  • Procurement and technical
  • Warehouse and production
  • Hygiene and production

Each department may understand its own responsibilities, but weaknesses can arise when one team assumes that another has obtained, checked or communicated important information.

A supplier substitution may look like a routine commercial decision, but is the allergen profile the same, and is there a process in place to manage the change? Even a small formulation change may appear insignificant, but what impact does it have on specifications, labelling, scheduling, rework and cleaning?

Another example is when teams assume that completing a cleaning process means that a line is ready to run. But has it been inspected and verified? The challenge is to identify the points where critical information could be lost, delayed or misunderstood.

Key Takeaways

  • Changeover points between teams and organisations are among the biggest allergen vulnerabilities
  • Weaknesses arise when one team assumes another has obtained, checked or communicated key information
  • Businesses should identify where critical information could be lost, delayed or misunderstood

The Procedure Trap

A procedure is evidence that a business has described what should happen; it is not evidence that what is written actually happens. A stronger assessment asks:

  • Is the procedure technically appropriate?
  • Does it reflect the real process?
  • Do people understand it?
  • Is it consistently applied?
  • Can it withstand operational pressure?
  • Does it achieve its intended outcome?
  • Is there evidence that it remains effective?

The same applies to training. A training record confirms that someone attended training, but it does not necessarily demonstrate that they can apply the process correctly, recognise a problem or make the right decision when something goes wrong.

Key Takeaways

  • A written procedure shows what should happen, not what actually happens
  • Procedures should be tested against the real process, operational pressure and their intended outcome
  • Ongoing evidence is needed to show a procedure remains effective

5 Warning Signs To Look Out For

Wendy highlighted several warning signs that could indicate weaknesses in an allergen management system.

1. Generic Risk
Assessments

Does your allergen risk assessment reflect your site’s actual allergen hazards and practices?

2. Allergens Are Seen as One Department’s Job

Effective controls depend on business-wide ownership, not just one department.

3. Training Without
Competence

Attending training is not the same as demonstrating competence.

4. Near Misses Are Corrected but Not Investigated

Near misses need to be investigated, not just corrected without further thought.

5. Controls Are Built for the Audit

Allergen controls need to work every day, not just when an auditor is expected.

Key Takeaways

  • Allergen risk assessments should reflect your site’s actual hazards and practices, not generic templates
  • Effective allergen management needs business-wide ownership, not just one department
  • Attending training isn’t the same as demonstrating competence
  • Near misses should be investigated, not just corrected
  • Controls must work every day, not just when an auditor is due

Food Safety Culture

An effective food safety culture is vital across the business, and a key part of this is building robust controls around hazards.

As mentioned throughout the webinar, an incident rarely occurs due to one isolated failure. Going back to the root cause and understanding the failure pathway can reveal a series of weaknesses across multiple teams that have ultimately allowed the incident to occur. This highlights that allergen control is a company-wide responsibility and should never be left to one department.

Building this culture requires a clear sense of ownership. People need to understand their roles and responsibilities, and these should be clearly assigned rather than assumed. Every department has a role to play, from procurement and product development through to production and warehousing.

Senior leadership also has an important role in ensuring people have sufficient time and resources to maintain robust allergen controls.

Key Takeaways

  • A strong food safety culture depends on robust controls around hazards
  • Tracing the failure pathway often reveals weaknesses across multiple teams
  • Allergen control is a company-wide responsibility, not one department’s job
  • Roles and responsibilities should be clearly assigned rather than assumed
  • Senior leadership must give people the time and resources to maintain robust controls

Moving From Confidence to Evidence & Procedures to Control

One of the strongest messages from the webinar was simple: Confidence isn’t evidence. A business can be confident that its allergen controls are working, but that confidence needs to be supported by evidence. Wendy outlined five important stages:

  • Validate: Can the control achieve its intended outcome?
  • Monitor: Is the control continuing to operate within its defined criteria?
  • Correct: How do we regain control if it has been lost?
  • Verify: Is the wider system working consistently in practice?
  • Review: Does the control remain effective after changes or new information?

Testing can support verification, but testing alone is not a control. One result represents one sample, in one location, at one point in time. The question is not simply whether the required activities have been completed. The question is whether the combined evidence shows that the controls consistently achieve their intended outcome.

Key Takeaways

  • Confidence isn’t evidence; allergen controls need proof they’re working
  • Controls should be validated, monitored, corrected, verified and reviewed
  • Validation confirms a control can achieve its intended outcome
  • A single test result reflects one sample, in one location, at one point in time. The real question is whether combined evidence shows controls consistently achieve their intended outcome

5 Actions To Take Today

To finish the webinar, Alex brought the discussion back to the factory floor with five practical actions that can be taken immediately at your site.

1. Risk Assessments

Every risk assessment should be specific to the line, product and site. Avoid simply copying an assessment from another area, as this can be a red flag for auditors.

Don’t just write the assessment at your desk. Walk around the factory and observe things like where products are transferred and where rework takes place. Watch a changeover and follow the movement of equipment and people.

Most importantly, involve the people who actually work there. An operative on the line may be able to identify a practical risk in two minutes that would never appear in a desk-based review.

2. Keep It Simple

Allergen management systems can become increasingly complicated over time. Every audit and near miss can add another form or sign-off. Eventually, the system may make perfect sense to the technical team but become difficult for everyone else to follow.

The test is simple: Does it still work at 3am under pressure? Use clear language, visual controls, colour coding and effective signage. If an operative cannot explain the process in their own words, the issue may be with the procedure rather than the person.

3. Bring In External Eyes

Familiarity can create blind spots. When you walk past the same equipment and processes every day, it becomes easy to stop noticing things that someone unfamiliar with the site would immediately question.

An external perspective does not necessarily mean hiring an expensive consultant. It could be another food manufacturer, technical manager, subject matter expert, or supplier. The important thing is to have someone challenge your assumptions.

4. Create Ownership

There is a difference between responsibility and ownership: responsibility can be written into a job description, but ownership is when people genuinely care about whether the control works. Allergen management should not sit solely with the technical team: production controls the schedule, engineering changes equipment, purchasing manages suppliers, NPD changes recipes, and hygiene manages cleaning. Every function has a role to play. Bring those teams together and make sure everyone understands why their actions matter.

5. Treat Allergen Management as a Journey

Allergen management is not something you fix once and move on from. Factories, products, suppliers, and people all change, and every change can introduce new risk. That means risk assessments need to be revisited, controls need to remain visible and teams need to keep looking for opportunities to improve. The aim is to keep allergen management alive rather than allowing it to become another set of documents that only receives attention before an audit.

Key Takeaways

  • Review risk assessments on the factory floor, not just at a desk
  • Keep processes simple enough to work under pressure
  • Use external perspectives to challenge assumptions
  • Build allergen ownership across the whole business
  • Treat allergen management as continuous improvement, not a one-off project

Webinar Q&A

Do you expect changes to allergen management in BRCGS Issue 10?

Alec explained that he expects developments around Codex allergen thresholds to be considered, although he could not confirm what changes may be made. The existing BRCGS Global Standard for Food Safety already contains requirements around allergen management and precautionary allergen labelling, including the need for a suitable risk assessment.

Why does the UK have more recalls than other countries?

The panel was asked whether higher UK recall numbers indicate weaker controls or a greater willingness to report incidents. Alec said he would not draw a direct conclusion from the numbers alone. Recall data can be influenced by many factors, including how businesses and regulators identify and report issues. The wider point was that businesses should focus on protecting consumers rather than setting a target of having no recalls. A target that discourages reporting could create the wrong behaviours.

Can AI help identify system vulnerabilities?

Wendy explained that AI is already being used in areas including labelling, packaging and raw material checks. The technology is developing quickly and could provide additional tools to support food safety management.

Additional Q&As

We ran out of time to answer all of the questions during the live session, so Alec and Wendy shared their thoughts on the remaining questions below.

Is the allergen control regarding the premix change, managed via the change control process, sufficient?

Having a change control process is an excellent start and ensures that a management system is in place to prompt the review of allergen controls when such changes happen. Of course, it is then a matter of ensuring that the actual review is conducted thoroughly, taking account of any new allergen hazards introduced by the new ingredient or process and ensuring that the correct controls are put in place. This must include processes, procedures and training, to name just a few of the many considerations needed to ensure that change is managed safely. The change should not be approved until the required controls have been implemented and verified across all affected functions.

We use a fish collagen ingredient that is fully hydrolysed (broken down) during processing. The usual fish allergen test we conduct is ELISA for fish protein. The lab says that ELISA only detects parvalbumin, but this protein is destroyed by the hydrolysis process (the raw material we use). So these tests always come back negative, even if we take swabs before cleaning.

Can this happen, and what other ways would you suggest to prove there is no fish allergen risk in an ingredient like this?

One approach that could be considered is to use a generic indicator of the presence of protein, as this will indicate the removal of wider protein residues. In such circumstances, it would be useful to establish that wider protein residues are present prior to cleaning.

For fish testing, we tend to apply DNA-based detection, as it helps with a spread of variety as well as damage due to processing.

Processing of fish products is known to cause issues with detection. Many companies where detection is an issue, due to processing or the presence of low levels, look for another marker to show that the allergen-containing product has been removed, rather than testing for the allergen itself. I would look at the product and see what else could be tested for.

It’s also worth noting that swabbing will always reduce the chance of detection, simply because it requires two transfer steps (onto and from the swab), and transfer efficiency at each step is known to be low. Therefore, it’s important to ensure that final product testing is performed as well.

From a statistical perspective, which parameters should be taken into account when assessing and recording allergen-related incidents? Besides the level of risk and the frequency of occurrence, should the number of people affected also be considered? And how should we account for cases that remain unreported?

If the question is about the considerations to take into account when determining the significance of an allergen contamination event, the key points should include the assessment of the hazard and the likelihood of exposure.

The hazard assessment should include consideration of the quantity present in a portion, given foreseeable consumption circumstances, and the levels likely to elicit a reaction in the susceptible population.

The exposure assessment should include the number of packs in the marketplace and the number of potentially affected packs. This can be informed by reported complaints, but complaint data is an unreliable measure, and the assessment should be driven by the extent of the contamination event in production. Complaint data is useful in triggering an incident investigation, but customers who do not have an allergy may not be motivated to report events that could be dangerous to those who do.

If you are a nut-free site, can you have products with “may contain” labelling?

If you are making a nut-free claim on pack at a nut-free site, then you should not be handling nuts on site, and neither should you use any ingredients that carry a nut precautionary allergen label (PAL) in the nut-free product. This is because, implicitly, if an ingredient has such a warning statement, it should carry through to the finished product, and it is not permissible to have a PAL on a product with a free-from claim, other than gluten-free (as that has a defined level).

There are conceivable circumstances where this might not be the case, based on a risk assessment of carry-over allergens in the ingredient with the PAL statement, and the introduction of allergen thresholds might further change this, but that would need a more detailed risk assessment.

Handling ingredients that have a nut PAL statement at a nut-free site may also be possible where such ingredients are not being used in a product with a nut-free claim. However, this requires a detailed risk assessment to ensure effective separation and management of any potential cross-contamination risks.

Watch the complete webinar recording

So there you have it, the key takeaways from the first webinar in our three-part series with Campden BRI: Allergen Management: Closing the Gap.

  • Part 1: Click here to access the webinar recording and presentation slides.
  • Part 2: Best Practice Cleaning Regimes takes place on Tuesday 20th October and will explore cleaning validation, shared equipment, cross-contact and what good cleaning looks like.
  • Part 3: Food Safety Culture and People takes place on 17th November and will look at the role people and culture play in effective food safety management.

If you have any questions about allergen management or the support we can provide here at Klipspringer, you can contact our team on 01473 461 800 or email sales@klipspringer.com. Alternatively, fill out the form below.

If you have any questions about allergen management and the support we can provide here at Klipspringer, share your contact details below.