Food allergy incidents are rising. How can food manufacturing improve?

Technology, regulation and a culture of safety all have a role to play as part of a joined-up approach.

Image: Adobe Stock.

By Lili Jia

The growing prevalence of food allergies has moved to the top of the food safety agenda, with prominent news coverage raising awareness of this worsening problem.

In the UK, admissions to hospital for allergic reactions to food increased from 1.23 to 4.04 per 100,000 people per year between 1998 and 2018. Among them, the largest increase of hospital admissions was children under 15 years old, jumping from 2.1 to 9.2 per 100,000 population per year.

There is no cure for food allergies and the only way to prevent allergic reactions is to avoid trigger allergens. This is why mandatory declaration of food allergies has been included in international and national regulations, requiring food manufacturers to provide full ingredient and food allergen information on the product label.

But despite the health and legal importance of declaring allergens, the number of incidents continues to grow.

As part of recent research, we collected all the food allergy alerts between 2016 and 2019 from the Food Standards Agency for England, Wales and Northern Ireland, and from Food Standards Scotland and National Archives websites.

Among the total 435 food allergy alerts collected, almost all were down to production problems: mislabelling accounted for 54% of the total, incorrect packaging for 19%, food allergen contamination for 14%, and a lack of English labelling (the allergens not mentioned in English on the label) for 8%. We found varying but similar errors in studies across Australia, the EU, Hong Kong, New Zealand and the US.

Unfortunately, the incentive to adopt good practice is generally low among food manufacturers: many are not willing to go beyond the minimal certification requirements, assess their food safety culture or implement food traceability technologies proactively. And cost pressures mean that quality control is suffering at a time when it is needed more than ever.

There are a number of initiatives that can help address these problems. Intelligent food traceability technologies that use machine learning can support operational improvements. Better and tighter regulation can increase the incentives for manufacturers to invest proactively. And a better food safety culture could help avoid errors where communication and coordination between operators has failed.

But these cannot work in isolation. For food manufacturers to improve, it is important to look at the incentives of all the operators involved, and how these incentives can reinforce each other.

Good leadership is an important place to start. Creating a clear vision of food allergen management, with recognition for good practice, can help operators develop shared values and collaborate more efficiently to avoid food allergy incidents.

Small changes to the operational environment can have a big impact — for example, colour-coded labels on production machines could make it easier and quicker for operators to identify the production lines dedicated to a specific allergen profile.

A good understanding of food allergen management can also prevent potential risks. If facilities are set to avoid the crossover of open production lines (such as conveyor belts), they can prevent food allergen contamination from spillage and reduce the risk of contamination.

But without a good knowledge of best practice, operators will have difficulty mitigating the risks of food allergen cross-contact and avoiding airborne cross-contamination. A survey in the Republic of Ireland has shown that only 13% of food operators could list the 14 allergens identified in the relevant EU regulation and 28% of these food operators did not receive any food safety training in Ireland. More than 50% of the workers surveyed did not know that celery, peanuts, mustard, sulphites, sesame seeds, lupin and molluscs are food allergens.

Inconsistent food allergen labelling regulations across countries have also created difficulties. There are eight food allergens that are mandatory to declare for international trade. In the US, there are also eight food allergens, yet they include wheat rather than different cereal sources of gluten like the UN’s Food and Agriculture Organization and the EU. EU regulations themselves specify 14 food allergens that are mandatory to label.

The different labelling regulations for imported ingredients or materials may cause confusion to operators and result in undeclared food allergens. More consistent regulation or a better food traceability system that eliminates confusion would assist manufacturers to manage allergens more effectively.

A food manufacturer’s allergen management is also constrained by its suppliers. For example, when the ingredient suppliers fail to label all the food allergens or use a confusing label, it is difficult or impossible for the food manufacturer to declare all the allergens correctly.

A broader, incentive-based system can help to increase awareness of potential risks. When some common errors are unknown to a food manufacturer, it may make the same mistakes that others made before. Allowing food manufacturers to examine their practices against common errors can raise awareness of poor practices and encourage them to take actions to improve their performance before the outbreak of any food allergy.

It can also facilitate collaboration. A shareable framework for all relevant stakeholders can contribute to the sharing of best practice and the improvement food allergen management along the supply chain.

To make progress across all these issues, technology can help. As manufacturers collect more operational data, this could be used to benchmark the performance of food allergen management, and machine learning algorithms could help to more efficiently tailor incentives for better performance.

This is different from the use of machine learning in conventional automation and food traceability technologies, where machine learning is used independently and aims to replace human beings. Instead, machine learning can be used as a prediction tool to provide information to operators, instead of making decisions for them.

Focusing on incentives for manufacturers to improve can also assist the design of more effective food allergen policies and regulations. Existing regulatory systems rely heavily on on-site inspections, which are costly, and the information collected is limited. An incentive-based regulatory system could transform the existing regulatory systems to more efficient, intelligent, and lower-cost systems.

With the increasing prevalence of food allergies and the severity of food allergy incidents around the world, improving food allergen management in manufacturing has become more prominent than ever. An incentive-based approach built around advanced food traceability technologies, more effective regulations and better food safety culture could begin to answer this growing problem.

IfM Insights