Controlling Microorganisms: Problems at the food safety / risk manager interface

The present approach to food safety relies upon implementation of good hygiene practices and the application of HACCP principles against hazards in food. In use for some time almost everywhere around the world, these principles are mandatory in many countries and familiar to the European food industry.

Now, a new approach to food safety is being discussed in the context of the Codex Alimentarius Committee on Food Hygiene. It is based on an entirely different principle: instead of being targeted at hazards in foodstuffs, it is focused on the risk, namely the likelihood and magnitude of impact on public health. Examples of risk articulation are: ‘the risk of disease D (or of mortality due to that disease) caused by the hazard H in country C is X.10-n cases per inhabitant per year’ or ‘Y.10-n per serving of food F’.

This paper discusses the problems that hinder good application of the present approach and explains the new approach, emphasising its originality and problems slowing its development. The difficulties described apply to the situation in Europe within the framework of Hygiene Package Regulations.

The present approach

The present approach is based upon the principle that:

“… where contamination is combated by the food business operators by means of appropriate good hygiene practice (GHP) and hazard control measures through the application of HACCP principles, then an appropriate level of protection of the public health is achieved….”

In other words: ‘GHP/HACCP = safe food’.

Consequently, as illustrated in Figure 1, if the intended level of public health protection is not achieved, GHP/HACCP must be strengthened. The international standard for this approach has been published and updated by the Codex Alimentarius Commission (CAC/RCP 1-1969): ‘Recommended international code of practice – General principles of food hygiene’ including the Annex ‘Hazard analysis and critical control point (HACCP) system and guidelines for its application’.

Implementation

A characteristic of this approach is its ‘downstream character’. It applies to the steps of the food chain taken one-by-one, ‘from farm to table’.

 The first step – from farm to table – is good agriculture practice (GAP), which includes many provisions for hygiene. For example, prevention of field contamination, irrigation with water devoid of faecal contamination, clean milkingparlours, hygienic storage of milk, feeding of animal with safe feed in clean housing, clean transport conditions, appropriate storage and disposal of waste, etc. Yet, there are no critical control points (CCP) at the farm level, which means HACCP cannot be implemented entirely and application of its principles is partial.

It should be noted that GAP is taught in veterinary schools, in animal production and population medicine, so veterinary inspectors can be expected to be aware of GAP.

The next step in the food chain is processing. Application of GHP/HACCP in the food industry is also taught to veterinary inspectors, in food technology and food hygiene, which means they are aware of both GHP and HACCP. However, misapplication of the Codex standards has occurred in some national regulations and food industries.

The first frequent mistake arose when the following were not understood: (a) HACCP Principle 1 ‘conduct a hazard analysis’ has to be done only once the GHP are in place; (b) the next six HACCP principles have to be followed only if the hazard analysis detects hazards remaining despite the implementation of GHP.

The second mistake was to describe as CCP steps that do not have all the requested CCP components: (a) being ‘essential to prevent or eliminate a food safety hazard or reduce it to an acceptable level’; (b) having ‘critical limit(s)’ that (c) are ‘validated’ and (d) monitored through ‘the scheduled measurement or observation relative to the critical limits’ that ‘ideally provides the information in time to make adjustments to ensure control of the process to prevent violating the critical limits’; (e) ‘specific corrective actions must be developed in order to deal with deviations when they occur’.

Few process operation steps can be qualified a CCP. Most often, the safety of food is achieved through GHP alone. Another international standard now recognises this and introduces the ‘operational prerequisite programs’ (oPRP), which are reinforced GHP targeted at specific hazards. This is the norm EN ISO 22000, ‘Food safety management systems -Requirements for any organisation in the food chain’.

A large number of guides to GHP and latterly to GHP and the application of HACCP principles have been published in EU Member States. They provide useful advice and help to improve food safety, regardless of whether they contain the mistakes described above.

These aspects of the EU Hygiene Package regard food processing from an operational perspective and miss a key element, namely hygienic equipment design and engineering. EU Regulation 852/2004 on hygiene of foodstuffs, Annex II, Chapter II, clause 1(f) only alludes to design and engineering: ‘surfaces (including surfaces of equipment) in areas where foods are handled and in particular those in contact with food are to be maintained in a sound condition and be easy to clean and, where necessary, to disinfect. This will require the use of smooth, washable corrosion-resistant and non-toxic materials, unless food business operators can satisfy the competent authority that other materials used are appropriate (including surfaces of equipment) in areas where foods are handled and in particular those in contact with food are to be maintained in a sound condition and be easy to clean and, where necessary, to disinfect’.

 A specific chapter devoted to the food industry in the EU Directive 2006/42 on machinery (Annex 1, section 2) gives more information, such as: ‘all surfaces in contact with foodstuffs or cosmetics or pharmaceutical products, other than surfaces of disposable parts, must:

• be smooth and have neither ridges nor crevices which could harbour organic materials. The same applies to their joinings
• be designed and constructed in such a way as to reduce the projections, edges and recesses of assemblies to a minimum
• be easily cleaned and disinfected where necessary after removing easily dismantled parts; the inside surfaces must have curves with a radius sufficient to allow thorough cleaning’

 The hygienic design and engineering of machinery is not widely taught in veterinary schools and therefore not known to inspectors. Consequently, there is little incentive for equipment manufacturers to comply with these requirements and provide ‘instructions that indicate recommended products and methods for cleaning, disinfecting and rinsing, not only for easily accessible areas but also for areas to which access is impossible or inadvisable’ as stated by the EU Directive. Fortunately, 40 or so EHEDG guidelines are available that deal with hygienic design (see www.ehedg.org).

Listeria monocytogenes may persist for months or years in food factories although it is not specific in terms of its adhesion properties, ability to form biofilms, adaptation to increased sub-lethal concentrations of disinfectants or resistance to lethal concentration. Indeed, persistence occurs in so-called harbourage sites, which are sheltered or inaccessible sites in food premises, and equipment where nutrients are available and the bacteria are able escape cleaning and/or disinfection. This is why it is so important to avoid hard-to-clean places in food premises. Floors and drains are covered by Hygiene Package Regulations and subject to inspection as well as cleaning and disinfection of equipment but their hygienic design is not considered.

The next steps in the food chain are retail shops and consumers’ homes. GHP/HACCP applies to retail but not the home. However, both have in common the temperature-time combination that should ensure the food is safe at the time of consumption. Implicit in the EU Directive 2000/13 on labelling is that all foods are perishable and therefore ‘become unfit for human consumption’ when the time and/or temperature conditions for shelf-life are inappropriate. Except for a few foodstuffs, labels must indicate a ‘date of optimum durability’ (best before). Some foods are more perishable and can contain pathogenic microorganisms, which make the food ‘injurious to health’. According to the EU Directive 2000/13, ‘in the case of foodstuffs which, from the microbiological point of view, are highly perishable and are therefore likely after a short period to constitute an immediate danger to human health, the date of minimum durability shall be replaced by the ‘use by’ date’. Whether ‘unfit for human consumption’ or ‘injurious to health’, foodstuffs are ‘unsafe’ according to EU Regulation 178/2002 Article 14, and shall not be placed on the market.

Can the durability and use by date be predicted with confidence? How can recommendations be established and communicated with competent authorities? How can it be ascertained that the recommendations are understood and respected by consumers?

The EU Regulation 2073/2005 on micro – biological criteria for foodstuffs provides microbiological limits that concentrations of pathogenic micro-organisms must not exceed in ready-to-eat foods, and indications for shelflife in order to respect these limits. Nevertheless, the question remains as to how these limits should be interpreted.

If m represents a mandatory limit concentration, R the result of a microbiological analysis and U the measurement uncertainty, should the rule be: R + U ≤ m, R ≤ mor R ≤ m+ U? The EU failed to provide a harmonised answer so the French agency responsible for food safety assessment, ANSES (formerly AFSSA) recommended R ≤ m, provided the measurement of uncertainty was within recommended boundaries both for safety and process hygiene criteria.

The same EU Regulation 2073/2005 includes process hygiene criteria for Bacillus cereus¸ and some industries have set criteria for Clostridium perfringens. Like Staphylococcus aureus, these species can produce toxins injurious to health when they grow at concentrations much higher than the limit of the microbiological criteria. The French authorities, therefore, established ‘alert limits’ but there is no progress at the EU level in establishing harmonised recommendations.

Criticisms at the present approach

From an industry standpoint, one may ask if GHP, oPRP or CCP could be less stringent, if they are all needed, or if some of them should be more stringent. For responsible authorities, it is difficult to establish a link between GHP/HACCP and the number of cases of the specific diseases that these measures aim at reducing. In other words, the industry effort towards hygiene is not optimal, and the link between this and public health is not clear.

The emerging approach

 The incentive for the new approach was the ‘International Agreement on the Application of Sanitary and Phytosanitary Measures’ (SPS Agreement) enforced on 1 January 1995 as part of the General Agreement on Tariffs and Trade (GATT). The SPS Agreement states the settlement of international trade disputes about food safety must be based on scientific and transparent assessment of the risk, according to the methodology standardised by the Codex Alimentarius Commission. The ‘appropriate level of protection’ (ALOP) for public health in the importing country, also termed ‘acceptable risk’, must not be jeopardised by imported foodstuffs.

The focus of the new approach to food safety is the risk, based upon the following principle:

“The level of risk of adverse health effects, as determined by the competent authority (the risk manager), must be attained by relevant good hygiene practice and hazard control measures.”

In other words, the appropriate level of public health dictates which GHP/HACCP must be applied.

Consequently, as illustrated in Figure 2, if the ALOP is not achieved or deteriorates, measures shall be taken to reach or recover the acceptable risk. After a few years lag, the Codex Alimentarius Commission (CAC/GL 63-2007) published ‘Principles and guidelines for the conduct of microbiological risk management (MRM)’ including the Annex ‘Guidance on microbiological risk management metrics’.

Implementation

A characteristic of this approach is its ‘upstream character’. It applies ‘from the public health target to the weapons’ (from aim to arms) that have to be used against the hazards.

The approach is based upon the knowledge of the relationship between:

• Dose – which is the amount (not the concentration) of hazard ingested in a single serving 
• Response – which is the probability of illness cases caused in the consumer population by the hazard ingested

The probability of illness should not exceed the acceptable risk. Because of the dose-response relationship, and knowing the portion size, the risk manager is able to derive the maximum dose of the hazard that is tolerable in a food at the time of consumption. This dose, expressed as a concentration or as a frequency of presence in the foodstuff is called the ‘food safety objective’ (FSO). The EU Regulation 2073/2005 on microbiological criteria for foodstuffs includes microbiological limits for Listeria monocytogenes and Salmonella spp. in ready-to-eat foods ‘placed on the market during their shelf-life’, which is until consumption. Thus, these limits are FSO, as acknowledged in the Annex ‘Microbiological criteria for Listeria monocytogenes in ready-toeat foods’ from the ‘Guidelines on the application of general principles of food hygiene to the control of Listeria monocytogenes in foods’ (CAC/GL 61-2007).

The FSO has to be achieved by retailing servings that contain less than a maximum concentration and/or frequency of the hazard. This maximum (performance objective, PO) is at the point where the foodstuff is put on the market. The EU Regulation 2073/2005 provides microbiological criteria to verify the PO is obeyed. The difference between PO and FSO is the predicted evolution of microbiological flora from the moment the food is placed on the retail shop shelf to the point it is consumed. Other POs may be established for steps earlier in the food chain. To achieve the FSO and the various POs, and taking into account the hazard load of ingredients and contaminants, the food business operator has to carry out appropriate process criteria (e.g. time, temperature of heat treatment or of storage) and/or product criteria (e.g. acidity, water activity). What is important is the link between hygiene practices and hazard control measures (CCP with their critical limits and oPRP), and the targets described in the PO and FSO. Compliance with some of the objectives can be verified with the help of microbiological criteria, as shown in Figure 3).

 The implementation of this emerging approach is based on dose-response relationships, several of which have been published. Those resulting from international agreement have been established by the Joint FAO/WHO Experts Meetings on Microbiological Risk Assessment (JEMRA, http://www.fao.org/ag/agn/agns/jemra_riskasse ssment_en.asp or http://www.who.int/foodsafety /micro/jemra/assessment/en/index.html).

Predictive microbiology is already used to model microbiological behaviour from retail to consumption but steps earlier in the process also need to be modelled. Several modelling techniques are available such as Monte Carlo simulation or Bayesian networks. These take into account the uncertainty (lack of solid knowledge) and the variability (inherent to any natural phenomenon). Models that have been published to-date can be accessed at: http://foodrisk.org/risk_analysis/RA/RAs.cfm.

It is important to note that any GHP and hazard control measures can be incorporated into these models: those applied by food business operators as well as intervention by national and international authorities such as inspection, sampling, microbiological limits, withdrawal, etc. The greatest advantage of risk assessment modelling resides in the ability to simulate a range of scenarios and, by the means of sensitivity or importance analysis, understand where control measure(s) will be most effective or cost-effective.

Outcomes of food safety models are used by risk managers. One example, among many, based on published risk assessments is the cited Annex ‘Microbiological criteria for Listeria monocytogenes in ready-to-eat foods’ from the ‘Guidelines on the application of general principles of food hygiene to the control of Listeria monocytogenes in foods’ (CAC/GL 61- 2007). It is worth mentioning that scientists in public institutions, notably agencies responsible for food safety, specialised in quantitative risk modelling in almost all European countries and collaborate with the European Food Safety Authority (EFSA). Administrations in the USA, Canada, Australia, New Zealand, etc., multinational companies and some professional associations have also established permanent risk assessor teams.

Criticisms at the emerging approach

The emerging approach involves many personmonths. It requires a lot of data to be collected on, for example, concentration or frequency of hazards in foods, microbiological growth, inactivation kinetics, and dose-response relationships for different susceptible populations. Specific modelling skills are also a prerequisite. Experience is needed to cope with uncertainty, make reasonable approximations or simplifications, devise easy communication tools targeted at the industry and risk managers, etc. The emerging approach is, therefore, expansive and time consuming, and is not adapted to circumstances where decisions must be made quickly. Only a small number of countries have articulated explicitly their ALOP. The number where new food safety goals have been enunciated is even smaller. As a consequence, few food business operators, and we dare say few state employees are informed about emerging metrics, ALOP, FSO or PO.

Conclusion

The control of food safety systems is on the verge of a revolution as it moves from a ‘downward’ to an ‘upward’ approach. It is not an exaggeration to say the paradigm is shifting from ‘hygiene empirically ensures safety’ to ‘an acceptable degree of safety is derived through scientifically optimised hygiene measures’. It should not be forgotten that the present approach has proven efficient, and there is a long way to go before common use of the new approach because of the technical and scientific challenges as well as associated costs. More university courses in risk assessment will have to be created to ensure the new approach becomes an obvious and natural way of thinking, and new skills taught to veterinarians and food equipment engineers, alike. However, the train is already in motion, and the revolution will take place.

Acknowledgements

he author thanks Huub Lelieveld and Siân Astley of the EFFoST Editorial Support Team for their help in revising and improving the manuscript.

About the Author

Olivier Cerf has an MSc in Agronomy and PhD in Dairy Science. For thirty years, he was a scientist at the National Institute for Agricultural Research (INRA) as Head of the Laboratory of Engineering of Food Process Hygiene, before becoming Professor of Food Hygiene at Alfort Veterinary School (ENVA). He is the Vice-Chair, Scientific Panel on Microbiology, French Agency for Food, Environmental and Occupational Health Safety (ANSES). Olivier participated in the work on the emerging metrics for the management of microbial risks of the Codex Committee on Food Hygiene.