- Print Friendly - Help

Food Safety in New Zealand: Application of a Risk Management Framework

NZFSA’s Risk Management Framework

Effective application of risk analysis in food safety is dependent on agreed principles and processes. The most important aspect of risk management in this regard is the consistent and transparent application of a Risk Management Framework (RMF) to all food safety issues.

While inputs may vary substantially for each issue, risk managers initiate the risk management process and see it through to completion. The benefits of systematic application of an RMF include:

• establishment of food control systems that are risk-based and achieve required levels of consumer protection

• regulatory decisions that are proportionate to the health risks involved

• providing for innovation and flexibility in application of control measures

• allowing due regard to be taken of the costs as well as benefits of regulatory activities.

An RMF facilitates interaction between government, industry, consumers and other stakeholders on many levels and NZFSA can act in a consultative manner that is independent of sector interests. Science can be appropriately merged with other inputs in the development of standards and other risk management activities.

Risk management decisions made by international organisations increasingly influence the NZFSA regulatory environment and impact directly on the domestic marketplace. This creates a strong impetus for an NZFSA RMF that uses risk analysis guidelines adopted by international agencies.

The RMF is a four-step process for risk managers to work through all food safety issues as they arise. This includes the periodic review of old standards and the evaluation of scientific information on new or ‘re-emerging’ hazards that may require regulatory action by NZFSA.

AN RMF provides a systematic means whereby knowledge on risk, and evaluation of other relevant factors, are used to choose and implement appropriate food control measures. While all stakeholders have a role, NZFSA as the risk manager is the central player. The process is cyclical, iterative and on-going, with monitoring and review likely to lead to new control measures over time (see Figure 3).

Figure 3: Components of the RMF

Sources of NZFSA’s RMF Food Administration in New Zealand: A Risk Management Framework for Food Safety, Joint Ministry of Health and Ministry of Agriculture and Forestry Food Harmonisation Project, 2000, MOH Wellington. Food Safety Risk Analysis: A Guide for National Food Safety Authorities, FAO Food and Nutrition Paper No.87, 2006, WHO/FAO Rome. Working Principles for Risk Analysis for Application by National Governments, CAC Procedural Manual, 17th Edition, 2007, FAO Rome. FAO Biosecurity ToolKit, 2007, FAO Rome. Principles and Guidelines for the Conduct of Microbiological Risk Management (MRM), CAC/GL 2007, FAO, Rome.

Role of science

An international consensus has now developed that, to the extent practicable, risk assessment should be functionally separate from the standard-setting process carried out by risk managers. The intent of this is to protect the integrity of risk assessment as an objective and unbiased scientific activity.

NZFSA, along with a number of other Food Safety Authorities, has reinforced this functional separation in its organisational structure. The Science Group is the key repository of scientific expertise in NZFSA and provides scientific advice to the other business groups. Where necessary, the Science Group contracts scientific inputs from external providers. Participation in international standard-setting organisations and their working groups is another important source of cutting-edge scientific information for the Science Group.

Scientific inputs are required at all steps in the RMF (as shown in Figure 4).

Figure 4: Scientific inputs to the RMF

Step 1: Preliminary risk management activities

This initial step in the RMF consists of a number of tasks (see Figure 5):

• identification of food safety issues

• risk profiling

• establishing broad risk management goals

• deciding on the need for a risk assessment

• if needed, setting risk assessment policy and commissioning of the risk assessment

• considering the results of the risk assessment

• ranking and prioritisation of the food safety issue for risk management action.

Food safety issues that require a risk management response arise from many sources but most are identified by NZFSA’s ongoing activities. External sources include new concerns raised by consumers, requests from industry for the evaluation of new food production and processing technologies, new hazards identified by the global scientific community, and equivalence discussions with trading partners.

Risk profiling provides an opportunity to gather relevant information on an issue and this provides a lead for further action. Each risk profile should be fit-for-purpose – in some situations a very elemental exercise. Components that may appear in a risk profile include:

• description of the food and the food chain scenario

• the biological or chemical characteristics of the hazard

• assembly of scientific information on possible risks

• identification of gaps in scientific knowledge

• description of current control measures

• WTO SPS Agreement implications.

Risk profiling may be used directly by risk managers to guide identification and selection of risk management options, for example where:

• rapid action is needed

• there is sufficient scientific information for action

• embarking on a risk assessment is impractical.

Figure 5: Individual tasks at each step in the RMF

NZFSA risk profiles Working closely with ESR, NZFSA has commissioned an extensive set of risk profiles for hazard/ food commodity combinations of importance in New Zealand. These include: • Bacillus spp. in rice • Campylobacter jejuni/coli in poultry • Ciguatoxins in seafood • Clostridium botulinum in ready-to-eat smoked seafood in sealed packaging • Listeria monocytogenes in soft cheeses • Salmonella (non-typhoid) in and on eggs • Shiga-like toxin-producing Escherichia coli in uncooked comminuted fermented meat products • mycotoxins in the New Zealand food supply • natural toxins in crop plants. These risk profiles have guided a range of standard-setting and other regulatory activities within the RMF process. The profiles constitute a ‘living library’ of up-to-date scientific information specific to New Zealand and are reviewed every five years. Risk profiles commissioned by NZFSA

Following the risk profile, risk managers need to decide on broad risk management goals. This is likely to occur in conjunction with a decision on whether or not a risk assessment is feasible and necessary but should precede actual commissioning of a risk assessment. The broad risk management goals will help direct the scope of a risk assessment and will likely be refined when the outputs of risk assessment are known.

At this stage of the RMF, it is important to recognise that many food safety issues can be actioned without commissioning a risk assessment. Rather, more intense scientific evaluation may be sought and other ways of developing available information on risk may be brought into play, eg food source attribution data derived from surveillance of foodborne illnesses.

Scientific evaluation of aspartame Public concerns arose during 2007 in New Zealand over the use of aspartame as an artificial sweetener in foods, with claims of adverse health effects being widely expressed in public media. A rapid response from NZFSA was needed to allay public fears and a full review of scientific information available from other Food Safety Authorities was quickly undertaken. Of particular importance was the scientific report on aspartame published by the European Food Safety Authority in 2002 and an evaluation by NZFSA’s Science Group of the studies published by the European Ramazzini Foundation in 2005. NZFSA found no scientific evidence in support of the new public claims of risks to human health and initiated a risk communication programme to this effect. Commissioning new risk assessment work in New Zealand was not needed.

During risk assessment, scientific judgements often entail a choice among several reasonable options. Thus gaps in scientific knowledge are bridged through a set of inferences that are called ‘risk assessment policy’. Risk assessment policies are usually generic and are established by risk managers in consultation with risk assessors. They preferably should be established before a risk assessment commences.

‘Safety factors’ in chemical risk assessment The application of internationally accepted default ‘safety factors’ by NZFSA when estimating acceptable daily intakes for chemical residues in foods is an example of risk assessment policy. Animal exposure studies are used to determine ‘no observed adverse effect levels’ and then a safety factor of 10 is applied in case there is any biological variation. An additional factor of 10 is applied in case there is any interspecies variation when extrapolating from the animal test species to humans.

If it is decided to commission a risk assessment, the risk manager should clearly define, in association with the risk assessors, the scope, purpose and expected outputs. The required resources should also be agreed, and in some cases simple projects will be able to be undertaken by individuals. NZFSA may have to contract scientific research to fill data gaps as the risk assessment proceeds.

Risk assessments commissioned by NZFSA Recent risk assessments have determined risks associated with: • Salmonella Brandenberg contamination of sheep meat • Cysticercus bovis in domestic and exported beef • Salmonella spp. in imported fresh broiler chicken meat • imported Roquefort cheese (made from unpasteurised sheep milk).

NZFSA Campylobacter in Poultry Strategy 2007–2010: A combination of scientific evaluations and risk assessment within an RMF process Following development of a comprehensive risk profile on Campylobacter jejuni/coli in poultry, NZFSA has implemented a detailed Campylobacter Strategy with the goal of significantly reducing foodborne risks to consumers from broiler chickens. NZFSA’s Campylobacter in Poultry Risk Management Strategy Because of the severity of the problem, preliminary risk management activities cover a range of scientific projects. Short-term responses include scientific collaboration with industry to develop codes of hygienic practice for both producers and processors. Medium-term responses include scientific evaluation of the likely level of hazard control associated with a number of hazard-based interventions. Meanwhile, a farm-to-plate risk assessment model is being developed so that risk-based controls that achieve agreed levels of consumer protection can be regulated in the longer term.

Proper interpretation of the outputs of the risk assessment by the NZFSA risk manager is a vital function. Risk assessors should clearly describe the uncertainty in a risk estimate and its origins. The overall strengths and weaknesses of the risk assessment should be discussed and documentation should include a general summary that is easily understandable by lay stakeholders.

Ranking of food safety issues for risk management action can take place at different stages during preliminary risk management activities. While ranking is essentially a scientific exercise, prioritisation of issues is an NZFSA management role. New work may be prioritised according to drivers other than the rank of food safety risk, eg consumer interest and/or political concerns within New Zealand, or as periodically happens, disputes over international market access.

Domestic Food Review NZFSA is undertaking a comprehensive long-term review of regulatory involvement in the domestic food sector. An important part of the project is application of risk ranking and prioritisation methodology in the early stages. The findings from the ranking and prioritisation models will provide a basis for the transition and implementation plan. NZFSA’s ‘Domestic Food Review’

Step 2: Identification and selection of risk management options

In the second step, potential risk management options are identified and then selected by risk managers according to appropriate criteria. NZFSA strives to involve all stakeholders to the extent possible, as well as providing a clear rationale for the final decisions taken. As a general principle, all parts of the food chain should be taken into account when selecting control measures.

Possible control measures for Campylobacter in broilers The NZFSA Campylobacter Strategy identifies numerous control measures that need to be evaluated in the New Zealand context and these include: • quarantine of the production environment • decontamination of drinking water • testing of flocks prior to slaughter • biological decontamination (bacteriophages) • improved process hygiene • chemical decontamination of carcasses • performance targets for chilled carcasses (see below) • leak-proof packaging at retail • consumer education.

During identification and selection of risk management options, risk managers will likely have asked the risk assessors to examine the impact of different control measures on minimising risks. This process may continue until one or more risk management options that achieve the desired level of consumer protection are chosen.

Level of consumer protection

Establishing the level of consumer protection to be achieved by the control measures is a core part of Step 2 in the RMF process. Decisions can be influenced by a wide range of economic, political, social and environmental factors.

Desired levels of consumer protection can be expressed in a number of ways. The tolerable number of cases of illness due to a particular hazard in a food in a particular population over a specific time period may be used, eg no more than 1 case of disease Y per 100,000 people per year in the general population. However, it is more likely that NZFSA will express the desired public health goal in terms of a percentage improvement over current (unacceptable) levels. In other situations, the risk per edible portion of a food is a useful parameter to anchor a decision on control measures. Measurement of the societal impact of a foodborne disease, eg using disability-adjusted life years (DALYs) as a comparative unit, provides a means of comparing risks from disparate sources when deciding on a desired level of consumer protection.

Examples of approaches to establishing levels of consumer protection The NZFSA Campylobacter Strategy incorporates a risk-balancing approach, ie cost-benefit analysis and an ‘as-low-as-reasonably-achievable’ (ALARA) level of risk reduction, in significantly reducing Campylobacter and Salmonella contamination of broiler chickens in modern processing systems. This approach is reflected in NZFSA’s goal to have reduced foodborne Campylobacter illnesses by 50% by 2010. NZFSA’s Campylobacter in Poultry Risk Management Strategy Direct comparison of risks using surveillance data and food attribution studies were used to apportion risks from different Salmonella serotypes in a range of foods and prioritise those at unacceptable levels for specific food chain interventions. NZFSA’s Science Report ‘Modelling of exposure of New Zealanders to Salmonella’ NZFSA uses a generic ‘notional zero-risk’ as the required level of consumer protection for chemicals that are intentionally added to the food supply such as food additives or veterinary drugs, ie standards are developed on the basis that any allowable residues can be ingested daily over a lifetime without any appreciable health risk. In the case of threshold approaches, eg potentially carcinogenic chemicals in the food supply, the generic level of consumer protection is no more than one additional case of disease above background per million consumers. This was how the NZFSA standard for residues of xylazine metabolytes in deer velvet was set.

Costs and benefits associated with a risk management scenario need to be evaluated in an understandable and transparent manner. As well as economic analysis, the technical feasibility and practicality of available risk management options must be appropriately evaluated. This includes the ability to verify and enforce any regulatory standards that may be decided upon.

Performance objectives as risk management options

Where microbial hazards exist continuously in a food chain, risk-based control measures can benefit from the establishment of regulatory ‘targets’ that are called performance objectives. A performance objective is a quantitative expression of the frequency and/or concentration of a hazard in a food at a specified step in a food chain that should not be exceeded if the required level of consumer protection is to be met. It is envisaged that use of risk assessments within an RMF will lead to decisions on performance objectives that provide considerably increased flexibility to industry in design of food hygiene programmes.

However, Food Safety Authorities around the world are finding it difficult to reach public policy decisions on acceptable levels of consumer protection for commonly occurring foodborne illnesses, eg those due to Campylobacter and Salmonella, which are a necessary input to setting performance objectives. Decision-making on acceptable levels of consumer protection for severe foodborne illnesses of very low frequency, eg those due to E. coli O157:H7 and E. sakazakii, is even more difficult. (This is in contrast to decisions on risks from certain chemical hazards in the food supply, such as agricultural compounds and food additives, where a predetermined ‘notionally zero risk’ policy is the norm.)

On the other hand, Food Safety Authorities in a number of countries are leaning towards setting food safety goals that reflect continuous improvement in levels of consumer protection. In striving to achieve these goals, regulatory targets that are hazard-based rather than risk-based are set at specific steps in the food chain. Systematic application of an RMF and improving attribution surveillance data allows the risk manager to monitor progress and modify targets as needed. If continuous improvement in consumer health is not achieved, the stringency of hazard-based targets can be increased. The disadvantage of this approach is that in the absence of risk assessment, there is no opportunity to compare the effectiveness of different targets (or different control measures) in achieving consumer health goals.

NZFSA Campylobacter Strategy 2007–2010 The NZFSA Campylobacter Strategy seeks to significantly reduce foodborne risks to consumers from broiler chickens. A number of risk management options are being developed, including GHP-based, hazard-based and risk-based control measures. In early 2008, NZFSA regulated a hazard-based target in the form of a quantitative limit for Campylobacter on chilled carcasses. This is an interim response to a severe public health problem. When the farm-to-plate risk assessment has been completed, a risk-based evaluation of available control measures will be undertaken and a performance objective established on this basis. NZFSA’s Campylobacter in Poultry Risk Management Strategy

International trade

In imported food situations, the WTO SPS Agreement places specific constraints on factors that can be included in decisions on ‘appropriate levels of protection’ (ALOP) that are chosen by NZFSA. Decisions should take into account the minimisation of trade effects and ensure that selected control measures are not more restrictive than necessary to meet an ALOP. NZFSA must also avoid unjustifiable or arbitrary distinctions in levels of ALOP chosen in different food safety situations.

NZFSA policy on equivalence Judgement of the equivalence of different food safety control measures for exported food is of vital importance to New Zealand. Where food standards in an exporting country differ from those in an importing country, the WTO SPS Agreement states that “Members shall accept the sanitary measures of other Members as equivalent, even if these measures differ from their own or those used by other Members trading in the same product, if the exporting Member objectively demonstrates to the importing Member that its measures achieve the importing Member’s appropriate level of sanitary protection”. Systematic use of the RMF has allowed NZFSA to achieve judgements of equivalence for a significant number of alternative, cost-effective food safety control measures for exported food. In the case of the cattle tapeworm Taenia saginata, traditional and labour-intensive post-mortem meat inspection procedures were historically imposed on New Zealand at high cost to industry. A risk assessment was able to show that alternative and highly cost-effective risk management options, ie lesser intensity meat inspection and a detailed trace-back and farm quarantine system, achieved the same level of consumer protection.

Uncertainty

Uncertainty is intrinsic to risk analysis and a precautionary approach to food safety is expressed in various ways during risk assessment and risk management activities. Precautionary positions may be intrinsic to risk assessment policy, eg use of safety factors in establishment of acceptable daily intakes for chemical residues in food, or may be introduced on a case-by-case basis, eg ‘worst-case’ modelling scenarios where pathogens have a low infective dose and/or severe adverse health consequences.

Different approaches may be taken to risk management in the face of scientific uncertainty in different political, social and economic contexts. In some cases, consumer fears have driven actual bans on trade even though this was not scientifically supported by international standard-setting processes, eg when the European Union banned the importation of hormone-treated beef from all countries including New Zealand. In other cases, a conservative approach to standard setting may be taken by NZFSA if the ramifications of a single detection of a high-profile pathogen (eg E. coli O157: H7) in exported product might include a worst-case reaction from trading partners.

NZFSA does not have a specific policy on application of a precautionary approach in the face of scientific uncertainty. Rather, there is an understanding that incorporation of precaution in the RMF will be rational, practical and based on scientific principles. If there is likely to be a significant risk to human health from a particular hazard or situation, NZFSA will take appropriate risk management action that is proportional to: the potential risk, the consequences of the risk management option(s) chosen, and the degree of uncertainty in the scientific evaluation. The regulatory response will prevent or limit exposure while more conclusive information is gained on the actual risks faced and the control measures that are likely to be most effective. For products in trade, there is an obligation under the WTO SPS Agreement to actively pursue additional scientific information when a precautionary approach is taken, with timely review of interim control measures.

Pyrrolizidine alkaloids in honey Monitoring by NZFSA as part of the National Residue Monitoring Programme recently showed that the speciality vipers bugloss honey was contaminated with pyrrolizidine alkaloids. There are more than 100 different types of these alkaloids arising from different flowers and while the data shows that the individual substances have quite varied toxicity, information across the whole spectrum is very limited. In view of these gaps, a precautionary approach to risk management was taken. The chemical risk assessment used data from the most toxic of the known alkaloids to establish an acceptable daily intake (ADI) and the monitoring information from honey was used to assess a worst-case exposure scenario. It was found that even with this precautionary approach, there was no appreciable risk to New Zealand consumers given the level of exposure in the New Zealand diet and therefore no standards have been set.

Shiga toxin-producing E. coli in uncooked comminuted fermented meat E. coli O157:H7 and other shiga toxin-producing pathogens in uncooked comminuted fermented meat have caused severe illness in a number of countries. However, surveillance data to date has revealed no such cases in New Zealand and contamination levels in fresh beef and pork are very low. Despite this, a risk-based processing standard that mirrors overseas standards for this product has been put in place as a precautionary measure. The processor is required to monitor the microbiological quality of raw materials and apply processing parameters that are sufficient to inactivate any pathogens that may be present.

Risk management of genetically modified foods also illustrates a precautionary approach. At present, only a limited number of these foods are traded across New Zealand borders and this is largely a consequence of precaution by risk managers because of the high perceptions of risk by some consumers. In some cases, partial risk assessments and incomplete substantiation of the benefits and risks of genetically modified food have resulted in controversy over their safe use and their safe release into the environment.

Step 3: Implementation of control measures

Industry working throughout the food chain has the primary role in implementation of control measures. Nevertheless, NZFSA may be directly involved in this step in the RMF, such as in supervisory meat inspection. More often, the Verification Agency of NZFSA will verify control measures implemented by industry. The Compliance Group of NZFSA carries out an independent audit of regulatory functions and applies sanctions where control measures have not been properly implemented by industry.

The Approvals and Registration Group of NZFSA also carries out an implementation function by registering food premises and approving food safety plans developed by industry.

NZFSA often develops implementation tools to assist stakeholders in implementing regulatory requirements. Examples are generic codes of hygienic practice for different food commodities, guidelines on quality assurance systems, accreditation systems for laboratories, and assisting with training and education.

Farm-to-plate approaches to food safety promote design of integrated food safety programmes that make the best use of industry and government food safety resources. The NZFSA Campylobacter Strategy is a good example of an alliance between the poultry industry and the regulator to implement the most effective control measures in an integrated manner.

Step 4: Monitoring and review

The aim of monitoring by NZFSA is to gather and analyse data on the level of control of specific hazards throughout the food chain and combine this with human health surveillance data to determine the effectiveness of regulatory activities. This may be carried out ahead of implementation of control measures so as to establish baseline levels or it may follow their implementation.

NZFSA has an extensive programme for monitoring hazard levels in the food chain compared with many countries.

Monitoring of the food chain by NZFSA Monitoring programmes carried out by NZFSA include: • National Microbiological Database for systematic and ongoing monitoring of premises slaughtering cattle, sheep, deer, goats, poultry and ostriches • Food Residue Surveillance Programme for compliance with chemical food safety standards across a wide range of foods • Total Diet Survey for evaluating the level of exposure of the New Zealand population to chemicals in the food supply • National Residue Monitoring Programme for exported foods.

Evaluating data on hazards and risks on a periodic basis provides NZFSA with information on the effectiveness of their risk management decisions and actions. It also helps to identify new food safety problems as they emerge. For imported foods, it is not possible to check every unit or lot in a consignment for the presence of hazards and monitoring programmes in the country of origin are sometimes used by NZFSA as a means to improve the limited assurance that can be gained from sampling plans applied at the border.

Human health surveillance National human health surveillance activities administered by the New Zealand Ministry of Health are an important part of monitoring and review by NZFSA. In some cases, NZFSA will initiate and fund sentinel site studies where data on specific hazards in the food chain is lacking and the resultant level of foodborne disease is unknown. In liaison with ESR, NZFSA assists in analysis of human health statistics on gastrointestinal illness and has developed food source attribution factors for the most important pathogens, eg an expert elicitation workshop convened by NZFSA developed attribution factors for foodborne campylobacteriosis, salmonellosis, listeriosis and norovirus in New Zealand of 0.6, 0.6, 0.8 and 0.4 respectively.

While not quantitatively linked to foodborne risks per se, there are many opportunities to use monitoring of the food chain to demonstrate that control measures have prevented the level of exposure to biological and chemical hazards from increasing. In other situations, planned reduction in levels of exposure to specifi ed hazards can be demonstrated. The NZFSA National Microbiological Database, continuously applied to all of the major meat slaughter species, is showing gradual improvements over time for process hygiene indicators and gradual reductions in levels of microbial pathogens originating on the farm.

Where monitoring of hazards or risks indicates that food safety goals are not being achieved, risk management strategies and/or control measures will need to be reviewed. Review may also be required when new information on hazards and/or risks arises or new market access requirements are imposed.

Monitoring for pyrrolizidine alkaloids Monitoring by NZFSA has shown that exposure to pyrrolizidine alkaloids in honey is highly unlikely to constitute a health risk to New Zealand consumers. However, it is possible that these compounds are also present in other food types, eg cereals, and exposure across a total diet might possibly breach the acceptable daily intake. Consequently, NZFSA has initiated a wider monitoring programme to support its risk management decision to not set a maximum limit for the speciality vipers bugloss honey.

Monitoring and review is greatly enhanced by effective communication networks and linkages with offshore Food Safety Authorities. Trading agreements often contain obligations on monitoring and where possible, NZFSA links up with international organisations which operate early warning systems for foodborne disease, eg the World Health Organization INFOSAN system.