SCIENCE FOR DECISION MAKERS - data.daff.gov.au

S C I E N C E F O R D E C I S I O N M A K E R S

Risk AssessmentR i s k A s s e s s m e n tin Food Safety Policy and Practice

S. U. Fabiansson

Bureau of Rural Sciences

BRS/Risk Assess Cover.FA 11/7/01 3:18 PM Page 1


S. U. Fabiansson


Risk Assessment Doc.FA 11/7/01 3:13 PM Page i

Hazards and risks are now commonly used words in industry and government when

dealing with food safety issues. To create uniform ground, the terminology has been

defined by international agreement, and several studies initiated to explore food

safety hazards and their associated risks in more depth. Results point to the

increased importance of a systematic approach when attending to food safety

problems. 'Risk assessment' can provide that systematic approach.

This report describes the relationship between hazards and food safety risks and

explains how objective risk assessment in the future will be able to assist in the

refinement of HACCP-based quality systems. Microbial risk assessment is in a stage

of rapid evolution. We do not suggest that individual companies should immediately

adopt the methodology before it has been proven in practice. However, it is

important both for industry and for governments to follow developments and

be ready to implement new techniques as they become available.

Risk assessment is an objective, science-based method. As such, specialists are

needed to gather information, construct appropriate models and complete the

analysis. However, custom-made solutions are starting to become available for

industry application as well. This report aims to provide sufficient information

to give an understanding of the intricacies of risk assessment, its potential for

practical application, and links to further information sources for more in-depth

descriptions. It provides some practical examples of tools for risk-ranking in order

to set priorities for further action.

Dr Peter O’BrienExecutive Director

© Commonwealth of Australia

This work is copyright. The Copyright Act 1968 permits fair dealing for

study, research, news reporting, criticism or review. Selected passages,

tables or diagrams may be reproduced for such purposes provided that

acknowledgment of the source is included.

Major extracts or the entire document may not be reproduced by any

process without written permission of the Executive Director, Bureau

of Rural Sciences, PO Box E11, Kingston ACT 2604.

The Bureau of Rural Sciences (BRS) is the science agency within

the Commonwealth Department of Agriculture, Fisheries and

Forestry – Australia.

Postal address:


PO Box E11

Kingston, ACT 2604

Internet: http://www.affa.gov.au/output/ruralscience.html

Preferred way to cite this publication:

Fabiansson, S.U. (2001) Risk assessment in food safety policy

and practice. Bureau of Rural Sciences, Canberra.

This booklet does not represent professional advice given by the

Commonwealth or any other person acting for the Commonwealth for

any particular purpose. It should not be relied on as the basis for any

decision to take action on any matter that it covers. Readers should make

their own enquiries and obtain professional advice, where appropriate,

before making any decision.

The Commonwealth and all persons acting for the Commonwealth

in preparing this booklet disclaim all responsibility and liability

to any person arising directly or indirectly from any person taking

or not taking action based upon the information in this booklet.

iiiii R i s k A s s e s s m e n t i n F o o d S a f e t y P o l i c y a n d P r a c t i c eB u r e a u o f R u r a l S c i e n c e s

F o r e w o r dForeword

Risk Assessment Doc.FA 11/7/01 3:13 PM Page ii

This is an overview of quantitative risk assessment and its use in the food industry

to manage microbiological hazards. The report is directed to policy makers and

industry practitioners and provides a link between objective scientific information

and the development and implementation of improved food safety systems.

Growing acceptance of the principles of risk analysis has led to its use expanding

beyond regulatory standard-setting. Applying risk analysis and risk assessment

to microbial and chemical food safety hazards has recently been advocated

to facilitate international trade, to serve as a basis for sound food safety policy,

and to validate and specify criteria for hazard analysis critical control point (HACCP)

programs. Risk assessment is always the scientific component of the process, but

methodology will vary according to the intended use of the information. A well-

researched quantitative approach is most often a requirement for developing

government food safety policy, but a semi-quantitative or qualitative estimate

could be sufficient for the hazard analysis step in developing an HACCP plan.

A formal quantitative risk assessment (QRA) is demanding of resources and

requires particular expertise. It can take several years to complete all the steps

of a QRA and it often involves international collaboration. It is probably not

feasible for industry to undertake a formal QRA in developing an HACCP program.

However, published QRA information can be used as a basis for assessing

the effect of suggested critical control point interventions in a production system.

A more expedient method to objectively quantify risks is to use expert knowledge.

A number of experts are asked for their opinion in relation to steps in the process,

and a composite picture is then developed by the expert panel to complete the

risk assessment.

At an industry level, a complex questionnaire has been developed with a scoring

system to assess relative risks and to prioritise necessary action.

A similar system is used in a spreadsheet model using a through-chain approach

to calculate a computerised risk ranking score.

Finally, decision tree flow charts can be used to lead the risk assessor through

a simple qualitative approach of assessing risks. This is the most common way

in industry to improve on simple experience and gut feel in analysing hazards

and their potential effect.

Risk assessmentcan be used torank food safetyhazards

vR i s k A s s e s s m e n t i n F o o d S a f e t y P o l i c y a n d P r a c t i c eiv B u r e a u o f R u r a l S c i e n c e s

Different tools forrisk assessment in practice

E x e c u t i v e S u m m a r yExecutive Summary

Risk Assessment Doc.FA 11/7/01 3:13 PM Page iv

Foreword iii

Executive summary v

Introduction 1

Risk assessment—a scientific approach 2

Linkages between risk analysis and HACCP 3

Risk assessment in safe food production 6

Risk assessment and international food standards 6

Risk assessment and national food standards 7

Risk assessment and individual food companies 8

Strengths and limitations of risk assessment 9

Methods for risk assessment 10

Quantitative risk assessment 11

Expert panel analysis 11

Questionnaire ranking tool 11

Computerised risk ranking tool 12

Decision tree flow chart 12

Experience and gut feel 12

International guidelines 12

Reports on quantitative risk assessment 13

International case studies 13

Domestic studies 19

Gap analysis 24

Lack of coordination 24

Lack of knowledge 24

Lack of information sources 24

Gap between science and practice 25

Lack of useful tools 25

Lack of practical industry examples 24

References 26

Appendix 1 – Quantitative risk assessment 31

Appendix 2 – Expert opinion 34

Appendix 3 – Questionnaire ranking tool 36

Appendix 4 – Computerised risk ranking 42

Appendix 5 – Decision tree flow chart 44

viiR i s k A s s e s s m e n t i n F o o d S a f e t y P o l i c y a n d P r a c t i c eB u r e a u o f R u r a l S c i e n c e svi

Several quantitative microbial risk assessments have been published or are well

under way. Assessments of this kind are set up to evaluate one particular hazard

in one type of product, and cannot be generalised.Specific microorganisms covered

include Salmonella spp., Listeria monocytogenes, Bacillus cereus, Clostridium spp.,

Vibrio spp., Shiga toxin producing Escherichia coli and Campylobacter spp. Products

covered so far include (number of studies in brackets):

• egg and egg products (6);

• ready-to-eat food (7);

• seafood (8);

• poultry meat (5);

• beef (4);

• smallgoods (2);

• dairy products (5); and

• plant products (2).

The above studies can be used as a starting point for government and industry

in implementing practical measures to secure a safe food production system. Despite

international efforts to harmonise the application of risk assessment methodology,

many questions remain unanswered.

A gap analysis was undertaken to identify issues and suggest actions to improve

the usefulness of risk assessment in HACCP-based industry quality systems. The

following points were noted:

• Risk assessment is still very much a scientific discipline, and support to industry

case studies would help evaluate the relevance of integrating risk assessment

with HACCP development. Development of practical, industry-focused tools

should be facilitated.

• Several Australian initiatives have begun, but fragmented efforts could benefit

from further coordination.

• Knowledge gained in applying risk assessments to practical problem solving

is still to filter through to industry practitioners. Information access is limited

and could be facilitated through formation of a web-based risk-clearing house.

C o n t e n t sContents

Quantitative riskassessments in

the literature

Risk Assessment Doc.FA 11/7/01 3:13 PM Page vi

Once the importance of food security is realised, a safe food supply is seen as essential.

Governments initially responded by introducing food legislation for visual inspection and end-

product testing. This was not sufficient, and recently quality systems, based on hazard analysis

critical control point (HACCP) principles, have been adopted by industry (Troutt et al. 1995,

Berends and van Knapen 1999, Fabiansson and Cunningham 2000). Ideally, safety can now

be engineered into the production process with a much improved outcome, but a detailed

knowledge of food safety hazards and associated public health risks is critical to success

(Buchanan and Whiting 1998, Morales 1998).

Risk assessment has always been part of human activities. Responses have been mainly

intuitive and countermeasures, based on previous experience, applied on a hit-and-miss basis.

Characterising the potential risk to human health and safety posed by modern food production

systems is a much more complex task. There is thus an urgent need to find an objective risk-

assessment process that can improve decision-making in food production. General principles

have been agreed to, although various aspects of the theory and practice continue to be debated

among scientists, risk professionals, policy makers, industry, and the risk-interested public

(Anon. 2000a).

Figure 1. Some international and national organisations involved in harmonising risk assessment methodology.

Guidelines for risk assessment have been issued through several international and national

organisations (see Figure 1):

• The US National Research Council published a report on the topic in the early 1980s

(Anon. 1983).

• The World Health Organisation, the Food and Agriculture Organisation and their joint body,

the Codex Alimentarius Commission (Codex), have issued principles and guidelines for risk

assessment (Anon. 1999).

Appendix 6 – Further reading 46

Web information 46

Useful literature 47

Glossary 48

Figures

1. Some international and national organisations involved in harmonising 1

risk assessment methodology

2. The triad constituting risk analysis 2

3. Integration of risk assessment into HACCP-system development 3

4. The difference between a harmful agent (hazard) and its likelihood 5

of causing disease and its severity (risk) …

5. Risk assessment continuum 6

6. The selective use of risk assessment tools by different participants

in the food chain 10

7. Decision tree flow chart 45

Tables

1. International quantitative risk assessment case studies sorted according 13

to product category covered

2. Australian risk assessment case studies of different products or product categories 19

1viii R i s k A s s e s s m e n t i n F o o d S a f e t y P o l i c y a n d P r a c t i c eB u r e a u o f R u r a l S c i e n c e s

I n t r o d u c t i o nIntroduction

Risk Assessment Doc.FA 11/7/01 3:13 PM Page viii

L i n k a g e s b e t w e e n r i s k a n a l y s i s a n d H A C C P

The growing acceptance of the principles of risk analysis has led to its use expanding beyond

regulatory standard setting. The application of risk analysis and risk assessment strategies

to microbial and chemical food safety hazards has recently been advocated to facilitate

international trade, to serve as a basis for sound food safety policy, and to validate and specify

criteria for HACCP programs (Kindred 1996, Käferstein et al. 1997). Risk assessment is always

the scientific component of the process but methodology will vary according to the intended

use of the information. A well-researched quantitative approach is most often a requirement for

development of government food safety policy, while a semi-quantitative or qualitative estimate

could be sufficient for the hazard analysis step in developing an HACCP plan (Morales 1998,

van Gerwen et al. 2000).

Risk assessment is a systematic approach for capturing information to describe and estimate

the likelihood and impact of adverse events (Kindred 1996). It is generally considered

to consist of four steps: hazard identification, exposure assessment, hazard characterisation

and risk characterisation (Lammerding and Fazil 2000). The knowledge in each step is combined

to represent a cause-and-effect chain from the prevalence and concentration of the pathogen

or chemical to the probability and magnitude of health impacts (see Figure 3). The process

can be structured in many different ways depending on the nature of the risk-

• The Australia New Zealand Food Authority published a framework for assessing

and managing food safety risks (Anon. 1996).

• A Scientific Steering Committee Working Group of the European Commission

has recommended harmonisation of risk-assessment procedures (Anon. 2000a).

• The Office International des Epizooties has issued risk analysis guidelines pertaining

to animal infectious diseases (Anon. 2000b).

R i s k a s s e s s m e n t — a s c i e n t i f i c a p p r o a c h

It is generally agreed that risk assessment is the process through which information on risks

is identified, organised and analysed in a systematic way to get a clear, consistent presentation

of the data available for practical decision making (Rodricks 1999). Risk assessment was initially

introduced in the early 1970s as a discipline using scientific data to quantitatively evaluate human

cancer risk (Potter 1996). The methodology has been refined and adapted for application in other

fields, and the concept of integrating scientific assessment with practical management

considerations and ongoing communication merged into an overall definition of risk analysis

(Anon. 1999). Codex is now defining risk analysis as a consistent, open and transparent process

following a structured approach consisting of three components: risk assessment, risk

management and risk communication. This report covers the topic of risk assessment, but the

triad composing risk analysis is intertwined (see Figure 2), and the scope of a risk assessment

will depend on its intended use in risk management and risk communication.

The three elements of risk analysis have been given specific meanings:

• Risk assessment is a process of evaluation, including the identification of the attendant

uncertainties, and of the likelihood and severity of an adverse effect occurring to humans

following exposure under defined conditions to an apparent hazard.

• Risk management is the process of weighing policy alternatives in the light of the result

of a risk assessment and of other relevant evaluations, and, if required, of selecting and

implementing appropriate control options.

• Risk communication is the interactive exchange of information and opinions throughout

the risk-analysis process concerning risk. It should involve not only risk assessors and risk

managers, but also consumers and a wide range of other actual or potential stakeholders.

As an integrated part of risk analysis, the scientific reliability and

relevance of the risk assessment is of crucial importance for the

validity of the whole risk-analysis process. Risk assessment, properly

defined and applied, is the most appropriate tool currently

available for quantifying the risks associated with microbial

and chemical contamination of foods (Morales 1998).

32 R i s k A s s e s s m e n t i n F o o d S a f e t y P o l i c y a n d P r a c t i c eB u r e a u o f R u r a l S c i e n c e s

RiskAssessment

RiskManagement

RiskCommunication

HACCP

RiskCharacterisation

HazardIdentification

ExposureAssessment

HazardCharacterisation

Risk Assessment

Hazard Analysis

Critical ControlPoints

CriticalLimits

Monitoring System

CorrectiveAction

Verification

Documentation

Figure 3. Integration of risk assessment into HACCP-system developmentFigure 2. The triad constituting risk analysis.

Risk Assessment Doc.FA 11/7/01 3:13 PM Page 2

HACCP planning is already an information-intensive process. Moving from a qualitative hazard

analysis to a quantitative risk assessment increases both the extent of the data needed and the

sophistication required of the team performing the evaluation. Such an undertaking is well

beyond the capabilities of many individual food companies. However, the development

of user-friendly software that takes the manufacturer through the process of developing HACCP

plans, evaluating microbiological risks, and establishing performance standards is already being

envisioned to help the small producer take advantage of these conceptual tools. Dynamic risk-

assessment models can be developed by industry organisations covering the entire farm-to-table

continuum, considering each step that influences the overall risk. Such models can sequentially

link a series of probabilistic process steps and control measures in the form of critical limits for

critical control points. The ability to include the variability associated with each of these steps,

to perform sensitivity analyses, and to conduct failure scenarios potentially provides a more

analytical approach to identifying critical control points. The effect both of critical points and

critical control points can be integrated to consider the overall effectiveness of the entire process.

A dynamic model would allow individual companies to amend the parameters to reflect their

individual production system (Buchanan and Whiting 1998).

The difficulty involved in integrating quantitative risk-assessment methodology into the HACCP

process should not be underestimated. Lack of readily available data can severely hamper any

attempt at constructing a through-chain model. Microbiological ecology of foods is still very much

an evolving field. Human dose-response data is often lacking, and variability in sensitivity levels

can be large between sub-populations. Similar problems faced toxicologists assessing chemical

risks; they were overcome by creating a number of default assumptions and safety margin

allowances. Microbial risk assessment has not yet arrived at that point, and thus acceptable risks

still need to be defined. This is particularly true for industry applications, and there is an urgent

need to develop industry case studies to test theories in realistic settings in order to recommend

practical solutions.

5R i s k A s s e s s m e n t i n F o o d S a f e t y P o l i c y a n d P r a c t i c e

The HACCP system identifies specific hazards, estimates risks and establishes specific control

measures to ensure the safety of food. It is a tool that emphasises prevention and control rather

than reliance on end-product testing and traditional inspection methods. The seven principles

of HACCP are: hazard analysis, critical control point identification, critical limit specification,

monitoring system, corrective actions, verification and documentation. Implicit in developing

an HACCP plan is the use of some form of risk assessment. This is normally at a cursory

level when identifying process and product hazards and setting control measures to avoid

unacceptable contamination. The HACCP methodology has greatly assisted the agrifood industry

in systematically designing programs to ensure the microbiological safety of foods. Unfortunately,

the scope of HACCP programs has been limited by reliance on qualitative measures of hazards

and their control. Qualitative measures cannot easily quantify the combined effect of multiple

control-point deviations or relate the successful operation of an HACCP system to measurable

public health impact. Recent advances in quantitative microbial risk assessment seem to provide

a means for overcoming those obstacles (Buchanan and Whiting 1998).

The integration of HACCP plans with dynamic risk-assessment models offers a means

of considering the entire farm-to-table continuum and for relating food-manufacturing

operations to public health goals. Such capabilities may be critical in establishing equivalence

among HACCP systems. Inclusion of objective risk assessment in the HACCP process can help

decide unacceptable levels of contaminants and provide a scientific rigour to the process to better

withstand third-party auditing. In particular, dynamic microbial risk-assessment models that

incorporate simulation techniques provide a means for dealing with the variabilities associated

with food raw materials. For producers of products for which there is no definitive lethal

processing step, like most production of raw agricultural commodities, or for those that have

formulations that are based on multiple barriers, integrating quantitative risk-assessment

techniques into HACCP plan development may be critical in demonstrating product safety

(Notermans and Mead 1996, Morales 1998, Buchanan and Whiting 1998, McNab 1998).

The use of HACCP and risk assessment within the food industry is still something of a novelty,

and many concepts and techniques need further explanation. While the thrust of HACCP

is simple, a number of underlying concepts are only poorly defined and often not widely

considered. This has lead to confusion concerning some of the very basic principles of HACCP.

The distinction between risk and a hazard is one of them. Commonly these words are used

interchangeably, but from a scientific standpoint they have different meanings (Ahl et al.

1993, Kaplan 1997). Risk assessment involves the identification of hazards and the methodical

description of a system and all its possible failures that could give rise to public exposure to the

hazard. A hazard is defined as any adverse event. The concept of risk, however, includes the

chance that the hazard will occur, and some measure of the consequences if it does (see Figure

4). Thus, the end-result of a risk assessment is an overall picture of the likelihood of the hazard

being realised and the severity of the impact if it is. The term “safety” is often used to describe

situations in which there is no appreciable risk

4 B u r e a u o f R u r a l S c i e n c e s

likelihood of harm and severity of impact

Risk

an agent with the potential to cause harm

Hazard

Figure 4. The difference between a harmful agent (hazard) and its likelihood of causing disease and its severity (risk).


The increased volume of the global food trade underscores the need for sound epidemiologic

information and international risk assessment. In this regard, Article 5 of the Sanitary and

Phytosanitary Measures Agreement explicitly requires WTO members to conduct scientific

and consistent risk assessments. Furthermore, the World Health Organisation (WHO)

has recommended the application of the HACCP system at every stage of the food chain

as an effective approach for governments to meet the terms outlined in the agreement.

The issue is receiving more attention from regulatory agencies, initially underlined by the Food

and Agriculture Organisation (FAO)/WHO Conference in 1991 on Food Standards, Chemicals

in Food, and Food Trade, requesting a scientific basis for the Codex standards. The conference

recommended that the Codex, in its norm-setting work on health and safety, place greater

emphasis on risk assessment. As a result, the Codex Alimentarius Commission identified risk

assessment of microbiological hazards in foods as a priority area of work. At its 32nd session

the Codex Committee on Food Hygiene (CCFH) identified a list of pathogen-commodity

combinations on which expert risk assessment advice is required (http://www.who.int/fsf/

mbriskassess/32thccfh.htm). In response, FAO and WHO are jointly launching a program

of work with the objective of providing expert advice on risk assessment of microbiological

hazards in foods to their member countries and to the Codex Alimentarius Commission.

As a consequence of the emerging international regulatory environment, microbial risk assessment

is now a common request in trade negotiations. Given that sanitary and phytosanitary measures

must be based on scientific evidence (or, where appropriate, scientific risk assessment), a formal

approach to the use of microbial risk assessment is required for the quantitative evaluation

of the microbial safety of foods

R i s k a s s e s s m e n t a n d n a t i o n a l f o o d s t a n d a r d s

Risk assessment and risk management underpin the government decision-making process

used for the protection of public health and safety in relation to the consumption of food.

It is important to understand these concepts when addressing the broader question about

acceptable levels of risk. In Australia, food standards allow the sale of a wide range of food

commodities, some of which may carry an inherent, albeit low, level of risk and, in this regard,

may be at odds with community expectations. In many cases, however, low levels of risk must

be balanced against the health benefit of a nutritious and varied diet. An informed debate about

acceptable levels of risk in relation to food can be achieved only through a greater awareness

of risk assessment and risk-management practices.

National food safety regulation is increasing in the degree of management complexity in many

countries. This increase is related to increased production volumes and product sophistication

and to changing consumer handling and consumption patterns. In addition, consumer

expectation has become progressively more demanding. Today, the public sometimes has virtually

unattainable expectations regarding the degree of safety that can be achieved through regulatory

action. Indeed, some consumers seem to desire 100% safety for many food products. Unfortunate-


Risk management is the field of decision-making, or at least its preparation. Its first objective

must be to safeguard health. Risk management appeals to the fields of technology, economy,

law, diplomacy and politics, particularly the regulation of public affairs (Anon. 2000a). Similarly,

in supporting risk-management decision-making, risk assessment will provide scientific support

to many disciplines at different levels of the food chain. Risk assessment can be applied at an

international level, at a national level, and at an industry or company level (see Figure 5).

R i s k a s s e s s m e n t a n d i n t e r n a t i o n a l f o o d s t a n d a r d s

As a result of food trade liberalisation, concern about the safety of imported food has grown.

The movement of ever-increasing quantities of food across borders has resulted in transnational

disease risks. Thus, the globalisation of the food trade and the open access to foreign markets

need to be accompanied by effective means of health protection for large international

populations. In the food sector, international regulatory instruments need to be integrated

with strengthened surveillance and monitoring (Käferstein et al. 1997).

Application of the Sanitary and Phytosanitary Measures Agreement, prescribed by the World

Trade Organisation (WTO), is designed to address these concerns (Vose 1998). Codex

Alimentarius Commission standards, guidelines, and recommendations are recognised as the

reference for national food safety requirements. Countries that are members of the WTO may

no longer be able to reject foods that meet Codex recommendations without providing

justification (Anon. 2000, Klapwijk et al. 2000).


R isk assessment in safe food produc tionRisk assessment in safe food production

Figure 5. Risk assessment continuum


the three steps are considered separately, when developing microbial risk assessments

for food processes and HACCP design there are more typically a series of risk assessment/risk

management iterations. A process is assessed, the results interpreted, the process modified, and

the new risk assessed until the process provides an acceptable level of control.

So far the primary focus of risk assessment has been to consider risks associated with entire

industries and broad groups of products; in fact, some proponents feel that microbial risk

assessment should be limited to these broad evaluations. However, when dealing with an industry

whose processes and products are as diverse as those encountered with foods, any meaningful

translation of risk-assessment models into risk-management decisions will require individual

assessments of specific products and processes. Just as HACCP is only truly effective when

applied on a plant-specific basis, ultimately risk-assessment models need to be flexible enough

to allow the user to modify parameters to reflect the specific attributes of their products, processes

and facilities. Performing risk simulations for a standard or proposed food process gives an initial

assessment of the magnitude of the hazard and more directly relates the process to public health

goals. The use of simulation techniques represents an important paradigm shift in that hazard

assessments begin to address variability inherent in any process, product and pathogen.

Simulations can be used to determine if the focus of an HACCP plan needs to be on

improvement of the entire process or a single operation, or whether preventing a few extreme

combinations of factors from occurring will make a process safer. With information from risk

assessments, it should be much easier to establish quantitative values for critical limits to produce

the degree of stringency required, and then to design appropriate monitoring systems. Results

can also suggest how much abuse the process can withstand both before and after the product

leaves the manufacturer’s control. This can assist manufacturers in practising “due diligence”

in the design of their process and HACCP program.

S t r e n g t h s a n d l i m i t a t i o n s o f r i s k a s s e s s m e n t

It is clear that risk assessment is a valuable tool in helping us make public health decisions

in the food safety area. Risk assessment helps to organise information about risk in a useful

manner. It also characterises the nature and likelihood of harm to the public. It helps to define

the uncertainties and provides some level of comfort with the inferences that are made. And

it points out data gaps that can help to prioritise research needs.

But there are some weaknesses to the risk assessment process as well. Risk assessment has

its roots in toxicology and carcinogenicity studies, and its application to other disciplines poses

significant challenges. For food safety, one challenge relates to the fact that bacteria can multiply

and produce toxins as conditions change while food moves through the farm-to-table continuum,

unlike chemical, environmental, or toxicological contaminants.

In addition to the difficulties in applying risk assessment to pathogens, risk assessment is also

subject to two types of uncertainties; those related to data and natural variability, and those

associated with any assumptions that are required when directly applicable data are not available.


ly, given current technology and production capabilities, industry cannot guarantee that a given

unit of a food product will be totally free of hazards. Some residual risk always remains, even

though it may be quite small in magnitude. Recognition of the need for a more systematic

process has prompted the adaptation of risk-analysis techniques to the field of food regulation.

Food regulators have always informally assessed the human health risks that may be associated

with the handling and consumption of food products. The informal qualitative assessment

of risk has assisted managers in decision-making for many years. This assessment is effective

and will continue to be used by managers to enhance food safety. However, for major decisions

it is important to gather all of the pertinent information and to consider it in the decision-making

process. This includes both available data and the uncertainties associated with the information.

Risk assessment is the appropriate tool to objectively evaluate this information. This deliberate

and comprehensive approach is valuable for major concerns in order to consider all available

information that could influence a management decision. The risk-assessment procedures also

document which information was considered and the assessment process itself.

R i s k a s s e s s m e n t a n d i n d i v i d u a l f o o d c o m p a n i e s

The second step in hazard analysis is to conduct a risk assessment. The type of risk assessment

that is used in hazard analysis for HACCP is different from the typical health-risk assessments.

The risk that is of concern to the food industry is not the probability of exposure of the consumer

to the hazard as it is in epidemiological risk assessments. Industry is more concerned with the

probability of microbial growth, survival, or contamination in the product during the process,

and how this can influence the microbiological status of their product. Although those conducting

an HACCP study must keep in mind the associated risks at the farm and consumer levels, the

food-processing industry must primarily focus on those hazards that they can control. Thus,

in the production of food, efforts are concentrated on controlling the occurrence of a hazard

because, in many cases, the consequences are beyond the company’s control.

Food microbiologists are called upon daily to make decisions concerning both the quality and

safety of foods, optimising the dual needs to provide wholesome products while maintaining

high quality at the least possible cost. Various conceptual tools have been developed to help

microbiologists make decisions of this nature. The most widely recognised and used system

is HACCP, which focuses on identifying and controlling the key process steps most significantly

affecting the safe production of a food. However, as HACCP has become more widely adopted,

it has become evident that there are areas within this approach that could be strengthened

if microbiologists were better able to quantitatively link product attributes with public health

concerns. In particular, conducting the initial hazard analysis and establishing critical limits would

benefit from more quantitative means of assessing the risks associated with the hazard identified

for a food product.

The desire to better estimate the potential impact of issues relating to safety and quality has

stimulated interest in the development of quantitative risk-assessment techniques that would

allow the food industry to perform reliable risk analyses. Although in traditional risk analysis



Q u a n t i t a t i v e r i s k a s s e s s m e n t

Quantitative food safety risk-assessment was initially applied for chemical hazards but is now

covering microbial hazards as well. There are some important differences in the nature of

toxicants and microbes that strongly affect the way in which risk assessments are carried out.

The principles of risk assessment are simple, consisting of four closely related steps permitting

a quantitative estimation of the risks associated with the consumption of a particular food

product. However, there are uncertainties, particularly in identifying hazards and in determining

dose-response relationships. Some causative agents of foodborne disease have yet to emerge

and be recognised. Dose-response relationships are weak points because reliable data are scarce.

This is especially true for certain sensitive groups.

Research is needed to determine the effect of infective dose in relation to the type of food and

the distribution of pathogens in a food product. Research is also needed to better understand

the critical assumptions involved in following the distribution and numbers of particular

pathogens throughout the food production and processing chain. The effect of individual

processing steps is often ill defined, and their inclusion into an HACCP plan can often lead

to arguments between processors and auditors. Increasingly, the safety of food will be based

on controlling all steps in the food production process. Such controls must be based on priorities

set after a proper risk assessment in order to be most effective.

Data assembled during the risk-assessment process is often modelled, and uncertainty and

variability scenarios tested, through mathematical simulations. A common method is Monte Carlo

analysis. The model can include a sensitivity analysis whereby the influence of assumptions can

be tested and the effect of corrective actions assessed. Critical knowledge gaps can be identified

and used to direct future research efforts. Quantitative risk assessment is described in Appendix 1.

E x p e r t p a n e l a n a l y s i s

Often there is a lack of available information to undertake an objective quantitative risk

assessment. The immediate alternative is to perform a qualitative assessment, the drawback being

that numbers will not be available to compare effects of remedial activities. As an interim solution,

several experts in the area can be asked to give their opinion on the different steps of the

process. A questionnaire could be used to ask the experts to rank effects. By averaging responses,

a semi-objective analysis will be achieved. This is sometimes used by scientists but is more

difficult for industry to undertake. Expert panel analysis is described in Appendix 2.

Q u e s t i o n n a i r e r a n k i n g t o o l

This method has been pioneered by Campden & Chorleywood Food Research Association

Group (CCFRA) in an industry guideline published in June 2000 (Voysey 2000). CCFRA

developed a “Risk Profile” questionnaire to cover all the steps involved in risk assessment.

The questionnaire includes 44 questions sorted in 14 sections, most including a five-point scoring

system. The questionnaire addresses all four areas of an objective quantitative risk assessment.

The questionnaire has been used as a starting point for the description in Appendix 3.


The ultimate risk assessment methodology is a scientific quantitative risk assessment (QRA).

A QRA requires much in the way of resources and particular expertise. It can take several years

to complete all the steps of a QRA, and it often involves international collaboration. It is probably

not feasible for industry to undertake a formal QRA in developing an HACCP program. However,

published QRA information can be used as a basis for assessing the effect of suggested critical

control point interventions in a production system.

A more expedient method to still objectively quantify risks is to use expert knowledge. A number

of experts are asked for their opinion in relation to steps in the process, and a composite picture

developed by the expert panel to complete the risk assessment.

At an industry level, a complex questionnaire has been developed with a scoring system

to assess relative risks and to prioritise necessary action.

A similar system is used in a spreadsheet model using a through-chain approach to calculate

a computerised risk-ranking score.

Finally, decision tree flow charts can be used to lead the risk assessor through a simple

qualitative approach of assessing risks. This is the most common way in industry to improve

on simple experience and gut feel in analysing hazards and their potential effect.

Figure 6. The selective use of risk-assessment tools by different participants in the food chain.


Quantitative Risk AssessmentExpert panel

Questionnaire Ranking

Government/Industry

Business

Decision TreeExperience/Gut feelProprietary models

Results/outcomes of industryrisk assessments should

provide knowledgefor food business

M e t h o d s f o r r i s k a s s e s s m e n tMethods for risk assessment



I n t e r n a t i o n a l c a s e s t u d i e s

International case studies listed here follow the quantitative risk assessment

methodology and cover a range of product and microorganism combinations.

The studies have been summarised in Table 1 according to product area covered.

Each of the areas covered and the studies completed are presented briefly in the

following text.

Table 1. International quantitative risk-assessment case studies sorted according to product category covered.

Product Microorganism Country Reference

Egg and egg Salmonella spp. Canada Todd 1996

products Salmonella enteritidis USA Whiting and Buchanan 1997

Salmonella enteritidis USA Baker et al. 1998

Salmonella enteritidis International Ebel et al. 2000

Salmonella spp. International Fazil et al. 2000

Salmonella spp International Fazil et al. 2001

Ready-to-eat Listeria monocytogenes Canada Farber et al. 1996

food Listeria monocytogenes International Buchanan and Lindqvist 2000

Listeria monocytogenes International Ross et al. 2000

Listeria monocytogenes International Buchanan et al. 2001

Listeria monocytogenes USA Whiting and Long 2001

Bacillus cereus France Carlin et al. 2000

Clostridium botulinum France Carlin et al. 2000

Vibrio parahaemolyticusListeria monocytogenesVibrio spp.

Vibrio spp

USA

Sweden

International

International

Miliotis and Watkins 2000

Lindqvist and Westö 2000

Walderhaug and Bowers 2001

Depaola et al. 2001

Seafood

Salmonella spp.

Salmonella spp.

Salmonella spp.

Salmonella enteritidisCampylobacter spp.

UK

International

International

International

International

Brown et al. 1998

Fazil et al. 2000

Fazil et al. 2001

Kelly et al. 2000

Anderson et al. 2001

Poultry meat

Escherichia coli O157:H7


EU

Canada

Dormont et al. 2000

Cassin et al. 1998Beef

Listeria monocytogenesListeria monocytogenesBacillus cereus

France

Canada

Netherlands

Bemrah et al. 1998

Farber et al. 1996

Notermans et al. 1997

Dairy

C o m p u t e r i s e d r i s k r a n k i n g t o o l

The risk assessment analysis seems ideally suited to using computerised tools; however, this

review only encountered one of those. This is a spreadsheet system developed at the University

of Tasmania (Ross and Sumner 2001). The spreadsheet asks 11 questions and, using a mix of

logarithmic and linear modelling, calculates a risk estimate for a given population to become ill

from a particular organism in a defined product. The spreadsheet is illustrated in Appendix 4.

D e c i s i o n t r e e f l o w c h a r t

This is the most common improvement to pure risk guessing. It takes the practitioner through

a few questions in a formalised decision tree structure. The benefit is that, rather than guess the

general risk involved in a process or for a particular hazard, it can be split up into components

and considered more carefully. An example is given in Appendix 5.

E x p e r i e n c e a n d g u t f e e l

In developing an HACCP plan, hazard analysis involves some sort of risk assessment

in order to prioritise development of critical control points and sample plans. In many cases

this is building on joint experience acquired by company staff and could be developed

in a round-table discussion format. This can work well in many companies; the big problem

is the lack of objectivity for the process to stand up to audit scrutiny. A more formal and

quantitative approach will facilitate the audit process.

I n t e r n a t i o n a l g u i d e l i n e s

• Codex Principles and Guidelines for the Conduct of Microbiological Risk Assessment

(CAC/GL-30 1999). ftp://ftp.fao.org/codex/standard/volume1b/en/GL_022e.pdf

• European Commission. Principles for the development of risk assessment of microbiological

hazards under directive 93/43/EEC concerning the hygiene of foodstuffs. Office for Official

Publications of the European Communities, Luxembourg (1997).

http://europa.eu.int/comm/food/fs/sc/scf/index_en.html

• Joint FAO/WHO Expert Consultation on Risk Assessment of Microbiological Hazards in Food

(WHO Headquarters, 15-19 March,1999).

http://www.who.int/fsf/mbriskassess/Consultation99/index.htm

• International Life Science Institute (ILSI) Revised framework for microbial risk assessment.

An ILSI Risk Science Institute Workshop Report, ILSI Press, Washington D.C. (2000).

http://www.ilsi.org/file/mrabook.pdf


R e p o r t s o n Q u a n t i t a t i v e R i s k A s s e s s m e n tReports on Quantitative Risk Assessment



E g g s a n d e g g p r o d u c t s

Quantitative risk assessment model for Salmonella enteritidis

in pasteurised liquid eggs

One of the first complete microbial risk assessments to be published covered

the risk of Salmonella enteritidis being spread through home-made mayonnaise

made from pasteurised liquid eggs (Whiting and Buchanan 1997). Using

a ‘unit operations’ and stochastic simulation approach, data on the frequency

of pathogens in raw ingredients, predictive microbiology models for growth

and inactivation, and dose-response models for infectivity were integrated

to create a quantitative risk-assessment model. The risk assessment indicated

that pasteurisation provided sufficient consumer protection from a high incidence

of infected birds and from temperature abuse between the farm and the

egg breakers. However, scenarios showed how inadequate pasteurisation

temperatures and/or temperature abuse during storage lead to a hazardous

product. This dynamic risk model aids in the identification and setting

of critical control points and of assessing the impact of alterations to food

formulations or processes.

Salmonella enteritidis risk assessment—shell eggs

and egg products

The previous project was expanded in a joint study involving scientists from the

US Department of Agriculture, the Food and Drug Administration, the Centre

for Disease Control, and universities (Baker et al. 1998). The joint study covered

a farm-to-table microbial risk assessment of Salmonella enteritidis in shell

eggs and egg products. It summarises the risk-assessment process from the

development of a conceptual framework through the careful organisation

of information obtained from published scientific literature and unpublished

academic, government and industry sources, to the incorporation of available

data into a comprehensive quantitative model. The model illustrates the public

health effects associated with the consumption of Salmonella enteritidis-infected

shell eggs and egg products. According to the baseline egg products model the

probability is low that any cases of Salmonella enteritidis will result from the

consumption of pasteurised egg products. However, it was concluded that the

current US Food Safety and Inspection Service time and temperature regulations

do not provide sufficient guidance to the egg products industry for the large

range of products involved. Time and temperature standards based on the

amount of bacteria in the raw product, how the raw product will be processed,

and the intended use of the final product will provide greater protection to the

consumers of egg products.

Risk assessment of use of cracked eggs in Canada

In 1992, Agriculture and Agri-Food Canada amended its Egg Regulations

to restrict movement of Canada C eggs (cracked) to federally registered processed


egg stations for pasteurisation. This was questioned by egg producers and some

provinces on economic grounds. It was also in conflict with long-standing

practices of marketing eggs in some provinces to retail stores, bakeries,

restaurants and institutions or at the farm gate. In order to determine how

much of a risk these eggs were to human health, the Health Protection Branch

of Health Canada conducted a risk assessment (Todd 1996). On the basis of

outbreak data, Salmonella spp. were identified as the main hazard in these eggs.

Salmonellae may occasionally be present on shell eggs even after washing, and

any microorganism reaching the membranes can be transferred to an egg mixture

through breaking, and will rapidly grow under incorrect storage conditions.

A relative risk analysis showed that cracked eggs are 3 to 93 times more likely

than uncracked shell eggs to cause outbreaks. A probability of illness of 1

in 3 800 was derived from the 40 million cracked eggs produced in Canada and

not pasteurised and a probable 10 500 illnesses arising from these. Even though

it is not possible to precisely determine the risk of salmonellosis through cracked

eggs, this assessment indicated that there was enough of a concern to merit

a management strategy. Eight options for managing the risk were considered

and ranked for acceptability.

Salmonella in broilers and eggs

Codex is currently attempting to coordinate international efforts in the food risk

assessment area. Initial work on hazard identification, hazard characterisation

(Fazil et al. 2000) and exposure assessment (Ebel et al. 2000) of Salmonella spp.

and Salmonella enteritidis in eggs, respectively, has been completed. The two

projects were combined with a risk assessment of Salmonella spp. in broilers for

the final risk characterisation (Fazil et al. 2001). The risk characterisation initially

set out to understand how the incidence of human salmonellosis is influenced

by various factors during the agricultural phase of egg production, marketing,

processing, distribution, retail storage, consumer storage, meal preparation and

finally consumption. Such models are appealing because they enable the study

of the broadest range of intervention strategies. However, as the work progressed

it became evident that quantitative information available was not sufficient to

allow the construction of a full model. The model now used will be expanded

as further information becomes available.

R e a d y - t o - e a t f o o d

Health risk assessment of Listeria monocytogenes in Canada

Certain ready-to-eat foods are vulnerable to contamination and growth

of Listeria monocytogenes. A Canadian study presents the major steps used

in the formulation of a health-risk assessment for Listeria monocytogenes (Farber

et al. 1996). Data is given on the numbers of human listeriosis cases reported in

Canada along with the current Canadian regulatory policy on L. monocytogenes.

Since no direct human dose response data is available for L. monocytogenes,


S e a f o o d

Listeria monocytogenes in salmon and rainbow trout

The objective of a Swedish study was to develop a quantitative risk-assessment

model in which the exposure and risk of acquiring listeriosis from consumption

of packaged smoked or gravad salmon and rainbow trout were estimated

(Lindqvist and Westö 2000). An Excel spreadsheet model was constructed

in which variables were represented by distributions based on surveys

of Listeria monocytogenes in these food products, and on demographic

and consumption data.

Public health impact of Vibrio parahaemolyticus

in raw molluscan shellfish

A draft risk assessment was produced by the U.S. Food and Drug Administration

to characterise the public health impact associated with consumption of raw

oysters harbouring pathogenic Vibrio parahaemolyticus (Miliotis and Watkins

2000). The risk assessment structures knowledge of the microorganism

in a systematic manner and includes sophisticated mathematical models to

estimate consumer exposure and dose-response relationships with the organism.

Vibrio spp. in seafood

Codex has also initiated work on Vibrio spp. in seafood. Preliminary findings

only on hazard characterisation (Walderhaug and Bowers 2001) and exposure

assessment (Depaola et al. 2001) have been presented for discussion.

P o u l t r y m e a t

Comparative risk of salmonellosis from chicken products

In the United Kingdom, models have been developed according to the Codex

principles to provide a quantitative risk assessment of salmonellosis involving

frozen poultry products (Brown et al. 1998). This model-based QRA takes into

account three types of information: occurrence and distribution of the agent,

sensitivity of populations to infection (e.g. normal or susceptible), and the effect

of cooking (in the factory or home) on concentration of the agent and hence

risks of infection after product consumption.

Salmonella in broilers and eggs

Codex has initiated international work on Salmonella in poultry meat. Initial

work on hazard identification, hazard characterisation (Fazil et al. 2000) and

exposure assessment (Kelly et al. 2000) of Salmonella spp. in broiler products

has been completed. The projects were combined with a risk assessment of

Salmonella in eggs for the final risk characterisation (Fazil et al. 2001). As the

work progressed it became evident that quantitative information available was

not sufficient to allow the construction of a full model. The model now used

will be expanded as further information becomes available.

17R i s k A s s e s s m e n t i n F o o d S a f e t y P o l i c y a n d P r a c t i c e16 B u r e a u o f R u r a l S c i e n c e s

a flexible dose-response model called the Weibull-Gamma model was used for

the hazard characterisation step. Paté was used as a prototype for the exposure

assessment step. The dose response output was combined with some predictive

growth modelling assuming an initial exposure of a single cell for food stored

at 4 and 8˚C. Storage of paté at 4˚C for more than 35 days resulted in a rapidly

increasing risk for the high-risk population. Storage at 8˚C produced a similar risk

after about 13 days.

Listeria monocytogenes in ready-to-eat foods

Codex has initiated an international assessment of the risk of L. monocytogenes

in ready-to-eat foods with initial work on hazard identification, hazard

characterisation (Buchanan and Lindqvist 2000) and exposure assessment

(Ross et al. 2000) completed. The risk assessment was concluded with the risk

characterisation component (Buchanan et al. 2001). Considering the resources

available the risk assessment was limited to selected ready-to-eat foods

representing various classes of product characteristics. The risk of these foods

serving as a vehicle for human food-borne listeriosis was estimated.

Relative public health risk from Listeria monocytogenes

in ready-to-eat foods

The U.S. Food and Drug Administration, the U.S. Department of Agriculture

and the Centre for Disease Control conducted an assessment of the relative risk

of listeriosis associated with various types of ready-to-eat foods (Whiting and

Long 2001). This draft risk assessment details the current state of knowledge

about this foodborne disease, and includes sophisticated models developed

to estimate exposure and dose-response relationships, and the variability

inherent in these estimates. A distinction is made between L. monocytogenes

infections limited to mild, flu-like symptoms (referred to as listerial gastroenteritis)

and those that are severe and life-threatening (referred to as listeriosis). This risk

assessment only addresses listeriosis.

Risk assessment of spore-forming pathogenic bacteria

in cooked chilled vegetable foods

A Flair-Flow Europe project aim to estimate the risk of food-borne disease from

hazardous spore-forming bacteria in cooked-chilled foods containing vegetables

(Carlin et al. 1999). Data from a literature review undertaken on hazard

identification and characterisation were used to identify dose-response

information curves for some of the hazardous spore-forming bacteria and their

toxins. Quality of the data was relatively poor and sparse. Spore-forming bacteria

detected in a range of processed products and raw vegetables belonged mainly

to the Bacillus genus. Clostridium spp. were also present but in very low

numbers. A non-isothermal measuring method indicated that the heat resistance

of spore-forming bacteria can vary greatly and validation of heat resistance in real

foods is, therefore, necessary.


General Listeria monocytogenes

Meat Inspection

Healy 2001

Pointon et al. 2000

Grains General Richardson 2000

Human listeriosis from consumption of soft cheese made

from raw milk

A French study identified microbial hazards in soft cheese made from raw milk

(Bemrah et al. 1998). Quantification of the resulting public health risk was

attempted within the framework of the Codex Alimentarius Commission 1995

approach to quantitative risk assessment, using Monte Carlo simulation software.

Quantitative data could only be found for Listeria monocytogenes. The complete

process of cheese making was modelled, from milking to consumption.

A risk assessment study of Bacillus cereus present in pasteurised milk

A Dutch study looked at the heat resistance of Bacillus cereus, its potential

pathogenic character, the capability to grow in milk and reported diseases upon

consumption of dairy products (Notermans et al. 1997). It concluded that the

organism should be considered as hazardous in pasteurised milk.

D o m e s t i c s t u d i e s

Australian case studies listed here use a combination of methods for the

risk assessment. Some include a full quantitative risk assessment of specific

product/microorganism combinations while others use a quantitative approach

to provide an initial risk-ranking of a range of products in an industry.

The studies have been summarised in Table 2 according to product covered.

Table 2. Australian risk-assessment case studies of different products or product categories.

Plant products General Richardson 2000


Shiga toxin Escherichia coli

Lammerding et al. 1999a

Lammerding et al. 1999dBeef


Campylobacter spp. in broiler chicken

Codex has also initiated work on Campylobacter spp. in broiler chickens.

Preliminary findings only on hazard identification and characterisation, and

exposure assessment (Anderson et al. 2001), have been presented for discussion.

B e e f

Oral exposure of humans to the BSE agent: Infective dose

and species barrier

Many risk assessments have been performed in relation to the pathogenicity

of the bovine spongiform encephalopathy (BSE) agent. The European

Commission Scientific Steering Committee has produced an opinion on the

infective dose and species barrier of oral exposure of humans to the BSE agent

(Dormont et al. 2000). This document, although not following the risk-assessment

model, includes an analysis of the risk to humans of the entry of BSE-infected

tissues into the food chain.

Escherichia coli O157:H7 in ground beef hamburgers

A Canadian study modelled the human health risk associated with Escherichia

coli O157:H7 in ground beef hamburgers (Cassin et al. 1998). Two models were

developed; the first was intended to describe the behaviour of the pathogen

from the production of the food through processing, handling, and consumption

to predict human exposure. The exposure estimate was then used as input to

a dose-response model to estimate the health risk associated with consuming

food from the process. The efficacy of three risk-mitigation strategies was

evaluated by modifying the values of the predictive factors and comparing

the new predicted risk. The average probability of illness was predicted to be

reduced by 80% under a hypothetical mitigation strategy directed at reducing

microbial growth during retail storage through a reduction in storage temperature.

D a i r y p r o d u c t s

Health risk assessment of Listeria monocytogenes in Canada

Soft cheese as a high-risk food in terms of listeriosis infection was used

as prototype in risk-assessment models (Farber et al. 1996). Using disappearance

data for cheese and 100g as a typical serving, the data suggested an average

of 102 servings per capita, per year in Canada. An equation to calculate the

average probability of acquiring human listeriosis in Canada from soft and

semi-soft cheese consumption was formulated. Computations derived from

this equation indicated substantial consistency between reported data and

assumptions of the risk-assessment model. The total annual estimated cost

of listeriosis illnesses and deaths in Canada was estimated to be between

C$11.1 and 12.6 million.


Product Microorganism Reference

Clostridium perfringens Lammerding et al. 1999cReady-to-eat

General

General

Listeria monocytogenes

Walsh 1999; Ross and Sanderson 2000

Sumner 2001

Healy 2001

Miliotis and Watkins 2000

Seafood

Dairy General

General

Ross 2001a

Safefood Production NSW 2000

Smallgoods Escherichia coli O157

Listeria monocytogenes

Lammerding et al. 1999b

Sumner 2001


Listeria monocytogenes in selected foods

A partnership between Australia New Zealand Food Authority, the New Zealand

Ministry of Health, Food Science Australia and the New Zealand Institute

of Environmental Science and Research Limited (ESR) to explore risks involving

Listeria monocytogenes in selected foods has been initiated (Healy 2001).

The study is currently concentrating efforts in the seafood industry since

zero tolerance legislation for the organism in cooked seafood products is being

introduced. The aim is to provide regulators, industry and consumers with

a comprehensive and transparent strategy, which identifies the general principles

to minimise risk to Australian and New Zealand consumers of a food-borne

Listeria infection. The strategy development will take into account a number

of national and international initiatives in the management of Listeria.

The project has five modules:

– Assessment of current Listeria management strategies

– Surveillance and epidemiology

– Resource identification

– Risk assessment and management framework development

– Strategy document

B e e f

Escherichia coli O157:H7 in fresh ground Australian beef:

Issues of equivalence

Escherichia coli O157:H7 remains an important pathogen to many of Australia’s

trading partners. This is particularly true of the United States, which is our largest

customer for frozen boxed beef destined for the hamburger market. Because

of the increased awareness of this particular pathogen a model was developed

to consider the concept of equivalence between food production systems

(Lammerding et al. 1999a). The study, incorporating the farm-to-fork continuum,

estimated the risk to the North American population of E. coli O157:H7

through the consumption of hamburgers manufactured from Australian meat.

This case study illustrates the potential application of risk assessment models

in determining equivalence as has been proposed under the Sanitary and

Phytosanitary (SPS) Agreement of the World Trade Organisation. The case study

highlights the potential application of risk assessment in trade disputes based

on food safety.

Shiga toxin-producing E. coli (STEC) in ground Australian beef:

Process improvement

There has been a large amount of data gathered on the prevalence of E. coli

O157:H7 in meat. However, other serotypes have been implicated in food-borne

outbreaks as well and these strains appear to be more important in Australia.


R e a d y - t o - e a t f o o d

Clostridium perfringens in Australian foodservice operations

Debate has recently arisen over the need for stricter cooling regimens after the

cooking step in roast beef manufacturing. Preliminary analysis of growth data

for C. perfringens during cooling conducted in collaboration between Food

Science Australia, the University of Tasmania, and Health Canada pinpointed

the lag time as one of the most important parameters in determining the amount

of growth of the organism (Lammerding et al. 1999c). The case study highlighted

the importance of data collection and predictive microbiology to support

scientifically valid processing guidelines. The primary focus of the model

was to investigate the impact that food service has on the risk following

both a well-controlled process as well as a process that may have experienced

some deviations.

S e a f o o d p r o d u c t s

Identification and prioritisation of hazards across the supply chain

This activity was commissioned by SafeFood Production NSW as a two-step

process involving Food Factotum Pty Ltd (Walsh 1999) and University of

Tasmania (Ross and Sanderson 2000) to serve as a basis to develop a food

safety scheme for the seafood industry. Raw finfish intended for cooking has

few associated hazards. The growth in raw fish consumption, however, may

have introduced a greater risk from parasitic infection and histamine poisoning.

Vacuum or modified atmosphere packaging (MAP) of raw fish and its more

common use for smoked fish introduces the risk from Clostridium botulinum

growth and the production of toxin and listeria growth, respectively. Cooked

seafood can be contaminated during the cooling and packaging processes. The

most hazardous seafood group is the bivalve molluscs, particularly when eaten

raw or undercooked. For each of the hazards identified as a high priority, the

University of Tasmania assessed the likelihood and severity of the hazard. Ten

hazards were ranked in the following priority order: viruses in bivalve molluscs,

algal biotoxins in bivalve molluscs, ciguatera toxin in fish, histamine in fish,

vibrio species in seafood, enteric bacteria in shellfish, L. monocytogenes in ready-

to-eat smoked fish products, botulism in vacuum packaged ready-to-eat seafood

products, parasites in seafood and mercury in fish.

Risk assessment tools for the seafood industry

A study for Seafood Services Australia to develop a simple model for risk

assessment in the seafood industry has recently been completed (Sumner 2001a).

The project supported the development of a computerised model (Ross and

Sumner 2001) and its application among regulators and operators in the

seafood industry.



Risk assessment of products made from unpasteurised goat

and sheep milk

A study supported by Safefood Production NSW has been completed to provide

the background for a risk-based Food Safety Scheme for the Goat and Sheep

Milk industries to be introduced by regulation in early 2001 (Anon. 2000c).

P l a n t p r o d u c t s

Plant product scoping study

A study by Food Science Australia for Safefood Production NSW was undertaken

to set priorities for further activities in the plant area based on an objective

assessment of food safety risks (Richardson 2000). This is considered a scoping

study for future work and includes qualitative risk assessments of a number of

plant products. The risks have been ranked from high to low depending on the

risk for initial contamination (high for edible product growing in soil) and usual

preparation methods (high for consuming raw).

G r a i n s

Risk management activities in the grains industry

A project was recently initiated by the Grains Research and Development

Corporation and undertaken by the University of Tasmania to set priorities

for risk-management activities in the grains industry (Ross 2001b). Results

are not yet available.

P i g c a r c a s e p o s t m o r t e m i n s p e c t i o n

Foodborne hazards in abnormal and normal tissues of pig carcases

A systematic quantification of food-borne hazards in abnormal and normal

tissues of pig carcases was undertaken to provide a risk-based assessment of the

effectiveness of traditional organoleptic meat inspection (Pointon et al. 2000).


In this study the risk of illness from the consumption of hamburgers

contaminated with pathogenic STEC was estimated using dose-response

information obtained for Shigella (Lammerding et al. 1999d). The model

developed looked at all stages of production beginning at the processing

of cattle and extending to the consumption of ground beef. The model

incorporates several parameters that are specific to Australian production,

in particular the prevalence and concentration of STEC in Australian cattle.

As well as providing an estimate of the risk to the Australian population, the

model focuses on possible process improvements using various hypothetical

risk mitigation strategies (i.e. decontamination, irradiation, consumer awareness,

better control at retail etc.).

S m a l l g o o d s

Escherichia coli O157 in fermented meat products in Australia

This case study is a process risk model for the manufacture of salami, although

the effect of raw material quality on the final risk estimate is also investigated

(Lammerding et al. 1999b). The process is modelled from receipt of raw materials

to final consumption and the risk to the Australian population estimated. The

case study incorporates aspects of predictive microbiology and presents data

generated from other research in the form of a mathematical model describing

the reduction in E. coli O157:H7 as a function of the process parameters. The

model identified E. coli O157:H7 concentration in the raw material and a high

pH after fermentation as being positively correlated to the final health risk,

while fermentation time and storage temperature were identified as important

negatively correlated processing parameters (i.e. increasing these values

decreased the final risk).

Listeria monocytogenes in smallgoods

A study has also recently been initiated by Meat and Livestock Australia to

quantify risks involving Listeria monocytogenes in smallgoods (Sumner 2001b).

Results are not yet available.

D a i r y

Priorities for risk management activities in the dairy industry

A project being undertaken by the University of Tasmania for the Dairy Research

and Development Corporation to set priorities for risk management activities

in the dairy industry has recently been initiated (Ross 2001a). Results are not

yet available.




G a p b e t w e e n s c i e n c e a n d p r a c t i c e

Because of the recent evolution of this area there is a substantial amount of scientific activity

but less activity in the areas of practical use and application of the methodology. Some of the

scientific through-chain work has been developed in a modularised fashion to allow industry

to apply the results for their part of the chain. Typically, practical applicability has been lacking.

S u g g e s t e d a c t i o n s

1. Support potential research that provides linkages between risk assessment and the practical

application of these findings within industry sectors.

L a c k o f u s e f u l t o o l s

Some risk assessment models developed are often very complex and difficult to communicate.

Considerable expertise is required to translate the models into practical food industry actions.


1. Evaluate some of the risk assessment tools already available.

2. Develop practical tools that apply to industry, which can be used by HACCP practitioners.

L a c k o f p r a c t i c a l i n d u s t r y e x a m p l e s

Most of the work so far in the area has been through scientific institutions building full

through-chain industry models or governments developing models to set priorities for policy

and regulations.


1. Industry case studies that apply risk assessment to HACCP-based management systems.

2. Explore in-house risk-assessment work. within large food companies.


L a c k o f c o o r d i n a t i o n

A gap analysis was undertaken to identify issues and suggest actions to improve the utilisation

of risk assessment in HACCP-based industry quality systems. The review found the following

points worthy of further consideration:

C o o r d i n a t i o n e f f o r t s

Within Australia there seems to be fragmented efforts in pursuing qualitative and quantitative

risk assessments.


1. Creating a risk assessment clearing house for Australian information.

2. Organise a conference/meetings to identify the current status of risk assessment

within Australia.

3. Ensure that Australia is tapped into the Codex work

L a c k o f k n o w l e d g e

Risk assessment methodology has progressed in the chemical area but less so in the

microbiological area. In recent years, general information on the methodology for microbial

risk assessment was published and terminology for the work further developed. This information

is still to filter through to industry and the practicality of the published information tested.


1. Organise hands-on industry workshops to disseminate information.

2. Develop an industry manual for quantitative microbial risk assessments.

L a c k o f i n f o r m a t i o n s o u r c e s

It is currently difficult to get general and specific information related to international risk-

assessment activities. There are some dedicated international websites covering this type

of information. There are also international risk associations dealing with food topics among

many other topics.


1. Conduct an international review of risk assessment activities.

G a p a n a l y s i sGap analysis


Buchanan, R. & Lindqvist, R. (2000) Hazard identification and hazard

characterization of Listeria monocytogenes in ready-to-eat foods. Food and

Agriculture Organisation, Rome.

Buchanan, R, Lindqvist, R., Ross, T., Todd, E., Smith, M. & Whiting, R. (2001)

Risk assessment of Listeria monocytogenes in ready-to-eat foods. Food and


Buchanan, R. L. & Whiting, R. C. (1998) Risk assessment: a means for linking HACCP

plans and public health. Journal of Food Protection, 61, 1531-1534.

Carlin, F., Girardin, H., Peck, M. W., Stringer, S. C., Barker, G. C., Martinez,

A., Fernandez, A., Fernandez, P., Waites, W. M., Movahedi, S., van Leusden,

F., Nautam M, Moezelaar, R., Torre, M. D. & Litman, S. (2000) Research on factors

allowing a risk assessment of spore-forming pathogenic bacteria in cooked chilled

foods containing vegetables: a FAIR collaborative project. International Journal

of Food Microbiology, 60, 117-135.

Cassin, M. H., Lammerding, A. M., Todd, E. C., Ross, W. & McColl, R. S. (1998)

Quantitative risk assessment of Escherichia coli O157:H7 in ground beef hamburgers.

International Journal of Food Microbiology, 41, 21-44.

Clark, A.J. & Brinkley, T. (2001) Risk management for climate, agriculture and policy.

Bureau of Rural Sciences, Commonwealth of Australia, Canberra.

Depaola, A., Karunasagar, I., Osaka, K., Sumner, J. & Walderhaug, M. (2001)

Exposure assessment of Vibrio spp. in seafood. Food and Agriculture Organisation,

Rome.

Dormont, D., Budka, H., Kretzschmar, H., Pocchiari, M., Sommerville, R., Ostrhaus,

A., Bradley, R. & Schlatter, J. (2000) Oral exposure of humans to the BSE agent:

Infective dose and species barrier. European Commission Scientific Steering

Committee, Brussels.

Ebel, E. D., Kasuga, F., Schlosser, W. D. & Yamamoto, S. (2000) Exposure assessment

of Salmonella enteritidis in eggs. Food and Agriculture Organisation, Rome.

Fabiansson, S.U. & Cunningham D.C. (2000) Evolution of Australian food quality

systems. Bureau of Rural Sciences, Canberra.

Farber, J. M., Ross, W. H. & Harwig, J. (1996) Health risk assessment of Listeria

monocytogenes in Canada. International Journal of Food Microbiology, 30, 145-156.

Fazil, A., Lammerding, A. M., Morales, R. A., Vicari, A. S. & Kasuga, F. (2000)

Hazard Identification and Hazard Characterization of Salmonella in broilers

and eggs. Food and Agriculture Organisation, Rome.


Anderson, S., Bak Christensen, B., Fazil, A., Hartnett, E., Lammerding, A., Nauta,

M., Paoli, G. & Rosenquist, H. (2001). Hazard identification, hazard characterisation

and exposure assessment of Campylobacter spp. in broiler chicken. Food and


Anon. (1983) Risk assessment in the federal government: managing the process.

National Research Council, Washington (DC).

Anon. (1996) Framework for the assessment and management of food-related

health risks. Australia New Zealand Food Authority, Canberra.

Anon. (1999) Principles and guidelines for the conduct of microbiological risk

assessment. CAC/GL-30 (1999) Codex Alimentarius Commission, Rome.

Anon. (2000a) First report on the harmonisation of risk assessment procedures.

European Commission Health & Consumer Directorate, Brussels.

Anon. (2000b) International animal health code. Office International des Epizooties,

Paris.

Anon. (2000c) Risk assessment of products made from unpasteurised goat and sheep

milk. SafeFood Production NSW, Sydney.

Baker, A. R., Ebel, E. D., Hogue, A. T., McDowell, R. M., Morales, R. A., Schlosser,

W. D. & Whiting, R. C. (1998) Salmonella enteritidis risk assessment—shell eggs and

egg products. Food Safety and Inspection Service, Washington, DC.

Bemrah, N., Sanaa, M., Cassin, M. H., Griffiths, M. W. & Cerf, O. (1998) Quantitative

risk assessment of human listeriosis from consumption of soft cheese made from raw

milk. Preventive Veterinary Medicine, 37, 129-145.

Berends, B. R. & van Knapen, F. (1999) An outline of a risk assessment-based system

of meat safety assurance and its future prospects. Veterinary Quarterly, 21, 128-134.

Brown, M. H., Davies, K. W., Billon, C. M., Adair, C. & McClure, P. J. (1998)

Quantitative microbiological risk assessment: principles applied to determining the

comparative risk of salmonellosis from chicken products. Journal of Food Protection,

61, 1446-1453.

R e f e r e n c e sReferences


McNab, W. B. (1998) A general framework illustrating an approach to quantitative

microbial food safety risk assessment. Journal of Food Protection, 61, 1216-1228.

Miliotis, M. & Watkins, W. (2000) Draft risk assessment on the public health

impact of Vibrio parahaemolyticus in raw molluscan shellfish. U.S Food and

Drug Administration, Center for Food Safety and Applied Nutrition, Washington.

Morales, R. A. (1998) Microbial risk assessment, economics, and food safety.

Journal of the American Veterinary Medical Association, 213, 1746-1749.

Notermans, S., Dufrenne, J., Teunis, P., Beumer, R., te Giffel, M. & Peeters-Weem,

P. (1997) A risk assessment study of Bacillus cereus present in pasteurised milk.

Food Microbiology, 14, 143-151.

Notermans, S. & Mead, G. C. (1996) Incorporation of elements of quantitative

risk analysis in the HACCP system. International Journal of Food Microbiology,

30, 157-173.

Pointon, A. M., Hamilton, D., Kolega, V. & Hathaway, S. (2000) Risk assessment

of organoleptic postmortem inspection procedures for pigs. The Veterinary Record,

146, 124-131.

Potter, M. E. (1996) Risk assessment terms and definitions. Journal of Food

Protection, Supplement, 6-9.

Richardson, K. (2000) Plant product scoping study. Food Science Australia,

Rodricks, J. V. (1999) Food Safety Policy, Science, and Risk Assessment: Strengthening

the Connection, Washington DC. National Academic Press, p 25-27.

Ross, T. (2001a) Personal communication on risk assessment in the dairy industry.

Ross, T. (2001b) Personal communication on risk assessment in the grains industry.

Ross, T. & Sanderson, K. (2000) A risk assessment of selected seafoods in NSW.

School of Agricultural Science, University of Tasmania, Hobart.

Ross, T. & Sumner, J. (2001) A simple spreadsheet-based food safety risk assessment

tool. International Journal of Food Microbiology, Submitted.

Ross, T., Todd, E. C. & Smith, M. (2000) Exposure assessment of Listeria

monocytogenes in ready-to-eat foods. Food and Agriculture Organisation, Rome.

Sumner, J. (2001a) Personal communication on seafood project.

Sumner, J. (2001b) Personal communication on smallgoods project.


Fazil, A., Lammerding, A. M., Kasuga, F., Ebel, E., Kelly, L., Anderson, W. & Snary,

E. (2001) Risk characterisation of Salmonella in broilers and eggs. Food and


Healy, M. (2001) Personal communication on risk assessment of Listeria

monocytogenes in food products.

Käferstein, F. K., Motarjemi, Y. & Bettcher, D. W. (1997) Foodborne disease control:

a transnational challenge. Emerging Infectious Diseases, 3, 503-510.

Kaplan, S. (1997) The words of risk analysis. Risk Analysis, 17, 407-417.

Kelly, L., Anderson, W. & Snary, E. (2000) Exposure assessment of Salmonella spp.

in broilers. Food and Agriculture Organisation, Rome.

Kindred, T. P. (1996a) Risk analysis and its application in FSIS. Journal of Food

Protection, Supplement, 24-30.

Kindred, T. P. (1996b) Risk assessment and its role in the safety of foods of animal

origin. Journal of the American Veterinary Medical Association, 209, 2055-2056.

Klapwijk, P. M., Jouve, J. L. & Stringer, M. F. (2000) Microbiological risk assessment

in Europe: the next decade. International Journal of Food Microbiology, 58, 223-230.

Lammerding, A. M. & Fazil, A. (2000) Hazard identification and exposure assessment

for microbial food safety risk assessment. International Journal of Food Microbiology,

58, 147-157.

Lammerding, A. M., Fazil, A., Paoli, G. M., Desmarchelier, P. & Vanderlinde,

P. (1999a) E. coli O157:H7 in fresh ground beef manufactured from Australian beef:

Issues of equivalence. Meat and Livestock Australia, Sydney.


P. (1999b) Escherichia coli O157 in fermented meat products in Australia. Meat

and Livestock Australia, Sydney.


P. (1999c) A quantitative risk assessment model for C. perfringens in Australian

foodservice operations. Meat Research Corporation, Sydney.


P. (1999d) Shiga toxin-producing E. coli in ground beef manufactured from

Australian beef: Process improvement. Meat and Livestock Australia, Sydney.

Lindqvist and Westö (2000) Quantitative risk assessment for Listeria monocytogenes

in smoked or gravad salmon and rainbow trout in Sweden. International Journal

of Food Microbiology, 58, 181-196.



H a z a r d i d e n t i f i c a t i o n

The identification of biological, chemical, and physical agents, capable of causing

adverse health effects, that may be present in a particular food or group of food.

Hazard identification is the first step of the risk-assessment process. It gives a qualitative

indication of the potential hazards arising from consumption of a particular food product.

For chemical agents, hazard identification establishes potential adverse health effects in

humans associated with exposure to a particular substance, a ranking of the likelihood of such

effects occurring and the certainty or uncertainty associated with such effects. Chemical agents

include intentionally introduced chemicals, inadvertent contaminants and naturally occurring

toxicants. Examples could be food additives; residues of pesticides and other agricultural

chemicals; residues from veterinary drugs; chemical contaminants from any source; and

natural toxins, such as mycotoxins and ciguatoxin.

For microbial agents, hazard identification consists of identification of the micro-organisms

and microbial toxins of concern. Information on potential hazardous micro-organisms can be

obtained from surveys of the microbial composition of raw materials and from epidemiological

surveillance of food-borne infections and intoxications. Analysis is complicated by the

enormous genetic variety shown even within identified species of micro-organisms and

their different reactions to food and processing conditions.

E x p o s u r e a s s e s s m e n t

The qualitative and/or quantitative evaluation of the likely intake of biological,

chemical and physical agents via food as well as exposure from other sources

if relevant.

After identification of the hazard, information is needed on the amount contained in the food

product, and on consumption patterns. Estimates of dietary intakes of food additives, residues

of pesticides and veterinary drugs and contaminants require information on the consumption

of relevant foods and the concentration of the chemical of interest in those foods. In general,

exposure estimates are based on known or anticipated dietary information for particular foods

together with an estimate of the level of the chemical in particular commodities.

For microbiological agents, information is needed on their numbers or the level of bacterial

toxin and the distribution in food components that are intended for processing. Such

information provides the basis for determining the effects of processing and product

composition on the level of contamination of the end-product at the time of consumption.


Todd, E. C. (1996) Risk assessment of use of cracked eggs in Canada. International

Journal of Food Microbiology, 30, 125-143.

Troutt, H. F., Gillespie, J. & Osburn, B. I. (1995) Implementation of HACCP program

on farms and ranches. In HAACP in meat, poultry and fish processing; A. M. Pearson

& T. R. Dutson, Eds.; Blackie Academic & Professional: Glasgow, 1995; pp 36-57.

van Gerwen, S. J., te Giffel, M. C., van’t Riet, K., Beumer, R. R. & Zwietering,

M. H. (2000) Stepwise quantitative risk assessment as a tool for characterization

of microbiological food safety. Journal of Applied Microbiology, 88, 938-951.

Vose, D. J. (1998) The application of quantitative risk assessment to microbial food

safety. Journal of Food Protection, 61, 640-648.

Voysey, P. A. (2000) An introduction to the practice of microbiological risk assessment

for food industry application. Campden & Chorleywood Food Research Association

Group, Chipping Campden.

Walderhaug, M. & Bowers, J. (2001) Hazard characterisation of Vibrio spp.

in seafood. Food and Agriculture Organisation, Rome.

Walsh, P. (1999) Identification and prioritisation of hazards across the supply chain.

Food Factotum, Tasmania.

Whiting, R. C. & Buchanan, R. L. (1997) Development of a quantitative risk

assessment model for Salmonella enteritidis in pasteurized liquid eggs. International

Journal of Food Microbiology, 36, 111-125.

Whiting, R. C. & Long, W. (2001) Draft Assessment of the Relative Risk to Public

Health from Foodborne Listeria monocytogenes Among Selected Categories of Ready-

to-Eat Foods. FDA/Center for Food Safety and Applied Nutrition, Washington.


A p p e n d i x 1 — Q u a n t i t a t i ve r i s k a s s e s s m e n tAppendix 1– Quantitative risk assessment


For chemicals, risk characterisation might be expressed as a margin of safety between the

acceptable level of intake of an additive or contaminant, based on the known hazard, and the

known level of human exposure via the diet. When no threshold is evident, a more qualitative

estimate of risk is necessary.

For microbial agents, risk characterisation is usually a qualitative description of the

circumstances under which a food may be contaminated at a level that would potentially pose

a risk to human health. It may include reference to a sampling plan, which should conform to

the microbiological limits.

In principle, if the exposure estimate is reliable and appropriate and the dose-response data

of good quality and covers this exposure range, risk characterisation is a quite straightforward

process. However, this is often not the case and the degree of confidence in the final

estimation of risk depends on the variability and uncertainty factors identified in previous

steps. The two components, variability and uncertainty, describing the degree of reliability

of the risk estimate, should be clearly and distinctly described. The variability is the effect

of chance and is a function of the system, whereas uncertainty is the assessor’s lack

of knowledge about a given parameter and may include parameter uncertainty, model

uncertainty and scenario uncertainty. The separate effect of variability and uncertainty

on the risk estimate should be made clear.


Several techniques, such as surveillance tests, storage tests and microbiological challenge

testing, can be used for this assessment. Predictive microbiology is also a useful tool for

determining effects of factors that control the safety of food products.

In the case of microbiological agents, there is an added dimension of micro-organism

replication under particular processing and storage conditions. For chemical agents, too,

there are additional considerations; for example, it is necessary to give proper consideration

to the role of degradation products and metabolites in either increasing or decreasing specific

adverse effects.

H a z a r d c h a ra c t e r i s a t i o n ( o r d o s e - r e s p o n s e r e l a t i o n s h i p )

The qualitative and/or quantitative evaluation of the nature of the adverse health

effects associated with the hazard.

Hazard identification is primarily a question of identifying the effects that are considered

as adverse. Hazard characterisation is centred on the quantification of these effects so that

a dose-response relationship can be established. Such a definition equally applies to chemical

and microbiological risk assessment. However, it has to be realised that the response of a

human population to a pathogen is highly variable. The characteristics of the pathogen will

affect its pathogenicity and virulence. The number of microbes that come into contact with

a host and overcome the natural barriers will determine infection. The general health or

immune status of the host may determine translation of infection into illness and its severity.

The attributes of the vector (in particular, in the case of a food product) can also influence

the outcome. The likelihood that any specific individual may become ill due to an exposure

to a microbial pathogen is dependent on the integration of the pathogen, the host and the

matrix effects. In addition to characterising the nature of the adverse health effects, analysing

and evaluating these interactions based on collating information on the pathogen, the host

and the matrix, are keys to microbiological hazard characterisation.

R i s k c h a ra c t e r i s a t i o n

The qualitative and/or quantitative estimation, including attendant uncertainties,

of the probability of occurrence and severity of known or potential adverse

health effects in a given population based on hazard identification, hazard

characterisation and exposure assessment.

Risk characterisation brings together information gained from previous steps and provides

a practical estimate of risk for a given population. Risk-management strategy is then

formulated on the basis of this determination.



E x t e r n a l c r o s s - r e f e r e n ce

Expert opinion should be cross-referenced against the opinions of other experts

and the body of published evidence (the scientific model of a risk). This can be

developed repeatedly throughout the course of the risk analysis.

I n t e g ra t i o n

The role of the analyst is to integrate a broad range of evidence, targeting it towards

the analysis criterion. This is an iterative process that requires careful questioning

of the evidence, re-assessment and finally the formation of an evidence-based

judgment.

Co m m u n i c a t i o n

Such judgment needs to be communicated to decision-makers. An appraisal

of the precision in the analysis and the research methodology is also important.


Documenting expert opinion on risk properties uses a structured research

methodology to minimise subjective biases. The role of the risk analyst is to

go beyond collecting expert evidence to developing a critical evaluation. This

description is based on the methodology published by Clark and Brinkley (2001).

It uses some or all of the following checks and balances:

E x p e r t c r e d i b i l i t y

To establish the credibility and qualifications of potential experts to provide opinion

relating to the specific risk. This may involve understanding the potential biases

an expert has in providing evidence.

Co n t ex t

To provide enough contextual information about the risk-management process

to the expert without highlighting the specific implications of any decisions.

This can maintain the independence of the expert.

I n f o r m a t i o n g a t h e r i n g

The experts can be asked to provide opinion:

• casually,

• in published reports,

• in formal interviews, or

• in a meeting of experts (comparative analysis).

Each of these has different checks and balances that help to minimise biases; for

example, published scientific work has been through a high level of critical review

by other experts in the field.

I n t e r n a l c r o s s - r e f e r e n ce

Experts can be asked to re-state or clarify their opinions at a later stage and this can

be cross-referenced with the original statement by the analyst. This accounts for

emotional or other biases.

A p p e n d i x 2 — E x p e r t o p i n i o nAppendix 2 – Expert opinion


7. How uncertain is this estimate?

Data from the specific food/microbe combination or similar sets.

Score

Sub-score occurrence

Exposure assessment (effect of processing/decontamination)

8. What is the effect of storage before processing

on the hazard contamination level?

Use in-house data and literature data as appropriate.

Document the data separately for future reference.

Score

9. What is the effect of processing and decontamination

steps on the contamination level of the micro-organism?

Consider initial numbers and the impact of processing and decontamination.

All assumptions must be clearly stated. Where there is no data, the presence

or survival of the hazard at a hazardous dose at consumption must be assumed.

Score

10. What is the combined effect of storage and processing

on toxin formation?

Consider toxin concentration as it relates to organism numbers

and environmental conditions conducive of toxin formation.

Score

11. What is the uncertainty of this estimate?

High levels of uncertainty may be associated with this type of assessment,

making it more-or-less useless—for example if the product is safe with high

uncertainty. Default conditions offering the highest level of consumer protection

should be used.

Score

Sub-score processing/decontamination

Exposure assessment (potential for re-contamination)

12. What is the frequency of re-contamination of the product

after processing?

This can include an estimate only, but it is better if the score is based

on real data and calculation of the actual frequency.

Score


This text and questionnaire contain the elements of a risk assessment and indicate the data

needed for it to be undertaken. It is based on a guideline published by Campden & Chorleywood

Food Research Association Group in 2000. Because many companies wishing to undertake risk

assessment will not have ready access to experts, values provided in this appendix represents the

likely ranges for the key determinants of risk. When using this questionnaire it is essential to

remember that the risk estimate will only be as good as the data used in the assessment. This

questionnaire will allow the user to recognise the features of their process and product exerting

the biggest influence on the level of risk, to distinguish between the seriousness of various

hazards, and to visualise the risk profile of the product under examination.

Going through the risk profile considerations, it can be seen that each step contributing to the

score can be rated from low to high and each ranking score is associated with a number from

1 (low) to 5 (high). When all the questions have been answered the ranking score is made up

of values between 1 and 5. The higher the value, the higher the risk. For infectious pathogens

the maximum score is 75 and for toxin-forming pathogens the maximum score is 10 with

a maximum quality score of 55.

Hazard Identification (descriptive part – no scoring)

1. What type of product and product characteristics Name:

are covered by this risk assessment? Type:

2. What pathogenic organisms can be associated with the product? Names:

3. Microorganism specifically covered by this risk assessment? Name:

4. Can the microorganism act through a pre-formed toxin? Yes/No

Please note the box placement in relation to the scoring category:

Exposure Assessment (occurrence of the pathogenic microorganism)

5. What is the frequency of contamination of the raw materials

making up the product?

Use in-house data and literature data as appropriate. Also consider

toxin formation.The conditions for toxin production need to be

known to assess this. Document the data for future reference.

Score

6. What is the contamination level range found

in the raw materials?

Use in-house data and literature data as appropriate. Document the data

for future reference.

Score

1 – never: 0

2 – very low frequency: 1/1000

3 – low frequency: 1/100

4 – high frequency: 1/10

5 – always: 1/1

1 – 0-10 cells/g

2 – 0-102 cells/g

3 – 0-103 cells/g

4 – 0-104 cells/g

5 – >104 cells/g

Quality Profile Toxin Score General Profile

1 – accurate data on relevant microbe/food

2 – accurate data on similar microbe/food

3 – quantitative general data on similar

microbe/food

4 – qualitative general data on similar microbe/food

5 – opinion/default, no hard data

1 – 0-10 cells/g

2 – 0-102 cells/g

3 – 0-103 cells/g

4 – 0-104 cells/g

5 – >104 cells/g

1 – inactivation: at least 6-log decrease

2 – inactivation: between 3 and 6-log

3 – no changes: survival

4 – slow growth: less than 3-log increase

5 – rapid growth: at least 3-log increase

1 – at least 6-log inactivation, no toxin

2 – between 3 and 6-log inactivation, no toxin

3 – no changes: survival, no toxin production

4 – low growth (<3-log), low chance of toxin

5 – rapid growth (>3-log), high chance of toxin



3 – quantitative general data on similar microbe/food



1 – never: 0




5 – always: 1/1

A p p e n d i x 3 — Q u e s t i o n n a i r e r a n k i n g t o o lAppendix 3 – Questionnaire ranking tool


19. What is the likelihood of toxin formation during storage?

This question has to consider both the likely number of organisms

(previous question) and the impact of storage conditions on toxin formation.

Score


Score

Sub-score storage

Exposure assessment (open shelf life to be considered only when relevant )

21. Is the product intended as single use or multi-use where

22. What is the effect of open shelf-life storage on the level

of microorganism?

This includes storage of the product throughout the chain

from production to consumption.

Score


Score

24. What is the likelihood of growth and toxin formation

during open shelf-life?

This question has to consider both the likely number of organisms

(previous question) and the impact of storage conditions on toxin formation.

Score


Score

Sub-score


13. What is the likely concentration of the pathogen

from re-contamination after processing?

This can include an estimate only, but it is better if the score

is based on real data and calculation of the actual level.

Score


Score

Sub-score potential for re-contamination

Exposure assessment (packaging)

15. What is the frequency of re-contamination at or after packaging?

This can include an estimate only, but it is better if the score is based on real data

and calculation of the actual frequency. Whether the product is packaged before

or after possible decontamination will be significant for this estimate.

Score

16. What is the likely level of re-contamination at or after packaging?

This can include an estimate only but it is better if the score is based on real data

and calculation of the actual level. Whether the product is packaged before

or after possible decontamination will be significant for this estimate.

Score


Score

Sub-score

Exposure assessment (storage)

18. How does the level of the microorganism change during storage?

This includes storage of the product throughout

the chain from production to consumption.

Score


1 – 0-10 cells/g

2 – 0-102 cells/g

3 – 0-103 cells/g

4 – 0-104 cells/g

5 – >104 cells/g




microbe/food



0 – not applicable

1 – never: 0




5 – always: 1/1


1 – 0-10 cells/g

2 – 0-102 cells/g

3 – 0-103 cells/g

4 – 0-104 cells/g

5 – >104 cells/g












1 – never

2 – unlikely

3 – possible

4 – likely

5 – certain




microbe/food



Single use Go to 26

Multi-use Go to 22











1 – at least 6-log inactivation, no toxin

2 – between 3 and 6-log inactivation, no toxin

3 – no changes: survival, no toxin production

4 – low growth (<3-log), low chance of toxin

5 – rapid growth (>3-log), high chance of toxin








Score

Sub-score food intake

Hazard characterisation

33. Who are the likely consumer groups for the product?

Examples include families/babies/elderly

34. Nominate a sub-group at risk!

Nominate each at-risk group for a separate analysis should the sensitivity differ.

If there is only one homogenous consumer group with similar sensitivity,

one analysis will suffice.

35. What is the severity of the hazard?

The sensitivity of each group should be considered or that of the most sensitive

consumers should be used for a single assessment.

Score

36. What is the hazardous level of the microorganism

covered by this risk assessment?

The generally accepted infectious dose, boundaries of toxin formation

or toxin persistence should be used.

Score


There will be high uncertainty with this type of risk estimate, it will only

be generally valid and take account of the consumers with ’text-book’ sensitivity.

Score

Sub-score hazard characterisation

Risk characterisation

Risk profile – general score Min 14 Max 75

Risk profile – toxin score Min 5 Max 25

Information quality profile Min 11 Max 55


Exposure assessment (consumer usage and preparation)

26. What is the effect of end-use preparation

and product usage on the level of the hazard?

Assessment should be based on use or portion size instructions,

unless there are indications that other practices can significantly

alter risk or dose (e.g. under-cooking). Likely abuse should

be considered.

Score


Score

28. What is the effect of end-use preparation and product

usage on toxin level and production?

Score

29. What is the probability of toxin presence at the point

of consumption?

Score


Score

Sub-score consumber usage & preparation

Exposure assesssment (food intake)

31. What is the likely quantity of the food consumed by a customer

on a specified occasion or over a period of time?

Score













1 – eliminate

2 – reduce

3 – unchanged

4 – slight increase

5 – large increase


1 – never: 0




5 – always: 1/1







1 – very low intake: 0 - 10 g

2 – low intake: 10 - 50 g

3 – medium intake: 50 - 100 g

4 – high intake: 100 - 200 g

5 – very high intake: >200 g






Groups:

Groups:

1 – mild symptoms, prompt recovery

2 – mild symptoms for a few days

3 – generally mild symptoms, hospitalisation

4 – severe symptoms, hospitalisation, deaths

5 – several deaths

1 – > 10,000 cells

2 – high minimum dose: 1000 – 10 000 cells

3 – minimum dose: 100 – 1000 cells

4 – low minimum dose: 10 – 100 cells

5 – very low minimum dose: 0 – 10 cells








This is a simple tool for food safety risk-assessment developed by Ross and Sumner

(2001) to assist decisions by food safety risk managers. The tool is in spreadsheet

software format and embodies established principles of food safety risk-assessment,

i.e. the combination of probability of exposure to a food-borne hazard, the

magnitude of hazard in food when present, and the probability and severity

of outcomes that might arise from that level and frequency of exposure.

The tool requires the user to select from qualitative statements and/or to provide

quantitative data concerning factors that will affect the food safety risk arising

from a specific food product, and specific hazard during the steps from harvest

to consumption. The spreadsheet converts the qualitative inputs into numerical

values and combines them with the quantitative inputs in a series of mathematical

and logical steps using standard spreadsheet functions. Those calculations are used

to generate indices of the public health risk.

The model underpinning the tool is a gross simplification of the harvest-to-

consumption pathway, but the tool offers a quick and simple means of comparing

food-borne risks from diverse products and is useful for ranking and prioritising

risks from diverse sources. It can be used to screen food-borne risks and identify

those requiring more rigorous assessment. It also serves as an aid to structured

problem-solving and can help to focus attention on those factors in food production,

processing, distribution and meal preparation that most affect food-safety risk and

that may be the most appropriate targets for risk-management strategies.

The risk model user interface, using Australian populations as an example, is shown

on the following page. Users mouse-click on their choice in each list box or provide

numerical values as required. As choices are made and values entered, the risk

estimates are automatically recalculated.


A p p e n d i x 4 — C o m p u t e r i s e d r i s k r a n k i n gAppendix 4 – Computerised risk ranking


Acceptable Level Dangerous Level

Description of Food Safety Problem and ContextWhat is the specific hazard to be explored?

Name:

Review criticalcontrol points

NEVEROFTEN

OK whenprocess is

under control

Make sureprocess is

under control

ExposureAssessment

Will end-use preparation reducethe hazard to acceptable

levels?

SOMETIMESALWAYS

Process undercontrol


ExposureAssessment

Will process steps (critical controlpoints) reduce the hazard

to acceptablelevels?

LOW

Less stringentprocess required


HIGH

ExposureAssessment

What is the likelihoodof contamination of raw

materials or process?

SEVEREMILD

Less stringentprocess required

Seek expertadvice


What are the consequencesand severity of adverse

effects?

HIGHLOW

Process undercontrol

Seek expertadvice

RiskCharacterisation

What is the integrated risk measureof hazard characterisation and

exposure assessment?

LOW

LOW

LOW

LOW

LOW

No Yes Don’t know

0 1 Find out

Mild Major Severe

0.1–0.5 0.5–0.8 0.8–0.1

Low Medium High

0.1–0.5 0.5–0.8 0.8–1

Often Sometimes Never

0 0.5 1

Range between 0 (low) and1 (high) risk (when multiplying

with 0 the answer is 0 whichindicates an absolute control point).

(number of CCPs as appropriate)

DON’T KNOWNO

Move tonext hazard

Seek expertadvice


Is the agent capable of causingadverse health effects?

LOW

Semi-Quantitative Analysis

Risk Estimate

1

2

3

CCP

1

4

0

0

Always

0

0

1

1

Sometimes

1

1

0.1–0.9

0.1–0.9

Often

0.1–0.9

0.1–0.9

X

X

X

X

X

Figure 7. Decision tree flow chart (to the left is a structured way to arrive at a qualitative risk estimate; below is an attempt to undertake a semi-quantitative risk estimate—the numbers in each row should be multiplied to arrive at the risk estimate number. When a range is given estimate specific number).


A decision tree is a structured way of evaluating risks by considering the different

components of risk assessment. The complexity increases further down the tree

when multiple influences are considered. Figure 7 (overleaf) is an attempt to

produce a possible decision tree covering the four areas of risk assessment.

One form should be completed for each identified hazard. The left-hand side

covers a qualitative approach with weighting of the responses at the end in the risk

characterisation. The more responses that end up to the left, the more acceptable

the risk. The right-hand side allows for a scoring system to be used. The numbers

should be multiplied to end up with an overall score. When multiplying with zero

the risk is negligible. By introducing multiple critical control points the risk can be

further reduced. Instead of a fixed scoring system, a nominated number between

the end points can be used.

The decision tree is a draft only and should be developed further. Microbiological

growth during storage or before consumption has not been included but could be

added by multiplying with numbers greater than one to increase the risk.


A p p e n d i x 5 — D e c i s i o n t r e e f l o w c h a r tAppendix 5 – Decision tree flow chart


Useful literature


Web information

Australia:

Australian Institute of Food Science and Technology http://www.aifst.asn.au

Food Science Bureau (AFGC) http://www.foodsciencebureau.com.au

‘Food Australia’ magazine http://www.foodaust.com.au

Supermarket to Asia http://www.supermarkettoasia.com.au

Food Connect Australia http://www.foodconnect.com.au

Food Info Australia (AFFA) http://www.affa.gov.au/foodinfo

Australia New Zealand Food Authority http://www.anzfa.gov.au/

Food Safety Information Council http://www.safefood.net.au

Food Law and Policy Australia http://www.ausfoodnews.com.au/flapa

Australian Institute of Health and Welfare http://www.aihw.gov.au

CSIRO http://www.csiro.au

International:

Institute of Food Technologists http://www.ift.org

Foodborne Illness Education Center http://www.nal.usda.gov/fnic/foodborne/risk.htm

US Government Food Safety Web Site http://www.foodsafety.gov

(President’s Council on Food Safety)

Center for Food Safety and Applied Nutrition (USFDA) http://vm.cfsan.fda.gov

International food Information Council http://www.ificinfo.health.org

Risk Communication Network http://www.uea.ac.uk/env/cer

(Centre for Environmental and Risk Management)

National Environmental Health Association http://www.neha.org

World Health Organisation (WHO Food Safety Programme) http://www.who.int/fsf/

FoodNet Canada http://foodnet.fic.ca

National Food Safety Database, University of Florida, USA http://www.foodsafety.org

Asian Food Information Centre http://www.afic.org


Microbial Food Contamination Charles L Wilson & Samir Droby

(editors)

CRC press, 2001

http://www.crcpress.com

Food Toxicology William Helferich & Carl K Winter

(editors)

CRC press, 2001


Fundamental Food

Microbiology (second edition)

Bibek Ray CRC press, 2001


Natural Toxicants in Food David H Watson (editor) CRC press, 2001


Interdisciplinary Food

Safety Research

Neal H Hooker & Elsa A Murano

(editors)

CRC press, 2001


Environmental Contaminants

in Food

Colin Moffat & Kevin J Whittle (editors) CRC press, 2001


Food Chemical Safety

Contaminants, vol 1

David Watson (editor) CRC press, 2001


Food Chemicals Codex, fourth

edition, CRCnetBase 2000

Institute of Medicine (editors) CRC press, 2001


HACCP in the Meat Industry Martyn Brown (editor) CRC press, 2001


EU Food law, A Practical Guide Kaarin Goodburn (editor) CRC press, 2001


The Microbiological Safety and

Quality of Food – Two Volumes

Barbara M Lund, Tony C Baird-Parker,

and Grahame W Gould (editors)

Aspen Publishers

Email: [email protected]

International Standards

for Food Safety

Naomi Rees, David Watson (editors) Aspen Publishers

Email: [email protected]

Food Protection Report (monthly newsletter) Pike & Fisher Inc, Maryland, USA

http://www.pf.com

Food Microbiology Newsletter Institute of Food Technologists & The

American Society for Microbiolog

http://www.ift.org/divisions/food_micro/

Title Author(s) Publishe

A p p e n d i x 6 — F u r t h e r r e a d i n gAppendix 6 – Further reading



BSE bovine spongiform encephalopathy

CCFH Codex Commission on Food Hygiene

CCFRA Campden & Chorleywood Food Research Association Group

Codex Codex Alimentarius Commission

CRC Cooperative Research Centre

FSIS US Food Safety and Inspection Service

HACCP hazard analysis critical control point

hazard any adverse agent; an agent with the potential to cause harm

ILSI International Life Science Institute

MAP modified atmosphere packaging

Monte Carlo analysis statistical simulation method

QRA quantitative risk assessment

risk likelihood of harm and severity of impact

SPS Agreement Sanitary and Phytosanitary Agreement

STEC Shiga-toxin producing Escherichia coli

Weibull-Gamma model flexible dose-response model

WHO World Health Organisation


G l o s s a r yGlossary




S C I E N C E F O R D E C I S I O N M A K E R S


S. U. Fabiansson


BRS/Risk Assess Cover.FA 11/7/01 3:18 PM Page 1

Documents

SCIENCE FOR DECISION MAKERS - data.daff.gov.au