Foodborne disease arises from the consumption of microbial pathogens and/or microbial toxins by a susceptible individual. The risk of foodborne disease is a combination of the likelihood of exposure to the pathogen, the likelihood of infection or intoxication resulting in illness, and the severity of the illness. In a system as complex as the production and consumption of food, many factors affect both the likelihood and severity of the occurrence of foodborne disease. Many of these factors are variable, and often there is great uncertainty about the system in general. To effectively manage food safety, a systematic means of examining these factors and gaining a better understanding of the system is necessary. Risk assessment is a process that provides an estimate of the probability and impact of foodborne disease.
Evaluations of the risks associated with foodborne hazards, in general or attributable to specific foods, have been predominately qualitative descriptions of the hazard, routes of exposure, handling practices and/or consequences of exposure. Quantifying any of these elements is challenging since many factors influence the risk of foodborne disease, complicating interpretations of data about the prevalence, numbers and behaviour of microorganisms, and confounding the interpretation of human health statistics. Consequently, policies, regulations and other types of decisions concerning food safety hazards have been largely based on subjective and speculative information. However, advances in our knowledge, analytical techniques and public health reporting, combined with increased consumer awareness, global trade considerations and better understanding of the real economic and social impacts of microbial foodborne illness have moved us toward the threshold of using quantitative risk assessment to support better prioritizing and decision-making in managing food safety (Altekruse, Cohen and Swerdlow, 1997; CAST, 1994; CAST, 1998; Vose, 1998; WHO, 1998).
It is only recently that risk assessment has been applied to microbial hazards, and techniques for its application are still evolving (ILSI, 1996; Kindred, 1996; Lammerding, 1997; NRC, 1996). Unfortunately, the entire field of risk assessment has historically been rift with assignments of different meanings to different terminologies. However, currently accepted food safety definitions and processes are shown in Table 1.1, as defined by the Codex Alimentarius Commission, the international regulatory body for foods (CCFH, 1998; Dawson, 1998).
Risk assessment is a process that is intended to facilitate the description, understanding and management of complex systems (e.g. bacterial spread through the food supply) or contentious issues (e.g. human health risk related trade disputes in agricultural commodities) by providing a framework that allows the evidence and information associated with the issue or systems to be objectively collected and combined to arrive at a conclusion. Risk assessment, in conjunction with risk management and risk communication forms the basis for sound science-based decision-making.
The following section introduces and defines some fundamental concepts that are central to the risk analysis arena.
Definitions of risk analysis terminology for foodborne hazards (Codex Alimentarius Commission)
|Hazard||A biological, chemical or physical agent in, or condition of, food with the potential to cause an adverse health effect.|
|Risk||A function of the probability of an adverse health effect and the severity of that effect, consequential to a hazard(s) in food.|
|Risk analysis||A process consisting of three components: risk assessment, risk management and risk communication.|
|Risk assessment||A scientifically based process consisting of hazard identification, hazard characterization, exposure assessment and risk characterization.|
|Hazard identification||The identification of biological, chemical and physical agents capable of causing adverse health effects and which may be present in a particular food, or group of foods.|
|Hazard characterization||The qualitative and/or quantitative evaluation of the nature of the adverse health effects associated with biological, chemical and physicals agents which may be present in food. For chemical agents, a dose -response assessment should be performed; for biological or physical agents, a dose response assessment should be performed if the data are obtainable.|
|Dose-response assessment||The determination of the relationship between the magnitude of exposure (dose) to a chemical, biological or physical agent and the severity and/or frequency of associated health effects (response)|
|Exposure assessment||The qualitative and/or quantitative evaluation of the likely intake of biological, chemical and physical agents via food, as well as exposures from other source\s if relevant.|
|Risk characterization||The qualitative and/or quantitative estimation, including attendant uncertainties of the probability of occurrence and severity of known or potential adverse effects in a given population based on hazard identification, hazard characterization and exposure assessment.|
|Risk management||The process of weighing policy alternatives in the light of results of risk assessment and, if required, selecting and implementing appropriate control options, including regulatory measures.|
|Risk communication||The interactive exchange of information and opinions concerning risk among risk assessors, risk managers, consumers and other interested parties.|
Risk is a concept encountered by most people on a regular basis. On a daily basis we often sub-consciously assess risks and decide to proceed or not. A very simple example could be the act of crossing the road. There is a risk associated therewith, specifically, a certain chance that we could be struck by a vehicle while crossing. We have already factored two associated dimensions of the risk upon its consideration: the impact of getting struck by the vehicle and the likelihood of getting struck. Risk, then, is defined as a function of two variables, the probability of an event occurring, and the magnitude or severity of the event should it occur. To take the road-crossing example further, if we were to assess the situation where there were hardly any cars on the road, then the probability of being struck would be very low and as such, we would consider the overall risk to be low. Similarly, if for some reason we were living in a world in which cars were made out of paper or travelled very slowly on this road, then the impact or severity of being struck would be low, and again we might consider the risk to be low. In essence, this example illustrates that for the risk to exist, both the likelihood of an event and the impact or severity of an event need to be considered.
An example from the microbial field might be a probability “x” of getting sick from Norwalk virus. In this case the risk is composed of the probability of an individual being exposed to the bacteria (chance of the event occurring) and the impact if the individual is exposed (the chance of the individual becoming sick). The Codex Alimentarius Commission defines risk as a function of the probability of an adverse health effect and the severity of that effect, consequential to a hazard(s) in food.
Risk analysis is a process for managing risk, which is compromised of three distinct but interactive components: risk assessment, risk management and risk communication. Risk analysis is the term used to define the overall process by which risks from foodborne pathogens are handled. The framework for risk analysis adopted by Codex is shown in Figure 1.1.
Codex schematic framework for risk analysis
Key to the risk analysis process is the overlapping and feedback mechanisms built into the three components. The three components are separate and should not exert undue influence on each other; however each should be designed to meet the needs of the other, and as such may have to be interactively and iteratively modified as the analysis progresses.
Risk management as the name implies, involves synthesizing the risk assessment information into some form of action so as to manage it (FAO/WHO, 1995; FAO/WHO, 1997; NACMF, 1998). The goal of risk management is not to achieve “zero risk” but rather to chose and implement scientifically sound, cost-effective, integrated actions that reduce or prevent risks while taking into account social, cultural, ethical, political and legal considerations” (PCCRARM, 1997). Different considerations are due when the risk management issue pertains to international trade, national policy-making, industry interventions, and/or consumer concerns. The tasks involved in risk management have been described as: the determination of what hazards present more danger than society is willing to accept; consideration of what control options are available, and deciding on appropriate actions to reduce (or eliminate) unacceptable risks. At the broadest level, risk management includes a range of management and policy-making activities: agenda setting, risk reduction decision-making, programme implementation, and outcome evaluation (ACS, 1998). Similarly, Codex has defined risk management as the process of weighing policy alternatives in the light of the results of risk assessment and, if required, selecting and implementing appropriate control options, including regulatory measures (FAO, 1997).
Risk assessment is a scientifically based process that systematically compiles and analyses the current data and knowledge about a risk issue. Ideally, the process should produce an objective evaluation of the probability and impact of an adverse effect, relying on published scientific research, surveillance reports, industry data, and, when necessary expert judgment elicited using appropriate methods. There are many frameworks that have been developed to describe the steps required in risk assessment, however most consist of four distinct steps: hazard characterization, exposure assessment, dose-response analysis and risk characterization (CCFH, 1998). The Codex Alimentarius Commission uses the term hazard characterization as the major step between exposure assessment and risk characterization under which dose-response analysis could be done if the data existed.
Risk assessment framework
The first activity in risk assessment is hazard identification, which usually involves an evaluation of the epidemiological data (CCFH, 1998). In addition, this step of the process looks at the issues such as acute vs. chronic disease, sensitive populations, and other complications such as long-term sequelae. During this initial step, the characteristics of the organism and its action should also be recognized, highlighting the mode with which the organism effects the host, such as through the action of toxins, either in the food before consumption or after consumption in the intestine, or, alternatively, through infectious mechanisms. Essentially, the hazard identification step is largely a qualitative evaluation of the available information, and serves to document the important information known about the pathogen, food product and host interface (Lammerding, Fazil and Paoli, 2001).
|Hazard identification||• evidence linking the food and pathogen to human illness|
|• Is there a problem?||• epidemiological investigations|
|• How much of a problem?||• national surveillance databases|
|• Details of the problem?||• microbiological research|
|• process evaluations|
|• clinical studies|
Exposure assessment is the step in the process during which most mathematical modelling takes place. Primarily, the exposure assessment is concerned with estimating the likelihood of being exposed to the hazard via the food product under consideration, and the amount or dose to which that the population or individual is exposed. Microbial risk assessment is faced with a much more dynamic hazard compared to traditional chemical risk assessment because of the potential for microbes to multiply in foods. Assessments concerned with exposures to microbial toxins are faced with a combination of the microbe’s characteristics, and the chemical-like effects of the toxin itself. This step should estimate prevalence and extent of microbial contamination of the product at the time of consumption, the likelihood that an individual consumes the certain food product in a given period of time, the circumstances under which the food was consumed (home prepared vs. institutional or food service, etc.), and the amount of the product consumed at each meal.
Since it is not possible to measure precisely the population of the pathogen present in a food at the time of consumption, models or assumptions need to be developed to estimate the likely exposure. For bacteria, the growth and death of the organism should be accounted for within the food and under predicted handling and preparation practices. Temperature, time, the food chemistry, and competing micro-flora may affect the growth and death rates of pathogens. For viral and parasitic agents that do not grow in foods, the effectiveness of decontamination and/or inactivation steps is of primary concern (Lammerding, Fazil and Paoli, 2001).
|Exposure assessment||•||sources of contamination: frequency, concentration, and an estimation of the probability and concentration that will be consumed|
|•||How many organisms are ingested||•||distribution, growth, inhibition or inactivation from primary by the consumer? contamination, through processing, handling at retail and consumer preparation practices|
|•||How often do they get ingested by the consumer?||•||growth studies, predictive models|
|•||food manufacturer data|
|•||food surveillance data - primary process and retail|
|•||animal/zoonotic disease data|
|•||food composition - pH, Aw, nutrient content, presence of antimicrobial substances and competing micro flora|
The hazard characterization step involves providing a qualitative or quantitative description of the severity and duration of adverse effects that may result from the ingestion of a microorganism or its toxin in food. The quantitative treatment of this step can be interpreted as a dose-response analysis. The purpose of this step is to describe the consequences of exposure to a pathogen as an estimate of the magnitude of adverse effects for an individual consumer or a population. This step also provides the measure for evaluating the value of food safety efforts; e.g. a decrease in the number of people becoming ill and/or the severity of illnesses as a result of an intervention (Lammerding, Fazil and Paoli, 2001).
|Hazard characterization||•||Pathogen: virulence parameters|
|•||How serious is the illness?||•||Food: factors that may protect the organism e.g. high fat content providing increased resistance to gastric acids|
|•||How long does it last?||•||Host susceptibility/resistance factors|
|•||If possible, perform dose-response analysis.||•||Population characteristics|
|•||How likely is infection based on||•||Animal studies the amount ingested?|
|•||Human feeding trials|
|•||Severity, long-term sequallae|
Risk characterization is the concluding task in risk assessment, and is typically the step in the process that puts context around all the prior analysis. The risk characterisation step combines the information generated in hazard identification, exposure assessment and hazard characterization to produce a complete picture of the assessed risk. Codex defines the risk characterization step as the process of determining 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 characterization and exposure assessment. It is important to remember that risk managers should articulate the outcome of the assessment and the types of questions to be answered by the risk assessors at the onset of the assessment process. However, the assessment should at the least strive to answer the following questions in risk characterization:
Risk characterization should also provide insights about the nature of the risk, which are not captured by a simple qualitative or quantitative statement of risk. Such insights include, for example, a description of the most important factors contributing to the average risk, the largest contributions to the uncertainty and variability of the risk estimate, and a discussion of gaps in data and knowledge. The risk assessor may also include a comparison of the effectiveness of alternative methods of risk reduction for consideration by the risk manager (Lammerding, Fazil and Paoli, 2001).
Risk communication involves the exchange of information between risk assessors, risk managers and stakeholders in the risk issue (FAO/WHO, 1998). Since the primary purpose of risk assessment is to inform the risk management decision-making process, it is important that both assessors and managers understand the specific issue of concern. Stakeholders in the risk issue may include, for example, producers, processors, or food handlers, the general public and/or specific subpopulations with increased risk. It is increasingly realized that the entire process of assessing and managing risks should be transparent and interactive. Of particular note in the communication arena is that the participation of all parties during the process tends to increase the acceptability of the final outcome, as opposed to decision-making in an atmosphere that lacks transparency. (FAO/WHO, 1997; FAO/WHO, 1998; PCCRARM, 1997) (Lammerding, Fazil and Paoli, 2001).
There is a quote that is often used in the modelling field that says: “All models are wrong, but some are useful” . This quote captures the essence of why we model a system: not to create a perfect and exact duplicate of reality (which would be an impossible task), but rather to create a tool that will provide insight into the system. In general, we are making simplifications where appropriate in order to represent reality in a way that can help us make informed decisions and allow us to explore the system we are managing
A quantitative risk assessment model for a microbial hazard in a food commodity can be very simple. For instance, it could be a model that uses data on the level of the hazard at the point of consumption and incorporates this with a dose-response model to estimate the risk. This type of model can estimate risk and explore a few limited options; however, it has limited utility, its decision-assistance capacity being limited to the scope it encompasses. In addition, it is often rare to have data on the level of contamination at the point of consumption. Perhaps at best, retail level surveys could provide contamination levels as close to the point of consumption as possible. Since data on the hazard close to the point of consumption is scarce (primarily as a result of low levels, low frequency, test-sensitivity and the resulting costs that would be associated with the number of samples required), it is often necessary to start the modelling downstream of consumption and estimate the changes that occur to the hazard as it progresses along the harvest to consumption chain. In addition, if the goal of the assessment is to detail the pathways leading to exposure so that the impact of the various elements in the pathway can be quantified in terms of their contribution to the overall risk to human health, then a Process Risk Modelling (PRM) approach (Cassin et al., 1998) can be used. This approach has the distinct advantage of assisting in the targeting of management actions, which have the greatest impact on the final health outcome.
The PRM approach is simply a modular approach to modelling a complex system. In the case of a food system, the approach has been to separate the system into modules that logically and sequentially progress in a similar order to that of the food system itself. In the case of seafood products this sequence could start at the harvest stage and progress to the consumption stage, while accounting for all the stages in between (for instance, pre-harvest, harvest, post-harvest, transport, storage, cooking and consumption). The starting point for the model and specific modules into which the system is classified, needs to be determined by the assessor. It considers the questions the assessment is trying to answer and the data and information available.
The process of describing a food system from the rearing of animals to the point of consumption has come to be known as a “farm-to-fork” risk assessment. However, this title is lacking in its ability to capture the diversity inherent in the global food production systems (the terminology reflects farm systems only) and cultural practices (reflecting primarily Western practices of forks for consumption). As a result, we herein refer to risk assessment models that attempt to describe the entire chain of the food production system as “harvest-to-consumption” models.
A schematic representation of the components of a harvest-to-consumption model is shown in Figure 1.3.
Generic schematic representation of harvest-to-consumption model
In addition to the modular concept for describing a system and the idea of characterizing a system from the harvest-to-consumption stage, Figure 1.3 also introduces another key characteristic of the modelling process. In order to estimate risk, there are basically two parameters that we need to consider: the prevalence (depicted in the top box) and the concentration (depicted in the bottom box). Regardless of the complexity and scope of the model employed, the goal of the process risk model is to estimate the changes that occur in the probability of contamination (prevalence) and the level of contamination if the product is contaminated (concentration). The final risk estimate, shown in Figure 1.3 in the last box, is a function of the probability of exposure, which is a direct result of the frequency with which the product is contaminated, and the probability of a response once exposed, which is a function of the level of contamination.
Figures 1.4 and 1.5 illustrate two hypothetical illustrations of how a PRM approach could be used to describe two different hazards in two different products, and the difference in the number and types of modules that may have to be considered. In the first illustration (ciguatoxin in marlin), there are only two stages that need to be considered: the level and prevalence of ciguatoxin in the marlin at harvest, and the consumer’s subsequent probability of exposure and illness. The second illustration shows the steps that might be considered if we were making a model for salmonella in shrimp. In this situation, four stages in the chain are considered: the level and prevalence of salmonella in shrimp after harvest, the level and prevalence of salmonella in shrimp after distribution, the level and prevalence after preparation in the home, and the subsequent probability of exposure and illness upon exposure.
Hypothetical schematic of harvest-to-consumption, PRM approach for ciguatoxin in Marlin
Hypothetical schematic of harvest-to-consumption PRM approach for salmonella in shrimp