Quantitative Risk Assessment

Food safety management can be viewed from different levels: broadly from the perspective of government food safety management systems or more narrowly from that of industry-wide food safety management systems or that of food safety management within an individual establishment. The role of the food industry is to provide safe foods; industry is responsible for establishing food safety management systems that ensure that foods present a minimal risk to the consumer. Risk analysis allows the development of risk-based metrics, such as food safety objectives (FSOs), performance objectives (POs), and performance criteria (PCs), which are addressed in this chapter. Seven hazard analysis and critical control point (HACCP) principles form the framework of a systematic approach to ensure the determination and control of significant food safety hazards associated with a product and process. Monitoring is the first line of defense in the preventive HACCP system. In addition to those related to GMP or sanitation regulations, prerequisite programs can include other programs, such as ingredient specifications, consumer complaint management, glass control programs, allergen management programs, microbiological monitoring of the plant environment, ingredient-to-product traceability programs, and supplier approval programs. PCs can easily be translated into product or process criteria by industry or by government. In integrating a farm-to-fork food safety system, there will be a variety of metrics at points along the food chain.

The handling and consumption of poultry is an important source of human campylobacteriosis. It is widely assumed that the control of Campylobacter in meat-producing poultry, with the aim of reducing the numbers of Campylobacter on poultry meat at the retail level, will reduce the public health burden of human campylobacteriosis. This chapter updates the knowledge of poultry infections and their control at the farm level and extends it to include the interface between the farm and the slaughterhouse. The focus of the chapter remains on the control of Campylobacter in primary production because in this food chain, the gut of living poultry is the only amplification point for Campylobacter. Epidemiological and exposure studies have implicated the handling and consumption of poultry meat as an important source for human campylobacteriosis. Colonization is largely in the ceca and is primarily confined to the intestinal mucous layer over the intestinal crypts of the villi. Reducing human campylobacteriosis is an important public health goal in most industrialized countries. Multispecies farming is a clear risk factor for Campylobacter-positive poultry flocks on the same site. The separation of Campylobacter-positive and -negative flocks and decontamination of the meat from positive flocks is an alternative strategy to reducing the risk of campylobacteriosis in humans. One practical control strategy that can be implemented at the farm-slaughterhouse interface is to separate colonized and noncolonized flocks during processing, and to subsequently treat the meat from Campylobacter-positive flocks.

Conflicting issues often emerge and must be resolved in order to implement an effective strategy to improve the microbiological safety of imported food, as illustrated in the two boxed examples that follow: (i) BSE and the risk of variant Creutzfeld-Jakob Disease (vCJD) from eating beef, and (ii) Salmonella in eggs and broilers. This chapter describes the range of tools adopted to improve the microbiological safety of imported food. The proliferation of private standards and related certification raises important questions regarding their roles in ensuring the microbiological safety of imported food. International organizations play important leadership roles in efforts to improve the microbiological safety of foods. Among these organizations, the Codex Alimentarius Commission (Codex) is particularly important. The Codex standard-setting process and the standards themselves play an influential role in microbiological food safety for imported foods on several levels. In assessing the potential value of risk-based programs to improving the microbiological safety of imported foods, it is also necessary to consider how economics may challenge their value. Risk management is the identification and selection of appropriate food safety controls based on risk assessment and other factors that may be important to the risk management decision. The Hazard Analysis Critical Control Points (HACCP) approach is a prime example of the application of risk management principles in a systematic manner. Government is responsible for introducing or strengthening existing emergency response systems to respond to food bioterrorism by identifying necessary components of an emergency response program.

Instituting controls over pathogenic microorganisms in foods reaches back to the beginnings of civilization with the development of salting, fermenting, drying, and cooking processes. The conceptual ‘’food safety objective’’ equation was proposed to overcome some of the limitations of the Hazard Analysis Critical Control Points (HACCP) plan. A bright line focuses on the risk per serving and means that the high doses are always unacceptable, even though they may be very rare. There are several approaches to setting a quantitative appropriate level of protection (ALOP); one begins with the risk manager’s determination that the rate of illness is unacceptably high for a food or class of foods and can be reduced. The decision about an ALOP, of necessity, often focuses on the susceptible human populations. Consumption of foods cannot be easily restricted from certain groups, and the ALOP will need to reflect the more susceptible populations. Traditional assessments of the microbial hazards in foods were based on data for the characteristics of the pathogen, contamination levels, effects of process steps, ability of the pathogen to grow in the food, and the epidemiological record. A dose-response model then relates consumption to the probability of illness or other measure of public health. Thus, the risk assessment links the process parameter values (product pH, storage temperature) to public health (illnesses per serving).

The use of models to quantitatively describe the transmission of pathogens over the food-production chain is increasing in quantitative microbial risk assessment (QMRA). Such models may cover the whole “farm-to-fork” food pathway or only the part that is relevant to the problem. This chapter explains the use of a methodology developed for this purpose: the modular process risk model (MPRM). First, the MPRM methodology is outlined, offering guidelines to perform a food chain QMRA; then, some examples are given to illustrate these guidelines and to show how MPRM has been applied in practical situations. The major advantage of modeling is that it is an instrument that allows an easy comparison of a broad range of alternative scenarios, the main objective of risk assessment. Risk assessment deals with risk, which is a function of probability and severity of an event. As variability and uncertainty can both be represented by probability distributions and the difference between the two is not always obvious, they are easily mixed up. In MPRM food chain risk assessment, the selection of the “best” model depends on the statement of purpose, process knowledge, and data availability.

This chapter reviews the risk assessment activities that have been performed on Enterobacter sakazakii and focuses on the relations between the various concepts. Other food products besides powdered infant formula (PIF), such as breast milk, starch, and commercially sterilized liquid infant formula, have been described as potentially causing E. sakazakii infections. One of the potential control measures to reduce the risk of E. sakazakii is setting microbiological criteria (MC) for PIF. When a microbial criterion is in place, batches of product need to be sampled at the end of manufacturing, e.g., at packaging, prior to storage. The risk assessment model (RAModel) predicts the number of illnesses due to E. sakazakii per million infant-days, resulting from feeding reconstituted PIF, and compares this number with the numbers resulting from a baseline scenario. Risk communication from users to risk managers is also required. To be able to weigh the various options for control measures, it is imperative that the actual preparation and handling practices become known. Risk assessors can then use the RAModel to compare the options, based on better assumptions regarding temperature, holding times, and cooling rates. Consequently risk managers may be able to design and implement control measures that will effectively reduce the risk, but at the same time take into account all the relevant factors to make sure that the control measures will be feasible and acceptable and can be applied for those infants who are currently most at risk.

Microbiological criteria provide the food industry and regulatory agencies with guidelines for control of food processing systems and are an underlying component of any critical control point that addresses a microbiological hazard in Hazard Analysis and Critical Control Points (HACCP) systems. Aerobic plate count (APC) or standard plate count (SPC) is commonly used to determine “total” numbers of microorganisms in a food product. Depending on the pathogen, low levels of the microorganism in the food product may or may not be of concern. Some microorganisms have such a low infective dose that their mere presence in a food presents a significant public health risk. For such microbes, the concern is not whether the pathogen is able to grow in the food but that the microorganism could survive for any length of time in the food. Food products frequently subject to contamination by harmful microorganisms, such as shellfish, may benefit from the application of microbiological criteria. Enterococci counts have few useful applications in microbiological criteria for food safety. Some microbiological criteria related to safety rely on tests for metabolites to indicate a potential hazard rather than direct tests for pathogenic or indicator microorganisms. The current acceptable limits are based on the criterion that there is an 80% probability that the plant is actually exceeding the target value if it exceeds the acceptable limit.

Quantitative microbial risk assessment (QMRA) for Legionella can provide a means of setting risk based limits on air concentration or risk based limits on the Legionella concentrations in water. Guinea pigs provide a reasonable animal model and dose-response data on which to base human risk projections due to similarities in the course of the disease in guinea pigs and humans and in vitro uptake and replication rates of Legionella in human and guinea pig alveolar macrophages. Most mouse and rat strains appear to be relatively resistant to Legionella infection due to less compliant alveolar macrophages. A two-zone box model was used to estimate the air concentrations in CFU/m3, with stochastic input distributions using Monte Carlo simulation to yield results as probability distributions. The respective exposure distributions were fed into the dose-response model using Monte Carlo simulation. The resulting estimated risk distributions were then compared to the reported rates for the outbreaks. The guinea pig infection data adequately predict the reported subclinical (seroprevalence) rates. For the outbreaks used for model evaluation, the reported rates of disease span an order of magnitude. Thus, the QMRA may not extrapolate to other Legionella species and strains.

The purpose of risk assessment is to help a manager better understand the risks (and opportunities) being faced and to evaluate the options available for their control. The chapter describes some antimicrobial resistance risk assessments that have been completed. The Food and Drug Administration (FDA) wanted to produce a model that was based on reliable data. The FDA deemed that the risk assessment had quantitatively demonstrated that the resistance in question presented a risk to human health, and subsequently moved to withdraw use of the antimicrobial. The major strengths of the model were its mathematical simplicity, its reliance on mostly federally collected data, its very limited set of assumptions, and the ease with which it could be updated as new data became available. The data problems underlined the difficulties of evaluating even relatively simple model parameters for antimicrobial risk assessments. The FDA’s quantitative risk analysis model was successful in that it provided robust information from which the FDA was able to make an important decision. A key to the successful involvement of risk assessment in decision making is the neutrality of the risk assessor. There is, unfortunately, a growing trend in published antimicrobial resistance risk assessments that are clearly selective about their sources or manipulate a risk assessment model to produce the desired answer. The FDA guidance takes an essentially qualitative approach to evaluating drugs for approval and continued use.