A quantitative risk assessment for campylobacters in broilers: work in progress

https://doi.org/10.1016/S0964-8305(02)00082-3Get rights and content

Abstract

Quantitative risk assessments estimate the probability of unwanted events occurring and stochastic modelling can incorporate real-life uncertainty and variability into these estimates. There is now a focus on whether these techniques can be applied successfully to the risks associated with food-borne microbiological hazards. With microbiological food-risk assessments, in order to assess the risk to human health, it is not only necessary to estimate the probability of the organisms being present at each stage of the food production pathway, but also to estimate the burden of organisms present at each stage. We are currently undertaking a risk assessment of the risks to human health consequent upon the presence of campylobacters in on-farm poultry. This paper will examine the initial model framework and the methodological issues arising from the complexity of the risk assessment pathway.

Introduction

Campylobacters are the commonest cause of human acute bacterial enteritis within the UK between 1988 and 1998, the number of reported cases in England and Wales had risen from 28,000 to 58,000 (Fig. 1). Discussions with the Communicable Disease Surveillance Centre (CDSC) have indicated that this is most probably a gross under estimate because many cases go unreported. Infection with campylobacters results in high human and economic costs and has been linked in epidemiological studies with the consumption of poultry meat (Harris et al., 1986). However, the epidemiology of campylobacter infection is complex and other foods have been linked to campylobacter infection such as un-pasteurised milk (Robinson and Jones, 1981), pork (Matttila and Frost, 1988), beef (Stern and Kotula, 1982) and even tuna salad (Roels et al., 1998). Production, and other stages of the supply chain are of significance in relation to campylobacter contamination in poultry. Therefore, an investigatory process involving all stages of the supply chain is required to identify critical control points. To this aim, a quantitative risk assessment is being undertaken to assess the risks of human infection with campylobacters from the consumption of broilers/broiler products.

Risk assessment represents an evaluation of the likelihood and severity of a known or potential hazard (Hathaway, 1997). When constructed in a quantitative manner the outcome is a mathematical statement, which describes the chance of an adverse outcome from exposure to a risk at some defined level (Covello and Merkhofer, 1993).

Quantitative risk assessment (QRA) has been used for several years in areas such as finance, engineering, and the nuclear and chemical industries. In contrast, its use in the veterinary sphere, and in particular the area of food safety, is relatively recent. Buchanan and Whiting (1996) have postulated two reasons for the limited number of food related risk assessments. Firstly, lack of knowledge concerning dose–response relationships for many micro-organisms and secondly, difficulties in estimating the numbers of organisms ingested by humans.

Although such problems are widely recognised, risk managers and policy makers are now beginning to appreciate the benefits of undertaking good quality, transparent risk assessments for food related issues. Moreover, the Sanitary and Phyto-Sanitary (SPS) Agreement of the World Trade Organisation, which allows countries to give priority to food safety over international trade only if a scientific basis for this priority can be demonstrated by means of defendable assessments (Wooldridge et al., 1996). This means that use of this techniques is essential to ensure safe food at both national and international levels. Consequently, QRA models for several micro-organisms in a variety of food products have appeared in the literature over the past 2 or 3 years. For example, Whiting and Buchanan (1997) have considered Salmonella enteritidis in pasteurised liquid eggs, Cassin et al. (1998) have modelled Escherichia coli O157:H7 in ground beef hamburgers and Nauta and Heuvelink (1998) have presented a model framework for E. coli O157:H7 in beef and beef products.

This paper presents an initial model framework for a campylobacter risk assessment and discusses the methodological and data collection issues associated with each stage of the general microbial risk assessment process. Some example results are presented and the future direction of the work is outlined.

Section snippets

Initial model framework

In line with other QRA models for microbial risks (see for example Cassin et al., 1998) our model will be stochastic in nature and as such will account for uncertainty and variability by appropriate probability or uncertainty distributions via Monte-Carlo (stochastic) simulation.

All stages in the farm-to-fork poultry supply chain will be addressed and at each stage, both the probability that any bird is infected or food product contaminated and the probable microbial dose will be estimated.

Data for risk assessment models: possible problems and solutions

For any QRA, as each stage of the assessment is considered, available data must be optimised and validated, however in many situations, data resources may be limited or even at first sight non-existent. This problem has been highlighted previously for the campylobacter risk assessment, and microbial risk assessments in general, particularly in relation to requirements such as quantitative microbial data and human dose–response data.

There are various approaches that can be used to deal with the

Example results

We now consider some example results which have been generated for Module 1 in Fig. 2, that is Rearing and Transport. In particular, we have obtained a preliminary estimate for the probability that any random bird within the GB broiler flock is campylobacter-positive at the time of depopulation, i.e. the time when the house is emptied and birds are taken to the slaughter house, based on the results of one epidemiological study which was undertaken to investigate campylobacter prevalence within

Summary

This risk assessment model will evaluate the risks to human health from campylobacters as a result of consumption of broilers/broiler products. The full poultry production supply chain will be considered within this model and for each stage of production, data will be collected and formulated within a Monte-Carlo simulation using appropriate mathematical and probabilistic methods.

An initial model framework has been developed and the various data requirements have been considered. Such

Acknowledgements

The authors would like acknowledge Rupert Low for assistance with the preparation of this manuscript. This work is funded by the Veterinary Laboratories Agency.

References (18)

  • S.C Hathaway

    Development of food safety risk assessment guidelines for foods of animal origin in International trade

    Journal of Food Protection

    (1997)
  • R.C Whiting et al.

    Development of a quantitative risk assessment model for Salmonella enteritidis in pasteurised liquid eggs

    International Journal of Food Microbiology

    (1997)
  • Buchanan, R.L., Whiting, R.C., 1996. Risk assessment and predictive microbiology. Journal of Food Protection (Suppl.),...
  • Cassin, M.H., Lammerding, A.M., Todd, E.C.D., Ross, W., McColl, R.S., 1998. Quantitative risk assessment for...
  • P.J Comer et al.

    Risk assessment of BSE infectivity in the environment from rendering of over thirty month scheme cattle

    Journal of Risk Research

    (1998)
  • V.T Covello et al.

    Risk assessment methods: approaches for assessing health and environmental risks

    (1993)
  • Evans, S.J., 1997. Epidemiological studies of Salmonella and Campylobacter in poultry. Ph.D. Thesis, London School of...
  • N.V Harris et al.

    The role of poultry and meats in the aetiology of Campylobacter jejuni/coli enteritis

    American Journal of Public Health

    (1986)
  • Horst, H.S., 1997. Risk and economic consequences of contagious animal disease introduction. Ph.D. Thesis, Wageningen...
There are more references available in the full text version of this article.

Cited by (9)

  • A review of quantitative microbial risk assessment and consumer process models for Campylobacter in broiler chickens

    2016, Microbial Risk Analysis
    Citation Excerpt :

    The transport portion of the model was designed to be complementary to the on-farm model, and organized colonized broilers, described by their location in the growing house, in three-dimensional space on transport vehicles to characterize limited Campylobacter contamination during transport to slaughter. A broad overview of the full QMRA was published by Hartnett et al. (2002). The industrial processing model was non-mechanistic; each stage of processing was described by a non-parametric, second-order distribution, which allowed for the separate and distinct calculation of variability and uncertainty.

  • Consumer acceptability of interventions to reduce Campylobacter in the poultry food chain

    2014, Food Control
    Citation Excerpt :

    The infection is asymptomatic and once in a flock Campylobacter is rapidly transmitted by the faecal-oral route throughout the broilers (Wassenaar, 2011). The bacteria can then survive during poultry processing (stages comprise of transport, slaughter, processing and preparation) through to human consumption (Hartnett, Kelly, Gettinby & Wooldridge, 2002) and causing subsequent illness. However at each of the process stages interventions either are in place or can be implemented to control Campylobacter.

  • Modified atmosphere packaging and the safety of poultry meat

    2005, Food Safety Control in the Poultry Industry
View all citing articles on Scopus
View full text