The process for the development of guidelines on hazard characterization for microbiological hazards in food and water began at a workshop held in Bilthoven, the Netherlands, 13-17 June 2000. The objective of the Hazard Characterization Workshop was to develop guidelines on hazard characterization for microbiological hazards in food and water. In order to accomplish this, the workshop 1) reviewed the state of the art in hazard characterization and relevant scientific disciplines (e.g. epidemiology, biomedical research, mathematical modelling); 2) categorized the principles and methods of hazard characterization; 3) provided guidance on the type of data needed and the means of assessing the adequacy of available data for developing dose-response relationships for specific pathogens; and 4) identified future research requirements to reduce uncertainty of dose response models and default assumptions for use in the short term.
Participants at the Bilthoven workshop compared and reviewed the approaches used in hazard characterizations for several pathogens (Salmonella spp., L. monocytogenes, enterohaemorrhagic Escherichia coli, Cryptosporidium parvum and Norwalk-like viruses). The outcomes of the workshop included a comparison of models that illustrated the benefits and weaknesses of different approaches. The workshop formulated general principles and guidelines for hazard characterization that were further discussed at this expert consultation.
The consultation discussed and reviewed the working principles and guidelines developed at Bilthoven with the intention of further developing the document in a wider arena of experts. In addition, practical lessons from the specific hazard characterization documents prepared by the expert drafting groups were included. The guidelines, in their current status, are a work in progress and will receive a variety of inputs in the medium term before completion.
The consultation welcomed the opportunity to participate in the continued development of these guidelines. General discussions indicated wide support for the purpose and scope of the work. Areas for further technical developments were specifically identified and comments on format, scope and structure are included in this summary.
General issues
The document should clearly and consistently state that the purpose and scope of a hazard characterization depends on the purpose and scope of the specific risk assessment that is required. This will generally be determined by risk managers in association with risk assessors and must facilitate achievement of identified risk management goals.
Hazard characterization of microbiological hazards will inevitably include consideration of biological characteristics of the microorganism, food matrix, and human host and their impacts on adverse health effects. This extends the understanding of hazard characterization as currently defined in the Codex Alimentarius: (The qualitative and/or quantitative evaluation of the nature of the adverse health effects associated with the hazard. For the purpose of microbiological risk assessment (MRA) the concerns relate to microorganisms and or their toxins. (CAC/GL- 30 1999)). Thus, hazard characterization for microbiological hazards in foods must be extended to cover all components of the pathogen, food, and host triangle.
Consideration of decisions on tolerable levels of risk are risk management functions and, therefore, not within the scope of the Guidelines Document.
Opinion was expressed that the working document is too closely linked to experience gained from hazard characterizations contained in relatively few pathogen / food commodity examples. Consideration of other approaches should be added.
Specific issues in the current document
The introduction of the document should clearly describe the way in which hazard characterization contributes to the process of risk assessment.
The introduction contains considerable repetition, which should be addressed in the next revision. Figure 1 should be presented in a narrative format.
In section 5.1.1.3, the orientation of the hazard characterization will be a consequence of the specific purpose of the risk assessment to which it contributes.
The representativeness and adequacy of data included in model inputs are major issues in MRA. Section 6 should be expanded to include further development of guidelines on these topics. Further to this work, the issue of qualitative risk characterisations remains controversial and will require further deliberation. Section 6 should also include strategies to acquire high-quality data.
Characterization of human adverse health effects (Section 7.1) should include reference to methods other than standard dose-response modelling. In particular the use of epidemiological methods (such as case-control, cohort or cross-sectional studies) should be emphasized in the guidelines. For instance, a comprehensive set of epidemiological studies could be used to elucidate the relative importance of various food types for a specific hazard. Establishing a dose-response function should be based on the consideration of an array of different mathematical models, as appropriate.
Information on the extent and severity of disease that is included in dose-response modelling (Section 7.1.2) should be appropriate to the purpose and scope of the particular risk assessment (e.g., differences in socio-economic and demographic characteristics of consumer populations in different countries).
The selection of models in Section 7.2.2 should not advocate building conservativeness into the model structure. This concept applies to treatment of inputs and outputs rather than the structure of the model of itself, which should be precise and unbiased.
The Bilthoven workshop rejected the idea of a threshold in infection for microbial hazards, as the weight of current evidence suggests that one ingested organism has some probability (however remote) of causing illness. Further debate on this issue may lead to the removal of any consideration of thresholds from the working document.
The expert consultation concluded that the working principles and guidelines developed at Bilthoven should be the subject of further review and development to incorporate the changes suggested above. The opinion of the Codex Committee on Food Hygiene should be solicited together with ongoing input for the Joint Expert Consultations on Microbiological Risk Assessment. In addition, comments will be solicited from the general public through publication on the WHO and FAO web sites. Development of the guidelines will be finalized at a second Joint Expert Consultation on Microbiological Risk Assessment in 2001.