IRAC-MRAF Workgroup

Food and Water Microbiological Risk Assessment Framework Study

A project of the Interagency Risk Assessment Consortium (IRAC), 
Microbial Risk Assessment Framework Work Group (MRAF)

Executive Summary
August 21, 2002

Background:
The goal of the project was to enhance the conduct of quantitative and qualitative microbial risk assessments by identifying critical elements and tools necessary or desirable for the conduct of microbial risk analysis (MRA). The project utilized the draft Environmental Protection Agency- International Life Sciences Institute (EPA-ILSI) Risk Analysis framework and the Codex Principles and Guidelines for the Conduct of Microbiological Risk Assessment. Conformity with the two risk assessment frameworks was evaluated from the perspective of RAC member agency experiences in conducting microbiological risk assessments.

The EPA, Office of Water provided funding through an interagency agreement to the Center for Food Safety and Applied Nutrition for the RAC to review the frameworks for the conduct of microbiological risk assessment of food and water. These frameworks describe elements that should be considered in the conduct of microbial risk assessments. EPA initiated this project with the RAC to assist them with the development of microbial risk assessment guidelines similar to EPA's chemical and environmental risk assessment guidance documents.

Risk assessors who have worked on US microbiological risk assessment efforts have valuable practical experience and information to share. To more fully document this knowledge, funding was used to hire a contractor (Dr. Charles Yoe) to interview seven federal risk assessors who worked on five different risk assessments. The risk assessments included:

• Salmonella Enteritidis Risk Assessment Shell Eggs and Egg Products, Prepared for the United States Department of Agriculture's Food Safety Inspection Service (USDA/FSIS) by the SalmonellaEnteritidis Risk Assessment Team, June 1998.
• Risk Assessment on the Human Health Impact of Fluoroquinolone Resistant Campylobacter Associated with the Consumption of Chicken, CVM/FDA U.S. Department of Health and Human Services, October 2000.
• Draft Risk Assessment on the Public Health Impact of Vibrio parahaemolyticus in Raw Molluscan Shellfish by CFSAN/FDA U.S. Department of Health and Human Services, December 2000.
• Draft Assessment of the Relative Risk to Public Health from Foodborne Listeria monocytogenesAmong Selected Categories of Ready-to-Eat Foods, FDA/CFSAN, USDA/FSIS, Centers for Disease Control and Prevention, January 2001.
• Draft Risk Assessment of the Public Health Impact of Escherichia coli O157:H7 in Ground Beef, Prepared for FSIS by the Escherichia coli O157:H7 Risk Assessment Team, September 2001.

The full report includes, general unifying themes from the interviews, responses to 15 questions posed by the RAC, summaries of the individual interviews and a database of MRAF elements. The Excel database lists elements from the ILSI framework and indicates whether or not the interviewee thought the element was included in a specific risk assessment. If an element was not included, the reason was given. This executive summary highlights some of the most interesting points from the report's general unifying themes. The risk assessors interviewed (in early 2002) stressed that each risk assessment is unique, so their comments were generally about aspects of frameworks that would benefit all risk assessments. They consistently agreed that a framework or set of frameworks should cover the process of conducting MRA, the possible elements in the model and useful suggestions for the documentation, reporting and communication of the MRA.

Report summary:

Framework flexibility. 
The steps and the elements of a framework should be more suggestive than prescriptive. Depending on the scope and purpose of the risk assessment the actual model may use some elements and not others.

Framework for risk assessment process should:

• Include or refer to a process for forming the management questions and other background information that is needed to determine the scope of the risk assessment.
• Recommend that concrete and specific questions in a written format should be developed iteratively and refined over progression of the project as other aspects of a risk assessment.
• Recommend establishment of a realistic timeline, data issues (sources, gaps, funded projects to collect missing data), and the appropriate level of resources for the risk assessment.
• Recommend the inclusion of provisions in project timelines in case considerable revisions in schedule and budget are needed to accommodate significant changes in the scope of an assessment.
• Recommend clear articulation of the authorities, responsibilities, expectations, coordination and interaction among team members (risk managers and assessors)
• Recommend a review process that is unbiased, thorough, competent, and constructive. Critique should be focused on the model used and the questions answered rather than on the model that "should" have been used and the questions that should have been asked.
• Recommend a round of initial review at the start of the process (during what the ILSI model calls the problem formulation step or following what Codex calls the risk profile).
• Stress that risk assessment questions must, initially at least, remain unconstrained by data availability. (Risk assessments are also useful to help guide future research efforts.)
• Underscore the essential role of stakeholders in the team approach to risk assessment.
• Articulate for the scientific community how data can best be collected to aid MRA and what data format is most useful for MRA.
• In the absence of data, consider methods of expert elicitation.

Framework for risk assessment modeling should:

• Recommend developing a conceptual risk model early in the assessment, long before the quantitative risk assessment is initiated.
• Catalogue model options that could meet the needs of risk assessment.
• Recommend adding a formal problem formulation phase to identify the risk assessment elements and model components and nodes that are practical and important.
• Recommend use of a mechanism for tracking assumptions and active searching for implicit assumptions.
• Articulate a mechanism for assuring the adequacy of a model's transparency.
• Recommend the development of complete documentation of each model and software developed to aid institutional memory.
• Identify potential software tools.

Framework for risk assessment documentation and communication should:

• Acknowledge that one document cannot serve all the potential audiences. (Multiple documents may be needed.)
• Remind risk assessors that definitions for sensitivity analysis, uncertainty, and variability should be fully articulated.
• Help the risk assessment and risk management teams build a common vocabulary (between epidemiologists, modelers, toxicologists, microbiologists, economists, physicians, and so on.)
• Recommend consideration of including risk communication professionals before releasing simulation results to stakeholders.

Conclusion: 
Application of the CODEX framework and the EPA-ILSI framework by risk assessors was evaluated after extensive experience conducting actual microbial risk assessments. Both frameworks demonstrated accurate foresight with respect to the risk assessment process. They are flexible, clearly written, and sufficiently descriptive to provide users with a list of considerations that can be tailored to fit assessor's needs. Consistent with the role of the frameworks, none of the five risk assessments rigidly adhered to the guidance of either framework. The Codex framework goes into greater detail in discussing food processing, while the ILSI framework is more detailed in describing human health factors. Because of the dynamic nature of microbial risk assessments, future efforts should not target revisions to these frameworks, but should make their utilization more accessible through worked examples, identification of useful tools, methods and approaches that can help assessors more easily perform risk assessments. 

IRAC-MRAF members
Bartholomew, Mary (FDA/CVM)
Coleman, Peg (USDA/FSIS)
Dennis, Sherri (FDA/CFSAN)
Ichida, Audrey (AAAS Fellow, JIFSAN)
Long, Wes (FDA/CFSAN)
Maczka, Carol (USDA/FSIS)
Miliotis, Marianna (FDA/CFSAN)
Roberts, Tanya (USDA/ERS)
Schaub, Steve (EPA/OW)
Tamplin, Mark (USDA/ARS
Turturro, Angelo (FDA/NCTR)
Walderhaug, Mark (FDA/CFSAN) 
Whiting, Richard (FDA/CFSAN)

Acronyms
AAAS: American Association for the Advancement of Science
CFSAN: FDA's Center for Food Safety and Applied Nutrition
CVM: Center for Veterinary Medicine
EPA: US Environmental Protection Agency
ERS: Economic Research Service
FDA: US Food and Drug Administration
FSIS: Food Safety and Inspection Service
JIFSAN: Joint Institute for Food Safety and Applied Nutrition
ILSI: International Life Sciences Institute
NCTR: National Center for Toxicological Research
RAC: Risk Assessment Consortium
USDA: US Department of Agriculture
OW: Office of Water

ILSI framework can be found at:
http://www.ilsi.org/publications/pubslist.cfm?pubentityid=%2D1&publicationid=20
The Codex Framework can be found at: http://www.who.int/fsf/mbriskassess/Reference/mra.pdf