The generally accepted paradigm for microbial, as well as chemical food contaminates, includes separating the assessment into four components: hazard identification, exposure assessment, dose-response assessment, and risk characterization. This process allows organization, characterization, definition of uncertainties, and identification of data gaps.
The questions below are not a prescriptive list but rather a list of possible questions that risk managers and risk assessors might discuss to clarify the scope of the risk assessment and to promote understanding of the results. This list of questions is intended to assist the risk manager and assessor in their deliberations and therefore no implications should be drawn regarding the rank order or the chronology in which the questions are presented. Finally, many of the questions might be posed by the risk assessment team leader to the risk assessment (RA) team.
Before Commissioning the Risk Assessment
Once the risk management (RM) team has provided the RA team with background information, and discussed the problem at hand, the questions to ask the risk assessment team prior to beginning the risk assessment include the following.
1a. Do you have any questions about the risk management problem or the subsequent risk assessment question posed by the risk management team?
1b. What type of risk assessment is needed? Would a threshold or safety assessment answer the risk assessment question posed by the risk managers?
1c. In consideration of the risk management problem, are there other questions that are important for the risk assessors to answer?
1d. Are there substitute risks that need to be evaluated because they will be impacted by any potential management decisions?
1e. Are any additional resources needed to adequately to address the problem?
1f. How should the risk management team communicate with the risk assessment during the project?
1g. How will the CFSAN risk assessment team work with our risk assessment partners (federal, academic, and industry), if appropriate?
During the Risk Assessment
General Data Questions
2a. Were studies that showed negative results or a lack of effect for the compound or food used in the assessment? If not, explain the reason for excluding those data.
2b. Were any data considered as potentially relevant to the assessment but not integrated into the calculations and conclusions? If so, why were those data excluded?
2c. Were qualitative data used? If so, how was it used it in the analysis?
2d. Were data collected outside of the United States used in the model? If so, does the United States import the relevant commodity or is there some other reason those data are relevant to the United States?
2e. Are the data representative of the subpopulation of interest? If not, how were the data weighted? Was the organization that collected the data consulted on proper weighting procedures?
2f. Were Bayesian approaches used in any aspects of the data computations? If so, how were the prior distributions obtained and weighted?
2g. What methods and assumptions were used to estimate the shape, spread, and location (e.g., mean or median) of the exposure or risk distributions? What are the most sensitive assumptions or data elements for the percentiles relevant to the risk management questions?
2h. How is the risk assessment team ensuring that all confidences (legal, proprietary) are being honored?
2i. Have any specific risk management options been eliminated because of the types of use of data or assumptions in the assessment?
Models
3a. What criteria were used to select the statistical model(s)? What models or types of distributions did you not use and why?
3b. Is there anything that can happen in the model that cannot happen in reality or the converse? (For example, unaccounted correlation between body weight and consumption rates or age that cause increased incidence at the tails.)
3c. Is the model flexible enough to be modified if new data become available?
Assumptions
4a. Which assumptions are the most influential on the results (that affect decisions)?
4b. Is there any specific modeling assumption(s) that the RM team should review?
4c. Which assumptions are science-based as opposed to policy-based?
Hazard
5a. Was the severity of the adverse effect and size of the exposed population adequately and properly described?
5b. Is the adverse effect reversible or irreversible?
5c. Is the hazard an acute one or chronic in nature?
5d. Does the adverse effect result from prenatal or early postnatal exposure of is it an effect seen latter in life?
5e. Is the primary evidence for the adverse effect based soley on laboratory animal experiments? Is there some correspondence for such an effect in humans?
5f. Are the adverse effects noted with only high levels of exposure or do they occur at levels more consistent with normal human exposure?
5g. Is there a mechanistic explanation for the adverse effects?
5h. Are there toxicokinetic/dynamic similarities/dissimilarities between laboratory animal models and humans?
5i. Does the adverse effect occur normally in the at risk population and is it a frequent or infrequent effect?
What other risk factors are normally associated with the adverse effect?
5j. If human studies are being used is the effect one considered to be a frank clinical one or one that is more subtle and subclinical in nature?
5k. Is there a spectrum or continuum of adverse effects?
Exposure
6a. If relevant, how were food groupings chosen?
6b. Are the contamination data from studies that use random sampling? If not, what adjustments were used to avoid bias?
6c. Were the food consumption data adjusted for geographical, socioeconomic and other factors?
6d. Does the risk assessment include food consumption data from more than one source? If so, what is the rationale for using data from more than one source or combining data from different data sets?
6e. Are there other, important sources of exposure to this pathogen that should be considered?
6f. Were outbreak data and/or surveillance used in the assessment?
6g. Is the exposure assessment derived based on food consumption and residue surveys or are they based on studies of biomarkers of exposure in human studies?
6h. Will the estimates of exposure be done in deterministically or probabilistically?
6i. Are the exposure data adequate to describe a distribution or will they have to be "bootstrapped" into a simulation?
Dose-response
7a. Was lab animal data used as the basis for the dose-response presentation? If so, how was the animal-to-human extrapolation addressed?
7b. Was human data used for the dose-response assessment? How many studies were used? If more than one was a 'meta' analysis performed or was the data pooled?
7c. Were the human studies clinical studies or environmental epidemiology studies? What type of study was used -- case-control, retrospective, prospective?
7d. Is the dose route of administration used in the dose-response assessment similar to the normal route of human exposure?
7e. Is the dose-response assessment within the observable range or does it involve low dose extrapolation?
7f. Is the dose range narrow or broad in scope?
Completion of Draft Risk Assessment
8a. What is (are) the answer(s) to the question(s) the risk management team asked?
8b. Was a sensitivity test conducted? Was it broad enough so that all major assumptions and data uncertainties were tested to evaluate how alternate assumptions or data would affect the results?
8c. Are sufficient data and information provided in the report so that another risk assessor could replicate the results?
8d. Which parts of the risk assessment will be most controversial?
8e. Have the major sources of uncertainty been identified? Is it possible to reduce the uncertainty?
8f. Explain which subpopulations are at the highest risk. How many are in each category and why they are in those categories (i.e., high levels of exposure or high sensitivity)?
8g. Does the risk assessment report include an explanation of the difference between uncertainty and variability?
8h. What is the impact of this risk (or risks) relative to other similar risks?
8i. Explain why the risk(s) were characterized in a particular way, e.g., risk per year, per lifetime, per eating occasion.
8j. What is the relationship in this risk assessment of morbidity to mortality? How many in each category and how severe is the morbidity?
Below is a summary of specific risk communication activities conducted within a risk analysis framework. The need for these activities was mentioned in other sections of this document and the focus of the discussion below is on specific risk communication activities.
Risk Communication Activities
During the planning step for the risk analysis project, the risk communication team will be formed by the appointed risk communication team leader(s) in consultation with the Risk Analysis Coordinator and Program Office/Staff directors.
Activity. Identify audiences for risk communication, such as the general public, scientists, the media, consumer and industry representatives, public health professionals, and regulators. Audiences may also include consumers, especially those consumers at-risk for foodborne illness, such as the elderly, pregnant women, young children, and people with weakened immune systems.
The audiences' needs for risk communication are as varied as they are. Careful analysis of the awareness and knowledge of the issues for each audience as well as the best method for reaching them is critical in preparing risk communications messages, materials, and determining the appropriate channels of communications.
Responsibility. The Risk Communications Team and the EOS Advisory Committee representative in consultation with the CFSAN Director and Deputy Director, the Risk Analysis Coordinator, and the Risk Assessment, Risk Management, and Risk Communication Team Leaders. Informal meetings with stakeholders, including meetings with professional colleagues and/or "sister" food safety agency representatives, should be held as well at the discretion of those mentioned above.
Activity. Once audiences have been characterized, the next step is to determine the most appropriate communication strategies (messages, materials and channels) for communicating that include both two-way and one-way communication.
The development of risk communication messages for information and education materials should be based on the results of the risk assessment and formal and informal information exchanges with various audiences. Common "talking points" or "themes" should be developed about the risk analysis process. An open exchange of information should insure that the broader risk communication messages are meeting the communication needs of the variety of audiences.
For example, messages to industry about the nature of the risks identified by a risk assessment and the methods of managing those risks may have a different emphasis than the message to consumers about safe handling practices necessary to avoid a specific risk.
Channels for communication may include:
- Public meetings
- Formal and informal working meetings
- Web pages
- Press releases
- Other CFSAN communication channels such as the Constituent Update
| Activity or Type of Material | Responsible Party |
|---|---|
| Identify talking points | Risk communication team |
| Research consumer advice, if any, based on the risk assessment | Consumer studies branch in consultation with the FSS education staff |
| Develop consumer advice, based on research | Consumer studies branch/FSS Education staff |
| Develop educational materials surrounding consumer advice | FSS education staff in consultation with Consumer studies branch |
| Prepare press release | Press Office in consultation with Risk Communications Team |
| Prepare Qs and As, if needed | Press Office in consultation with Risk Communications Team |
| Type of Material | Responsible Party |
|---|---|
| Risk assessment materials, including interpretive summary | Risk Assessment Office/Team |
| Risk Management Materials | Risk Management Team/Program Offices |
| News media information materials | Press office |
| Stakeholder information materials | OCO in consultation with Risk Communications Team leader, FSS staff director, and OCD/EOS |
| Consumer Educational materials | FSS Education staff in consultation with OCO |
Evaluate Success of Risk Communications Throughout and at the End of the Risk Analysis Process
Activity. Determine formal and informal means of evaluating effectiveness of risk communications with the various audiences during and at the end of the risk analysis process. The evaluations should consider whether all appropriate target audiences were identified and the extent of the target audiences' awareness, knowledge, and if appropriate, behavior change as a result of risk communication.
Responsibility. The Consumer Studies Branch, Division of Market Studies, OSAS should undertake consumer evaluation, in consultation with the FSS education staff through questions on the national consumer food safety survey.
The following are example questions to consider in compiling a list of candidate microbial risk assessments. The questions
address the nature of the risk in the context of identifying the risk (or a particular hazard) of concern to a defined
population group from consumption of defined product(s).
Is this a significant public health problem in terms of incidence?
Is this a significant public health problem in terms of severity of illness, morbidity, and mortality?
How geographically widespread is the hazard?
Is this hazard well recognized or a new or latent problem? Is the situation urgent or not?
Is the risk--worsening, stable, improving, or unknown?
What is the relative importance of this hazard in terms of the intrinsic properties (i.e., virulence, pathogenicity, host specificity)?
Is there a significant potential for human exposure to the hazard due to either high consumption or high contamination levels of specific commodities?
Are some people more susceptible to infection, such as the elderly, infants, and immunocompromised persons?
Does the public perceive this as a health risk?
Are there regulations, programs, or control measures that may be used to control the hazard?
Is there a need for a risk assessment because of a problem in trade, or economic, regulatory, or social concern?
The evaluation of the technical feasibility of conducting a risk assessment must consider the availability of data, the uncertainty created by lack of data, or the impact of the lack of specific data on the risk estimates and on the feasibility for conducting the assessment. Since the data are never perfect and rarely collected specifically for risk assessment, many types of data must be considered for inclusion into risk assessment. The criteria may depend on the risk management and/or risk assessment questions that are being addressed. The following are examples of criteria to include in an evaluation of data for the purpose of determining the feasibility of conducting a risk assessment.
Data availability. Are the data available and relevant to conduct a risk assessment?
Specific information to obtain include:
Evaluation of data
Other.