Managerial Framework Revisited

The motivation for change, as described above, appears to focus on stakeholders who want to understand risk and to participate in the process of risk assessment. The consequence of the expanded concept of RC is to change the structure of risk analysis to include the stakeholders from the beginning.

The figure developed by the U.S. President's Commission on Risk Management (1996; Fig. 3.2) illustrates a framework for risk management that reflects this new concept. In this figure, the stakeholders are in the center and the risk management activities circle around them. This is meant to imply that stake-

Figure 3.2. Framework for risk management (Presidential/Congressional Commission on Risk Assessment and Risk Management, 1996; www.riskworld.coml'Nreportsj).

holders are involved in each phase of the risk analysis. Such a concept places a tremendous burden on stakeholders to understand the process and the procedures of a risk assessment. The diagram depicts stakeholder input at the first stage, "problem context." In recent years, to implement this objective, U.S. agencies have commonly convened public meetings to introduce risk assessment projects and solicit data at the start of major risk assessments. The diagram also depicts additional processes ("risks," "options," "decisions," "actions," "evaluation"). Integrating stakeholders into these activities, particularly the "risks" and "options" activities, presents a challenge because of the highly technical nature of these areas. Some groups of stakeholders may lack the expertise to fully participate and the financial ability to hire risk analysis experts to provide input to these processes. The burden on U.S. regulators is to promote a fair and balanced process, even for those stakeholders who may be unable to afford to hire expert consultants to represent their interests. A consequence of this strategy is another point of tension or potential conflict and would impose a burden on the regulators to provide a fully independent and transparent risk assessment amenable to input.

Expansion of Risk Analysis

The expansion of the concept of RC to include deliberations, enhancing understanding, and implementing practical solutions, extends to risk analysis itself. The consequence of expanding RC is that risk analysis expands to include non-risk assessment procedures that would lead to an understanding of hazards and to practical solutions of managing them. The requirements for performing a complete risk assessment are sometimes not met, but yet solutions to problems are needed. Constraints of time, funding, expertise, and data available to support risk assessment modeling may not permit a full quantitative risk assessment. The data gaps could be so extensive that the credibility of a full quantitative risk assessment could be questioned.

There are other common approaches to evaluating hazards that, although not full risk assessments, use the tools of risk assessment. These procedures, which we refer to as "quasi-risk assessments," offer mangers practical solutions to problems and thus can fit under the expanded concept of risk analysis. Some of these "quasi-risk assessment" procedures are introduced below.

"Qualitative Risk Assessment"

The term "qualitative risk assessment" describes a process of ranking or categorizing hazards and risks that stops short of estimating risks and attendant uncertainties. This term is inconsistent with the definition of risk described herein. A "qualitative" risk assessment is incomplete because it lacks the calculations of likelihood of adverse consequences. Perhaps a new term such as "qualitative risk accounting" might be more useful to describe a legitimate process of ranking or categorizing risks that stops short of estimating risk with attendant uncertainty but still has value in problem solving and decisionmaking. An HI would be performed, and parts of an EA and DRA would be addressed, without providing quantitative results (USDA, 1998). A complete RC as described by the NRC (1996) would not be possible for a "qualitative risk accounting."

The possible benefit of a quantitative risk assessment is that the sensitive variables that most strongly influence the risk could be identified and used for priority setting for a research agenda that might fill essential data gaps. Although a qualitative risk accounting would not include such a sensitivity analysis, a systematic discussion and ranking of hazards and risks can provide useful understanding and reasonable "guesstimates" of variables that might be influential for further study.

"Safety Assessment" and "Worst-Case Scenarios"

Another procedure, which has been termed "safety assessment" (Wilson, 1999), involves calculating "safe levels" of hazards for all in the population. Often the focus is on simplifying default assumptions typified by application of a series of 10-fold "safety factors," such as for inter- and intraspecies extrapolations and high- to low-dose extrapolations, rather than explicitly estimating "risk with attendant uncertainty." The safety factor approach has become a standard practice of government regulators in the U.S. and abroad for managing some chemical hazards. On the basis of these derived "safe levels," tolerances for chemical levels in food are established.

Another approach is based on establishing the "worst-case" level for a hazard for an identified amount of product, using information from surveys or epidemiological studies. Standards are established to ensure that the hazard associated with the worst case would be eliminated from the product with high probability. This approach assumes some knowledge of the "lowest" dose that would result in adverse consequences if ingested. Both of these approaches, although serving an immediate regulatory need, lack the quantification of risk that would be part of a risk assessment.

Other Modifications to the NRC Risk Assessment Framework

Since 1983, the NRC paradigm of risk assessment has been adapted by many risk scientists, including Covello and Merkhofer (1993); National Research Council (1993, 1994, 1996); International Life Sciences Institute (ILSI, 1996); Presidential/Congressional Commission on Risk Management (1996); FAO/ WHO Consultation on Risk Management (1996); Kaplan (1997); Marks et al. (1998); Codex Committee on Food Hygiene (1998); FDA (1999); Rand and Zeeman (1998); NACMCF (1998); ICMSF (1998); and McNab (1998). Some suggest that different risk assessment frameworks might be helpful for different types of hazards (chemical, biological, and physical). Some may view the differences between the various proposed frameworks as merely semantic.

The NRC framework (1983) is applicable for quantitative risk assessment. However, as discussed above, the decision not to conduct a quantitative risk assessment, but to rely on a "quasi-risk assessment" may be triggered by the lack of relevant data for either exposure or dose-response assessments. One modification, suggested by the Codex Committee for Food Hygiene (1998), is the replacement of "dose-response assessment" with "hazard characterization," a more general term that emphasizes the value of qualitative approaches when dose-response data for the pathogen and food of interest do not exist for the populations at risk. In a similar vein, Covello and Merkhofer (1993) suggest replacement of "dose-response assessment" with "consequence assessment."

More substantive modifications have been proposed that reflect the need for greater interaction between research and policy areas. Covello and Merkhofer (1993) proposed considering "hazard identification" as a preliminary step before conducting a risk assessment rather than the first step in conducting a risk assessment. To enhance understanding for the stakeholders and to identify scenarios, the hazards must be identified before the calculations of risk. This approach is consistent with our definition of risk and the expanded concept of risk characterization discussed above.

Another deviation from the 1983 NRC four-element risk assessment framework actually predates that framework. An additional phase of modeling exposure termed "Release Assessment" originated in the mid-1970s under the Nuclear Regulatory Commission to simulate unintended releases of radiation by the nuclear power industry (Cohrssen and Covello, 1989). A body of work has accumulated in probabilistic risk assessment describing accidental releases of many other types of hazards. "Release assessment" could be appropriate for chemical spills in animal feeds or bacterial contamination in foods.

Selection of the managerial framework to structure the risk assessment process does not validate or invalidate the process. Of greater importance is that the analysis is based on the best available science and that key principles and guidelines such as those listed in Tables 3.1 and 3.2 are followed. The Codex Committee on Food Hygiene (1998) chose not to provide a methodological recipe for conducting a risk assessment but provided guidance that does not limit methodological approaches, thus encouraging use of the best available science.

0 0

Post a comment