Commission on Risk Assessment and Risk Management Risk Assessment and Risk Management in Regulatory Decision-Making DISCUSSION DRAFT fel \ 9 January 1996 COMMISSION ON RISK ASSESSMENT AND RISK MANAGEMENT Gilbert S. Omenn (Chairman), Dean. School of Public Health and Community Medicine, University of Washington, Seattle, WA Alan C. Kessler (Vice-Chairman), Partner, Buchanan Ingersoll, Philadelphia, PA Norman T. Anderson, Director of Research, American Lung Association of Maine, Augusta, ME Peter Y. Chiu, Kaiser Permanente, Department of Urgent Care, Milpitas, CA John Doull, Professor, Department of Pharmacology, Toxicology and Therapeutics, Kansas University Medical Center, Kansas City, KS Bernard Goldstein, Director, Environmental and Occupational Health Sciences Institute and Chairman, Department of Environmental and Community Medicine, Robert Wood Johnson School of Medicine, Piscataway, NJ Joshua Lederberg, President Emeritus, Rockefeller University, New York, NY Sheila McGuire, President, Iowa Health Research Institute, Boone, [A David Rall, Former Director, National Institute of Environmental Health Sciences, Washington, DC Virginia V. Weldon, Senior Vice President for Public Policy, Monsanto Company, St. Louis, MO Staff Gail Charnley, Executive Director Sharon Newsome, Associate Director Joanna Foellmer, Program Specialist and Designated Federal Official il Executive Summary 1.0 2.0 3.0 4.0 5.0 Table of Contents Introduction... 2... ee eee Uses and Limitations of Risk Assessment in Regulatory Decision-Making ..... . 2.1 Cancer Risk Assessment ....... 20.0.0... 0. eee eee eee eee 2.2. Noncancer Health Risk Assessment .................-20-0+0000055 2.3 Ecologic Risk Assessment .... 2... .. 2.000. ee eee 2.4 Sensitive Subpopulations Requiring Special Consideration .......... 2.5 Uncertain Risk Estimates .........0.. 20.00.0000. eee eee eee 2.6 Peer Review .. 0.0... 0.00 0c eee Risk Management and Regulatory Decision-Making ..................... 3.1 Bright Lines ..... 0.0... 0s 3.2 Communicating Risk .... 2.2.2... 000 ee 3.3. Comparative Risk Assessment ........ 0.0.0.0... eee eee eee 3.4 Economic Analysis ..... 20.2.2... 00.0 cee ee eee 3.5 Judicial Review ...... 0.0... 0. cee eee 3.6 Inter- and Intra-Agency Consistency ............-..0-0-020000- Framework for Risk Management ............... 0.000 e eee eee eee Recommendations for Specific Regulatory Agencies and Programs .......... 5.1 Environmental Protection Agency 5.1.1 Clean Air Act: Residual Risk and Maximum Available Control Technology [Norm/Alan/Bernie/Gail] 5.1.2 Superfund (should include ATSDR) [Gil/Gail] 5.1.3 Office of Water [Peter/Sharon] 5.1.4 Office of Pesticide Programs [John/Sharon] 5.2 Occupational Safety and Health Administration (should include NIOSH) {[Norm/John/Gail] 5.3. Food and Drug Administration [Dave/Gail] 5.4 Consumer Products Safety Commission [Alan/Sharon] 5.5 Department of Agriculture [Sheila/Virginia/Sharon] 5.6 Department of Energy [Bernie/Gil/Sharon] 5.7 Department of Defense [Josh/Gail] ii References... 0. ene ee eee R-1 Appendices A.l Mandate of the Commission ............... 0.0000 e eee eee Al-l A.2 Federal Agency Risk Assessment and Risk Management Practices ..... A2-4 Introduction... 2... ee A2-4 Survey of Practices... 0... ee es A2-7 Food and Drug Administration ............ 00... ...004. A2-7 Occupational Safety and Health Administration............ A2-11 Consumer Product Safety Commission.................. A2-17 Environmental Protection Agency ..................--- A2-20 Office of Pesticide Programs ................... A2-22 Office of Pollution Prevention and Toxics .......... A2-27 Office of Air and Radiation .................-4. A2-30 Office of Water 2... ee ee A2-37 Office of Solid Waste .................2 020005 A2-41 Office of Emergency and Remedial Response ........ A2-44 A.3. Comments on Science and Judgment in Risk Assessment .......... A3-47 A.4 Individuals Who Presented Testimony at Commission Meetings ..... A4-61 iv Executive Summary [to be written after issues and recommendations are agreed upon] tN te Oo eo NH WH 16 17 18 19 20 21 22 23 24 25 26 1.0 Introduction Since the 1983 National Research Council report, Risk Assessment in the Federal Government: Managing the Process (“the Red Book’), established risk assessment and risk management as separate activities (“two distinct elements’), scientists and policy-makers have debated the role of risk assessment in regulatory decision-making. While the authors of the Red Book did not intend risk assessment and risk management to be practiced in isolation from one another, the use of risk assessment as a tool in support of decision-making has had limited implementation. The report recognized the importance of communication between the risk assessor and the risk manager, but did not offer guidance to facilitate such interaction; as a result, the practice of risk assessment has evolved essentially in the absence of a risk management context. Reacting to this isolationist evolutionary tendency a decade later, the authors of the 1994 National Research Council report, Science and Judgment in Risk Assessment, concluded that science-policy judgments made in the course of risk assessment would be improved if they were more clearly informed by a regulatory agency’s priorities and goals in risk management. Protecting the integrity of risk assessment, along with building more productive linkages to make risk assessment more accurate and relevant to risk management, were both considered essential. As P.F. Deisler (1988) put it, “The ideal separation should not be taken to mean that the two activities be isolated from each other until the grand cataclysmic communication of risk characterization.” The National Research Council has described risk assessment as “the use of the factual base to define the health effects of exposure of individuals or populations to hazardous materials and situations” (NRC 1983), and as a process that “entails the evaluation of information on the PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 1-1 Ww & Ww oO Oo NH ON hazardous properties of substances, on the extent of human exposure to them, and on the characterization of the resulting risk” (NRC 1994). Risk assessment is a systematic approach to organizing scientific information about potentially hazardous situations. Risk assessments are based on logically compelling scientific information when it is available and on scientifically informed policy judgments when it is not. A review of cancer risk assessments by the National Research Council identified a minimum of fifty places in a risk assessment that could not be based on data and that required science-based assumptions and judgments (NRC 1994). Because there are extensive uncertainties and assumptions inherent in any risk assessment (NRC 1 een OstP 1995), estimates of health risk obrained by performing a risk assessment are not scientific estimates of actual risk. They are nding en? estimates of the actual risk that that could Peis Ferg He oS aptions: and se weet be adrenal Hye pact Sac sib! a 367 ‘ decisions about ne redu ro aaa eh few Se eting wD tre most cok oh [ apes ler Kea prima bese muluih Cncse Unptctneas ot laaaed The National Research Council has defined risk management as “the process of weighing policy alternatives and selecting the most appropriate regulatory action, integrating the results of risk assessment with engineering data and with social, economic, and political concerns to reach a decision” (NRC 1983). To some extent, risk assessment has evolved in the absence of a risk management context, due to the distinction that wee bee between these activities (“two distinct elements”) by t the National Research Council in 1983. There was a fear that the many assumptions relied upon in risk assessment would be corrupted by the politics of risk management. The result has been a tendency to produce risk assessments that often have poorly served the goals of risk management. [The Seurt dy lemma eta fo M2 caraplrs of wkeUe\eas ki whim seek ptrey. { In practice, the results of a risk assessment are integrated with other information—such as political, social, economic, and engineering considerations—to arrive at decisions about the need and methods for risk reduction (NRC 1994). Performing sound risk assessments is important; however, the results of a risk assessment constitute only one of many considerations in a regulatory decision. Simply performing risk assessments and other PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 1-2 ios) Oo on A WN + 17 18 19 20 21 22 23 24 25 26 27 28 analyses such as cost/benefit analyses, and certifying that they were conducted, does not address critical challenges in assuring rational and cost-effective risk management under the fo $a, tata awncetes complex statutes designed/to protect an health and. the environment that regulatory agencies must satisfy. One problem with focussing on a risk assessment-based approach to risk management decisions about health protection in an environmental context is its lack of a public-health base. The public-health foundation of environmental health protection has been obscured by legalistic regulatory command-and-control approaches and by technically based, centralized decision-making processes that can be unrelated to the real causes of public health risk or to the problems faced by local communities. For example, the U.S. EPA now has far more lawyers than public-health professionals—at its inception in 1970, EPA had 650 U.S. Public { fae F - Health Service commissioned officers; it now has fewer than 200. In contrast, EPA now employs about ??? attorneys. This focus on legal and regulatory expertise has obscured the public health principles and goals that are the foundation of our environmental health laws. Another problem with risk assessment-based risk management is pervasive public distrust, which has led to increased politicization and conflict. Over the last 25 years, the United States has achieved a significantly cleaner environment and an increasingly healthy population. Life expectancy continues to increase and non-tobacco-related cancer incidence to decrease. Yet the American public becomes increasingly concerned about risk, believing our air, water, and food to be more contaminated with toxicants than ever. Public perceptions of health and environmental risk clearly differ from scientists’ and regulators’ perceptions, and can be attributed to a sensitivity to technical, social, and psychological qualities that are not well-modelled by technical risk assessments (Slovic 1993). The important role of public perception in risk assessment and risk management has become apparent, but is not yet well accounted for. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 1-3 tad - Oo Be NH Ww 10 1] 12 14 15 16 17 18 19 20 21 22 23 24 25 26 As a result of concern about meeting the critical challenges currently facing risk assessment- based risk management and regulatory decision-making, the Commission on Risk Assessment and Risk Management was convened in May 1994, in response to Section 112(0) of the 1990 Amendments to the Clean Air Act, to address the role that risk assessment and risk management play in regulatory decision-making. The ten members of the Commission were appointed by the president, by the majority and minority leaders of the House and Senate, and by the president of the National Academy of Sciences. The Commission has met 15 times since then, in Washington, DC and in several other cities across the United States, to hear testimony from a variety of individuals, organizations, and interests, on issues related to its mandate. ' Congress first decided to turn to a commission when, while drafting the 1990 Amendments, agreement could not be reached on the best way for the U.S. Environmental Protection Agency (EPA) to determine whether any risks to human health remained after Maximum Available Control Technology was implemented to reduce contaminant emissions to air from industrial facilities, and if so, what to do about them. There was a concern that after technological solutions to pollution control were in place, some risks to health might remain, but there was disagreement about the risk-assessment techniques and assumptions that should be used to estimate those risks, about the benchmarks that should be used to distinguish between negligible and unacceptable risks, and about the risk-management methods that should be used to mitigate them, should they exist. The Commission’s mandate was not restricted to air pollution, the EPA, or the particulars of the Clean Air Act, however. The mandate required the Commission to address the broader issues of exposure assessment, uses and limitations of risk assessment, the uncertainty and variability underlying risk estimation, risk management policies with regard to comparing and 'A copy of the Commission’s mandate is included as Appendix A.1. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 1-4 aa oOo Oe NH NN 10 1] 12 14 15 ‘17 18 19 20 21 22 23 24 communicating risks and choice of risk-based standards, and the desirability of consistent standards of negligible risk across agencies and programs.” The Commission was also asked to comment on the conclusions of Science and Judgment in Risk Assessment (NRC 1994) (see Appendix A.3). The Commission’s mandate was particularly timely because of the regulatory reform debate that began in the 103rd Congress and that reached full swing in the 104th Congress. The regulatory reform legislation proposed in those Congresses called essentially for an overhaul of the methods used to perform and to communicate the results of health and ecologic risk assessments, and specified criteria for rulemaking that would require the benefits of an agency rule affecting health, safety, or the environment to be reasonably related to its costs. where the results of a risk assessment would provide direct input to estimating those costs and benefits. Congress’ concerns reflected the views of many that risk-management decisions by regulatory agencies were overly stringent, were based on risk assessments that overstated and exaggerated actual risks to health and the environment, and were made behind closed doors by agency bureaucrats with no accountability. Congress’ response to those concerns was in turn viewed by many as an attempt to legislate science and to reverse twenty-five years of successful environmental protection. It has been the goal of the Commission to resolve some of those issues under dispute so that future risk assessments and risk-management decisions will be science-based where possible, and based on informed and reasonable policies and judgments when scientific support is scarce. This report is the product of the Commission’s deliberations and evaluations, and responds to 7A survey of federal risk assessment and risk management practices is included as Appendix A.2. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 1-5 tod HR vn + the concerns of those who provided testimony to the Commission to the extent possible.’ This is a draft report intended for public review and comment. The Commission welcomes written comments on the report, and asks that they be sent to the Commission’s office at 529 14th Street, NW, Suite 452, Washington, DC 20045. Comments should be received in that office by June 15, 1996, if they are to be considered in the preparation of the final report. The Commission’s final report will be issued in August 1996. 3A list of the individuals who testified at Commission meetings is included as Appendix A.4. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 1-6 Oo oN DH Nn 10 12 13 | 4 15 16 17 18 19 20 21 22 2.0 Uses and Limitations of Risk Assessment in Regulatory Decision-Making Health risk assessment has evolved from an aid to regulatory decision-making conducted informally by technical experts behind closed doors, into a somewhat standardized process that is the subject of research, symposia, graduate school courses, and Congressional debate. Parties affected by risk-based decisions demand an open and accessible regulatory process including risk assessments that reflect their views. Many academic scientists believe risk assessment can and should not be done because of their necessarily subjective basis. Man environmental activists think risk assessments are inappropriate because they crag sete spectrum of individual sensitivities and multiple exposures that occur in a population. Critics of risk assessment are justified in their criticisms; however, decisions must be made about the mst fet a ; ; ; _pest wa to protect and improve the quality of human health and the environment, and despite its many limitations, risk assessment has emerged as a useful adjunct to such decision- making. This chapter makes recommendations about the conduct of health and ecologic risk assessments that are hoped will improve the tool of risk assessment and its utility and relevance in decision-making. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 10, 1996 Page 2-1 nh 1v%) Oo oN KH KH 10 11 12 13 14 16 17 18 19 20 21 22 23 24 25 26 2.1 Cancer Risk Assessment 2.1.1 @ ISSUE: Tremendous efforts are devoted to the identification and application of mathematical dose-response models used for low-dose extrapolation of the effects of suspect human carcinogens. The accuracy of those models at low doses is not known. RECOMMENDATION The Commission recommends that a margin-of-safety approach, like that currently used for noncarcinogens, be explored for the purpose of setting standards for carcinogens. RATIONALE A large part of the debate about cancer risk assessment has focussed on identifying the correct mathematical models to apply to bioassay or epidemiologic data to extrapolate below the range of effects that can be observed at high doses. Because an effect below that which is observable is, by definition, unobservable, the accuracy or validity of those models at low doses cannot be known. Consequently, the accuracy or validity of the potency estimates obtained on the basis of those models is not known. The purpose of identifying exposure concentrations associated with negligible risk is public- health protection. Public-health protection is not served by endless debates about mathematical dose-response models that delay regulatory agency’s abilities to set standards. A simplified method of identifying appropriate standards for carcinogens is needed. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 2-2 oOo oN DH NO FF YW WN N NHN NY NH NY N N NY NY YY KF FP KF FP ee ee oN KA A BB WY NHN KH OO wWwN DH FF WY NHN © Many investigators have explored the potential relationship between toxicity and carcinogenicity, and found that there is a high correlation between the maximum tolerated dose (MTD) used in cancer bioassays and measures of carcinogenic potency (Bernstein et al. 1985, Crouch et al. 1987, Rieth and Starr 1989a,b, Zeise et al. 1984, 1985, 1986). Using the method of Gaylor (1989), Krewski et al. (1993) showed that an estimate of the upper-bound dose corresponding to the 95% upper confidence limit for an increased cancer risk of 10° based on the linearized multistage model can be made in the absence of a standard bioassay by dividing the MTD by 380,000. That is, the dose of a carcinogen that is associated with negligible risk in humans can be estimated by dividing the dose approximating a Lowest- Observable-Adverse-Effect Level (LOAEL) for toxicity by 380,000. Other authors have demonstrated similar associations (cite). The distinction between “nonthreshold” carcinogens and “threshold” noncarcinogens is becoming progressively blurred, and the resources available to investigate the mechanistic activity of carcinogens or other toxicants are progressively eroded. A method for setting negligible-risk standards that is less sensitive to understanding exactly how a substance elicits toxicity, but that can be relied upon to protect public health, is needed. Methods for setting standards for carcinogens on the basis of LOAELs or benchmark doses, and a margin of safety, should be explored. 2.1.2 @ ISSUE: When tested using chronic rodent bioassays, a number of chemicals elicit only tumors that are unlikely to have human relevance due to mechanistic or physiologic considerations. Regulating all substances that are positive in rodent bioassays as human carcinogens, without considering mechanisms of tumor induction and their human relevance, will not result in significant health benefits. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 2-3 Lo) Oo Se ND NH 10 12 13 14 15 17 18 19 20 21 22 23 24 26 27 ‘@ RECOMMENDATION The Commission recommends that when chemicals tested in rodent bioassays induce only Cec he he ly . tumors that are not relevant to humans, they should not be regulated on the basis of carcinogenicity. Criteria are needed to facilitate decisions regarding human relevance, so that risk assessments of such substances are no longer needlessly delayed. € RATIONALE Approximately half of the over 600 chemicals tested for carcinogenicity in rodents by the National Cancer Institute or National Toxicology Program yielded results considered positive in at least one sex of one species tested. Many of those chemicals were common food components such as vitamins, essential elements, and sugars, that have no evidence of carcinogenicity in humans. Some of those chemicals induced tumors in rodent organs that have no human equivalent, such as the forestomach or Zymbal gland. Some induced tumors using biologic mechanisms that have no human equivalent, such as o-2-globulin-mediated male rat kidney tumors. And some, like saccharin, induced tumors only at doses that were so high that the tumors resulted from high-dose toxicity and not from any inherent carcinogenic properties of the chemical. As currently practiced, cancer risk assessment produces statistical estimates of risk that are us for regulatory purposes but that have little biologic basis. Mechanisms of carcinogenesis are considered in a weight-of-evidence context, but their relevance to human cancer risk is not explicitly evaluated. The revised cancer risk assessment guidelines currently under development specify that during the hazard identification phase, three categories of classification are possible: likely or possible human carcinogen, not relevant to human cancer risk, and unknown. The guidelines include no explicit criteria for classifying a substance as PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 2-4 0 ON DA NH PR W NY | ee at aA & Ww NO —& & irrelevant to human cancer risk, however.' At least ten years have passed since the human relevance of «-2-yglobulin-associated kidney tumors was first questioned, yet it was only relatively recently that EPA made the decision to consider them irrelevant (EPA 1991). Over fifteen (twenty??) years have passed since the human relevance of saccharin carcinogenicity was doubted, yet packages of sugar substitutes including saccharin still must carry the legally required warning that its use may be hazardous to health because it has been determined to cause cancer in laboratory animals. {insert thyroid follicular cell reference when obtained] The relevance of a variety of other tumors has also been questioned for at least ten years—male B6C3F, mouse liver tumors, Swiss mouse lung tumors, rodent Zymbal gland tumors, rodent forestomach tumors—and decisions regarding their use in risk assessment have yet to be made. Delaying such decisions can only lead to wasted time and resources. Criteria must be developed for classifying substances and tumors as irrelevant to humans so that future decisions can be made as quickly and efficiently as possible. 'The guidelines do indicate that to depart from a standard default assumption, there must be an accepted theoretical basis for an alternative mechanism, and adequate evidence to demonstrate that a particular case fits that alternative. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 2-5 0 on Au FF WwW LY 10 11 12 13 14 16 17 18 19 20 21 22 23 24 25 26 27 2.2, Noncancer Risk Assessment 2.2.1 @ ISSUE: Current quantitative methods for evaluating the likelihood of adverse health effects other than cancer cannot be used to estimate the magnitude of those risks above the benchmark used to distinguish unacceptable from negligible risk. Communicating information about noncancer risks and identifying appropriate risk-management options would be more effective if quantitative information on the magnitude of noncancer risks were available. @ RECOMMENDATION The Commission recommends that methods be developed to estimate the magnitude of noncancer risks when toxicant doses exceed those associated with negligible risk, and that quantitative estimates of noncancer risk be accompanied by qualitative information on the nature and severity of the health effects that might be expected. € RATIONALE Currently, regulatory agencies evaluate risks to human health other than cancer using benchmarks such as Reference Doses (RfDs) or Acceptable Daily Intakes (ADIs). If toxicant doses do not exceed their benchmarks, risks to health are considered unlikely; when doses exceed their benchmarks, risks to health are considered possible. Such comparisons do not generate quantitative estimates of risk, nor do they provide any information on the nature and severity of the health effects to be avoided. Decision makers, affected parties, and the public need more information than a simple benchmark comparison if useful, defensible, and cost- PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 7, 1996 Page 2-5 io oN DH Or Ff W NY NN NHN NWN NH HN WN ee mee ~PPRPRRBPRBG&Gae AAA RD HH DS effective decisions about methods for risk reduction are to be identified and implemented. 2.2.2 @ ISSUE: Current quantitative risk assessment methods for health effects other than cancer have a number of limitations, but appear to be adequately protective of human health. @ RECOMMENDATION To help overcome many of the limitations inherent in current noncancer risk assessment methods, the Commission endorses the benchmark dose approach’ for assessing risks to human health when adequate data to support its use are available. @ RATIONALE Less effort has been directed towards developing methods to assess the risks of noncarcinogens than of carcinogens. One reason for the disparity is the heterogeneous nature of health effects other than cancer. Another is a lack of consensus about how to account for inconsistent experimental design. -- Sfecdaohie (eneawy The method currently in use to set standards for regulating noncarcinogenic toxicant exposures, the No-Observed-Adverse-Effect-Level (NOAEL)/uncertainty factor approach, does not make full use of available data, ignores dose-response information, is constrained by experimental design, and lacks a biologic basis. Identification of NOAELs is subject to a great deal of judgment and inconsistency—a recent review of an OECD pesticide project compared the NOAELs identified by regulatory agencies of five OECD countries, and found them to differ 20- to over 30-fold. Additional variation in the application of uncertainty factors to NOAELs to set standards for acceptable levels of exposure contributed to ‘A benchmark dose is a statistical lower confidence limit for a dose that produces a predetermined change in response rate of an adverse effect compared to background. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 7, 1996 Page 2-6 = 0 OND KA A WwW WN. eet pet mW BP WwW NY KF OO 17 18 19 20 21 22 23 differences that ranged up to three orders of magnitude. In contrast to the NOAEL/uncertainty factor, the benchmark dose approach takes advantage of dose-response data, incorporates data variability, offers flexibility with regard to the response level of concern, and accounts more accurately for experimental design. The benchmark dose approach also lacks a biologic basis, but is at least consistent with relationships between dose and response. It also has the disadvantage of relying on the use of uncertainty factors to calculate RfDs or other standards from benchmark doses (although none would be necessary to account for use of a Lowest-Observed-Adverse-Effect Level in the absence of a NOAEL). Further application and development of the benchmark dose approach is encouraged, to improve its scientific basis. It should continue to be applied to a variety of end points of toxicity, including ecologic end points. Its application to nongenotoxic carcinogenic responses should be pursued. Methods to incorporate mechanistic or biologic-based information should be developed. Adopting a common response level approach to assessing the risks of diverse end points, such as that provided by benchmark doses, should contribute to a greatly improved ability for risk managers to compare potential actions and to a greater consistency among risk-management decisions. NOAEL/uncertainty factor-based standards currently in place are sufficiently protective of human health, however, and should be changed only if available data indicate that a benchmark dose-derived standard would more accurately reflect the likelihood of a substance’s toxicity. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 7, 1996 Page 2-7 nN La Co Oo ND HD NH 10 1] 12 14 15 16 17 18 19 20 21 22 23 24 25 26 2.3 Ecologic Risk Assessment 2.3.1 € ISSUE: Is the EPA framework for evaluating ecological risk appropriate? @ RECOMMENDATION The Commission supports the use of the EPA ecological risk assessment framework with the critical addition of stakeholder involvement in the initial problem formulation stage. Clear, explicit guidance is needed for several aspects of the framework. ‘@ RATIONALE Ecological risk assessment has been used informally for many years to make decisions about resource management and pollution control. However, it is only within the last few years that a concerted effort has been made to define the characteristics of ecological risk assessment and to establish a common language for discussing approaches and results. At the same time, there are a greater number of ecological risk assessments being done by an increasing number of federal agencies. The growing consensus around the EPA ecological risk assessment framework makes it especially important that it fulfill this wide range of needs. In particular, the framework should include stakeholders in the initial planning stage of the process and there should be clear, specific guidance on the framework’s implementation. The EPA ecological risk assessment framework is an appropriate template for organizing and evaluating information on risks to non-human living systems (see figure A). In the problem PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-8 Ecological Risk Assessment PROSLEM FORMULATION } Characterization | Characterization ot | of Exposure Eesiogical Exects HupoyUuow pus uopjedjeA ‘uosinboy e1eq Discussion Between the Risk Assessor and Risk Manager (Resuits) i Risk Management al] — —s ee eee eee L__—-~-~—--» Figure A: USEPA Ecological Framework b&b WwW oO SO 1 HD NH 10 11 12 14 15 16 17 18 19 20 21 23 24 25 26 27 28 formulation stage, the environmental values to be protected or the goals of the assessment are defined. In addition, the appropriate level of ecological organization (such as individual specie, population or community), the endpoints or potential receptors of stress, and ways to measure those endpoints are identified. In contrast to human health risk assessment, in which stakeholders, risk assessors, and risk managers tend to share an essentially common view of the value of individual human beings and the health of the general population, ecological risk assessment has no commonly accepted starting point. For example, some may focus on the need to maintain biological diversity, others may be drawn to protecting particular plants or animals. while still others may relate to aesthetic quality. Balancing these disparate goals is the challenge of the problem formulation stage and the likelihood of success will be increased by including stakeholders in the process at this early stage. Figure B reflects the Commission’s proposal to add stakeholders to the participants in the problem formulation stage. There may be many small or well-defined assessments that are part of established regulatory programs where it may not be practical to involve stakeholders in each and every case. In particular, stakeholder involvement should be considered for larger local or regional assessments where affected parties hold a range of interests and values. The collaboration between risk assessors, risk managers, and stakeholders provides an opportunity to bridge the gaps in understanding, language systems, and values. If the affected parties do not participate in the early decisions about goals, endpoints, and measurements, then the analysis is likely to fail to provide useful information for decision-making. Consideration of economic and legal issues will also be facilitated by the early inclusion of stakeholders. Stakeholder involvement in the planning and problem formulation stage of the ecological risk assessment has been endorsed by a range of organizations, including the Risk Science Institute, the American Industrial Health Council, the Environmental Defense Fund, the State of California, and Environment Canada. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-9 FigureB: Modified Ecological Framework Discussion Amang the Risk Assessor, Risk Manager, and Stakeholders (Planning) Ecological Risk Assessment PROBLEM FORMULATION i | of €xposure [| Charactenzation Charactenzation RISK CHARACTERIZATION HupoyjUoW PUB HONBIPOA :uaysinbay e,eC) Other Factors and Stakeholders (Resuits) Discussion Among the Risk Assessor, Risk Manager, - I —> Risk Management (pepeeu su) suojesey/seset N v>} In a review of ecological risk assessment case studies, EPA said that the strengths and weaknesses of the studies seemed to originate, in large part, from decisions made during the probiem formulation stage. However, there is very little guidance on how this process should occur and who should be involved. The addition of stakeholders in this stage requires guidance on who, when, and how to include affected parties. The analysis stage of the EPA ecological risk assessment consists of two distinct but interrelated activities, exposure characterization and ecological effects characterization. During the exposure characterization, the spatial and temporal distribution of a stressor or stressors and their contact with ecological components are predicted or measured. During the characterization of ecological effects, the adverse effects elicited by a stressor or stressors and, potentially, the cause-and-effect relationships are evaluated. One method for analyzing cause-and-effect relationships is the index of biotic integrity developed by Karr (Karr 1991) that is now in use by more than 30 states in their water quality programs. The index of biotic integrity is a multi- metric index that documents the equivalent of ecological dose-response curves. Guidance is needed on when to use this tool and others of varying complexity, such as fate and transport models, toxicity tests, and field studies, and which tools are most appropriate for a given problem. Finally, in the risk characterization stage, the exposure characterization and the ecological effects characterization are integrated to evaluate the likelihood that adverse ecological effects can be associated with exposure to a stressor or stressors. The assumptions and uncertainties of the assessment are explained and the strengths and weaknesses of the analyses are described. Risk characterization for ecological risk assessments is an area with little standardization. For example, there are many sources of uncertainty in ecological risk assessment and guidance is needed for the use of qualitative and quantitative descriptions of uncertainty. Guidance with explicit directions and examples would greatly improve the conduct of this important stage in the ecological risk assessment. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-10 to N oO wm N DN & YD G2 Oo we N HRN & In some cases, risk characterization is interpreted simply as a restatement of test results. In other cases, risk characterization is viewed as the final stage of a weight-of-evidence evaluation that relates the analysis results to the assessment endpoints. However. there is no consensus on the definition of “weight-of-evidence” evaluation or how it should be applied. Often the approach reflects an individual’s professional judgment and the conclusions may not be transparent to others. There are three ways in which this tool can be improved. A definition of a “weight-of-evidence” evaluation should be established for use in ecological nsk assessment. An effort should be undertaken to examine the professional judgments that underpin weight-of- evidence evaluations and how they can be made more explicit. Finally, guidance for conducting quantitative and qualitative weight-of-evidence evaluations should be developed. As the final step in the framework, the risk characterization should synthesize and provide information that can be understood and applied to risk management decisions. The EPA ecological risk assessment framework has been most successful in analyzing risks from chemical stressors because that scenario is the most similar to a human health risk assessment. However, the framework is being used with greater frequency for more complex problems with modifications well within the overall framework. This maturation of the framework tool is critical if it is to assist in solving the important problems of protecting biological diversity, maintaining ecosystem health, and guiding sustainable development. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-11 w 2d i Oo CO ~F HD WN 2.4 Sensitive Subpopulations Requiring Special Consideration 2.4.1 € ISSUE: Differences in individual susceptibility, concurrent exposures, and cultural practices make some populations more sensitive to the effects of toxicant exposures. € RECOMMENDATION The Commission recommends that risk assessments be conducted so as to identify increased risk to potentially sensitive subpopulations by involving affected parties in the early stages of the assessment, evaluating all known sources of exposure to a particular toxicant and to toxicants with similar or synergistic modes of action, and characterizing exposure factors specific to particular subpopulations. € RATIONALE There are a number of potentially susceptible and sensitive subpopulations that may be of special concern when conducting risk assessments and making risk-management decisions. Susceptibility may be determined by a number of factors, including age, gender, genetic predisposition, ethnic origin, socioeconomic status, geographic location, and lifestyle. Current regulatory approaches for controlling toxicant exposures generally do not reflect those differences in individual susceptibility, nor do they account for elevated levels of contaminant exposures that may occur in minority communities or areas of lower socioeconomic status. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-12 — —_- © —" Ne — Wa 14 15 16 17 0 oN HUN Fw NY Increased risks of adverse health effects from contaminant exposures can result from increased exposures or from an increased ability to react to a given exposure. Exposure is a function of the concentration of a substance in the environment and the degree of contact an individual has with that substance. Susceptibility to the effects of exposure depends on the sensitivity of an individual’s response to changes in the dose. The following charts present examples of factors that can place particular populations at potentially higher risks. High Risk Based on Exposure Population Factors Affecting Exposure Level Industrial and agricultural workers Elevated exposure to airborne and dermal toxicants; increased activity resulting in increased dose of inhaled toxicants relative to someone at rest Sports and subsistence fishermen Elevated consumption of contaminated fish Low income and minority communities Elevated exposure to lead PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-13 l High Risk Based on Susceptibility 2 3 Population Factor Affecting Response to Exposure 4 Asthmatics . | Increased airway responsiveness to allergens and respiratory irritants a2 % AL 5 Infants/young children Increased sensitivity to the neurological effects of lead exposure Alpha | antitrypsin-deficient individuals Innate pathological changes within the lung aggravated by exposure to airborne irritants Elderly Diminished detoxification mechanisms 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 The Clinton Administration, the 103rd and 104th Congresses, and several interest groups have made attempts to address the issue of sensitive or high-risk populations in several ways. The Clinton Executive Order 12898 on environmental justice is aimed at ensuring that federal programs protect minority and low-income populations from disproportionately high exposures and adverse human health and environmental effects. In Congress, amendments have been proposed to the Safe Drinking Water Act, regulatory reform legislation, the Federal Insecticide Fungicide and Rodenticide Act, and other bills that would require standards to be set so as to protect such subpopulations as the elderly, children, and women of childbearing age. EPA has responded to the potentially greater susceptibility of one subpopulation, children, by issuing a new policy that will, “for the first time [require that] assessments of environmental risks will consistently take into account health risks to children and infants from environmental hazards in the air, land, food, and water” (EPA 1995). The new policy followed a National Research Council report that concluded “variations in dietary exposure to pesticides and health risks related to age and to such other factors as geographic region and ethnicity are not addressed PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-14 Go & Oo On HD NH 1/ 18 19 20 21 22 23 24 25 26 27 28 in current regulatory practice” (NRC 1993). The use of safety factors in standard-setting is an attempt to account for and protect sensitive populations in the absence of specific knowledge about the nature or extent of that sensitivity. Generally, risk assessments use conservative exposure assumptions and either uncertainty factors or conservative dose-response modeling assumptions to account for variations in exposure and response among different individuals. Those methods of attempting to consider potentially high-risk populations are rarely sufficient to address site-specific concerns, and are increasingly criticized. As knowledge and information increase, there is an opportunity to move away from those default assumptions. For example, characterizing exposure factors specific to a particular subpopulation can target a risk assessment and broaden risk-management options. In one particular case, the Commission learned at a hearing in Seattle from Asian and Pacific Islanders regarding the importance of considering their fish consumption patterns. The diets of this population consist of a much higher level of fish consumption and consumption of parts of the seafood that concentrate pollutants than the general population, placing them at higher risk from contaminants in fish. Incorporation of this exposure information into the risk assessment of Puget Sound enhanced its quality and provided valuable information for the risk-management decision. Another situation in which using specific information gathered from the community and stakeholders could reduce the need for default assumptions and improve the quality of a risk assessment might be that of a community with a disproportionately high number of polluting operations such as a municipal incinerator, a chemical plant, and an abandoned hazardous waste site, placing it in a category of higher environmental risk relative to other communities. Involving that community and other stakeholders in the planning stages of a risk assessment in that community would help identify sources of toxicant exposure, age and occupation of citizens, and other factors that might influence risk. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-15 Finally, there are opportunities to identify and evaluate risks to sensitive individuals. Asthmatics, for example, comprise 5-10% of the general population in the United States. Some types of air pollution can pose a greater risk to this subpopulation than to the general public. By identifying the size of the population at risk and characterizing the risk specific to that population, it is possible to make a more realistic characterization of the risk than if it were based on the general population. on Eu vivonmend cavedl he rendeud fe pe Need. Jjakeuts- aN fo alherpaa or vfecTtpes MrT. bt flew E 2 ry , ~s Cd vv Re GC cl ee "Ot cl ‘iy oe. a susceptible population’ s'risk will necessarily result 1 an more stringent regulatory-restrictions. i 0 Spécific subpopulations can- lead toa risk management decisten that emphasizes educatibn, a6 it did in the case of the Asian and Pacific Islanders in Seattle, where risk management consisted of distributing educational brochures and sign postings around the affected water bodies. Contaminated urban industrial sites, sometimes referred to as brown fields, offer another opportunity for designing, with the involvement of minority and low-income stato risk-management strategies with less stringent clean-up standards( omic redevelopment Enforcement and evatation ategies suc Se 18 ¢ ¢€ reduction im risk is PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 2-16 nN ue Oo fo 7 HD Nn Ff 10 12 13 14 16 17 18 19 20 21 22 23 24 25, 26 2.5 Uncertain Risk Estimates 2.5.1 @ ISSUE: While there is general agreement as to the value of qualitative statements describing critical uncertainties in health risk assessments, formal quantitative approaches to uncertainty analysis are difficult to perform, potentially inaccurate, and may be unnecessary. @ RECOMMENDATION The Commission recommends that qualitative descriptions of the primary sources of uncertainty associated with a risk assessment should be included in a,risk characterization, but a formal quantitative approach Sconsi SN outine risk assessments. € RATIONALE Most estimates of potential human health risks from chemical exposures in the environment are plagued by: incomplete sampling and analysis of contaminated media; mathematical models of those incomplete data instead of measurements of actual exposure levels; generalized demographic information from which assumptions about actual exposure conditions, frequencies, and durations must be made; default assumptions about population characteristics that presume all members of the population to be identical; and information on chemical toxicity that is derived from poorly characterized workplace exposures or high-dose experiments in rodents. Most of the assumptions used in risk assessments incorporate some conservatism to be health-protective (e.g., using an upper limit on contaminant levels instead of average levels; assuming that the most sensitive species represents human sensitivity), so PRIVILEGED DRAFT—DO NOT CITE, QUOTE, OR DUPLICATE January 10, 1996 Page 2-17 i) ad Oo won Hn nN that many believe risk estimates are generally higher than actual risks. Because risk assessments concatenate multiple conservative assumptions, estimated risks might be several orders of magnitude greater than actual risks. Alternatively, lack of understanding of a _ substance’s underlying toxicity or its low-dose mechanisms, for example, could leadNo significant underestimation of actual risks. It is important that risk assessments incorporate some evaluation of the degree of uncertainty associated with risk estimates, so that the level of confidence that may be placed in those estimates is known. Support for routine, formal quantitative analysis of uncertainty is based on the desire to move away from poorly supported default assumptions and point estimates of risk that convey a sense of false accuracy and that fail to convey any sense of the confidence that the msk assessor has in the estimates or their inherent complexity. Providing a numerical range of risk estimates reflecting uncertainty and variability is thought to allow more informed and transparent decisions than are possible when only a single point estimate is generated. However, communicating a range of risk estimates may be misconstrued by those unfamiliar with quantitative methods as implying that each of the numbers in the range is equally likely or plausible, and therefore valid for regulation. Many risk estimates are crude yardsticks for decision-making. In this context, the routine provision of a range of risk estimates may only confuse and delay the regulatory process. Generating ranges or probabilistic distributions of risk estimates instead of point estimates is thought to portray more accurately the range of possible risks experienced by an exposed population. When data are scarce, however, assumptions about the underlying shape of the risk distribution will be needed—that is, when uncertainty is greatest, a range of probabilities based on assumptions would replace point estimates based on assumptions. Approximating uncertainty introduces yet another source of uncertainty. Providing distributions of risk estimates is also thought to counteract the perceived pro- PRIVILEGED DRAFT—DO NOT CITE, QUOTE, OR DUPLICATE January 10, 1996 Page 2-18 wo > Oo CO ~~ HD WN regulatory bias inherent in compounding conservative default assumptions. Any range of risk estimates will inevitably include an upper-bound confidence interval at least as stringent as currently provided by point estimates. however. When confronted by an array of estimates, regulators and community groups are likely to choose from the more stringent portion of the range. Using formal uncertainty analysis to support less stringent regulation is unlikely to succeed. If the risk-management process is perceived to be overly stringent, then the nsk- management process should be modified, not the risk assessment method. Advancing risk assessment as a tool for public and environmental health decision-making should be seen primarily as a problem of biology and public health, not of applied mathematics. lastead-of-devotine able-resources-to-devetopi -for_routine-mathematieattinecertainty-anatysis>determining the tox underlying disease causation snould Sepursucd. Ae . vevants bile ou 2.5.2 ‘@ ISSUE: Few risk-assessment issues easily lend themselves to validation, and many uncertainty issues in risk assessment are inherently unresolvable. @ RECOMMENDATION The effectiveness of risk-reduction strategies should be monitored wherever possible.' Health and environmental data should be linked more closely to create a more integrated public- health context for risk and to provide a more fruitful basis for addressing uncertainties in risk assessment than is possible using quantitative uncertainty analysis. € RATIONALE 'Useful models in this regard are the ACGIH’s ongoing review of occupational health standards and the Harvard Six City Study. PRIVILEGED DRAFT—DO NOT CITE, QUOTE, OR DUPLICATE January 10, 1996 Page 2-19 N ae Science-based policy decisions are generally made in the absence of requirements for testing or validation. The nineteenth-century epidemiologist John Snow developed a well- documented hypothesis concerning the genesis of several cholera outbreaks in London during the mid-1800s. Based on that hypothesis, he convinced city officials to remove the handle on the Broad Street pump, a major source of contaminated water. Following this action, he evaluated its effectiveness, and noted a dramatic decrease in the incidence of cholera. Modern examples in which studies to measure effectiveness have proven useful are in the areas of occupational health and in evaluating the impact of criteria air pollutants. Generally, standards in those areas focus on acute health effects that can be measured by existing health data bases (e.g., vital statistics, hospital discharge data). Those standards are also supported by environmental surveillance information, thus enabling the study of the relationships between dose and effect. the aren OF safety between actual exposure levels and the health effect of concern is usually quite narrow, if it exists at all, so the effectiveness of an intervention is potentially subject to measurement. Few issues in risk assessment lend themselves easily to that sort of validation, however. Many current regulatory decisions focus on reducing public-health risks that are already relatively low. For example, risk assessments omit discussions of the health consequences of cigarette smoking, alcohol consumption, occupational injuries, or motor-vehicle accidents. Also, most health effects considered by risk assessments are chronic and multifactorial in nature (e.g., cancer, developmental effects, neurotoxicity). It is therefore very difficult to measure the extent of risk reduction achieved by an intervention, or to identify the impact of one specific intervention relative to others that are being implemented in the same time frame or relative to the background variability in disease incidence. For example, f intervention lowers the incremental risk of developing cancer from exposure to emissions from a local industrial facility from 1 in 10,000 to 1 in 1,000,000. No health PRIVILEGED DRAFT—DO NOT CITE, QUOTE, OR DUPLICATE January 10, 1996 Page 2-20 be Oo won nH Nn 10 1] 12 14 15 17 18 19 20 21 Comer ur llAtnnrtun a Count of nna, m1 azo0Vo incremental change of that magnitude cannot be measured. Conclusions abc effectiv study could be designed to measure the effectiveness of that intervention. because an Hoe a eal must rely exclusively on exposure information and the assumption that some proportional decrease in risk occurs when exposure is reduced. Considerable amounts of money are being spent to prevent or reduce risks whose existence can be neither confirmed nor denied, giving rise to arguments over cost and efficiency that cannot be resolved scientifically. In contrast to risk assessment, which focusses on specific risk factors, studies of public health focus on the prevalence of a particular health effect and how it can be influenced by incremental changes in risk factors (e.g., lowering the speed limit from 65 to 55 miles per hour to reduce the number of motor vehicle accidents, increasing the excise tax on cigarettes to prevent smoking among youths). The success of public-health interventions, from John Snow to the present day, has been due to the ability to demonstrate their effectiveness in improving health status. As the ongoing challenges to those interventions demonstrate, however, there are implementation difficulties even when the underlying data base is supportive. Developing good baseline and surveillance information about disease incidence, linking health and environmental data, and determining regional differences in disease prevalence, their trends over time, and their relationships to risk factors of concern, would improve our ability to implement effective interventions and be confident that they are, in fact, effective. PRIVILEGED DRAFT—DO NOT CITE, QUOTE, OR DUPLICATE January 10, 1996 Page 2-21 io fo YT DH WN Ff YI N N NO NY WN NHN WN KN NO KR YR Re BF KF PF Ee PE lS Ya na FB WwW HO —&— OO WON DH FP YW NY © 2.6 Peer Review 2.6.1 @ ISSUE: Peer review plays a critical role in risk assessment, economic analysis, and regulatory decision-making. ‘€ RECOMMENDATION The Commission recommends that clear, written, and easily accessible guidelines for peer review should be established by regulatory agencies and programs. Those guidelines should distinguish among the three stages of peer review in the regulatory process: addressing the validity of technical data, addressing their interpretation, and addressing the use of those data or their interpretaton in decision-making. € RATIONALE The development and evolution of scientific knowledge requires effective communication among scientists. Peer review is the most important and effective mechanism for facilitating this communication. It is also a mechanism for establishing priorities and for determining the accuracy or validity of data, observations, interpretations, conclusions, and policy recommendations. Peer review can vary from the simple act of seeking the advice of a colleague over the phone to a more formal procedure that incorporates many features of the judicial system. In the context of risk analysis, peer review can do more than increase the credibility of and confidence in an assessment—it can serve as a basis for building consensus among affected parties by including stakeholder representatives in a substantial and contributory role. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE December 18, 1995 Page 2-22 & WwW wo won nn MN 10 1] 12 14 15 >) / 18 19 20 21 22 23 24 25 26 27 28 The first stage of peer review in the regulatory process evaluates the accuracy, representativeness, and quality of technical data such as health and ecologic effects data, exposure data, or economic data. Technical data used to support risk assessments or economic analyses should be drawn from peer-reviewed literature. be subjected to peer review by independent scientific experts, and have been generated by studies that followed a published and generally accepted protocol for quality assurance. Raw data from studies that play key roles in an analysis should also be reviewed and evaluated at this stage. The second stage of peer review, interpretation of technical data, might involve issues such the choice of dose-response model used to extrapolate rodent tumor data for a particular substance to humans, the choice of endpoints used to evaluate the impact of contaminants on an ecosystem, or the choice of benefits used as part of an economic analysis and the basis of their cost estimates. This stage of peer review could also address broader policy data- interpretation issues, such as the choice of default assumptions generally used in risk assessments or decisions about departing from those default assumptions. Establishing guidelines for the final stage of peer review is problematic. Most peer-review panels are useful for evaluating highly focussed topics, but tend to lack an understanding of the history and philosophy of an agency’s decision-making process. Quality control of regulatory decision-making has traditionally been accomplished through the judicial system. Effective use of peer review as a collaborative decision-making process (which is really more quality control than peer review), that involves stakeholders or affected parties, can decrease the likelihood of controversy over the outcome and thus reduce the extent to which the courts must be relied upon. Implementing the framework for risk-management decision-making described in this report would be an effective way to address this category of peer review. Administrative details such as whether to use internal or external peer reviewers, how peer reviewers are selected, how consistency among an agency’s programs should or should not be PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE December 18, 1995 Page 2-23 toa a oOo Co NY HD Wn ensured, and how the outcome of the peer review will be implemented, should be addressed by an agency’s peer-review policies. Peer review should not be conducted simply to seek legitamacy for agency decisions and positions but should be used to improve the quality of decisions and positions. Bypassing the i whith wa hres y standard routes of validation via press releases, other media events, AO can short-circuit the self-correcting mechanisms of science and damage the process and image of peer review and quality control. @ RECOMMENDATION The Commission recommends that the level of peer review should be commensurate with the level of scientific importance and regulatory impact of the decision. @ RATIONALE Full peer review is unlikely to be needed for every regulatory decision. The most effective and efficient use of peer-review panels should be made on a case-by-case basis, taking into account issues such as the economic impact that a decision might have, the extent to which the information on which a decision is to be based might be considered controversial, and agency resource constraints. Peer review should not be used as a device to stall controversial policy decisions. @ RECOMMENDATION The Commission recommends that members of peer-review panels should be chosen on the basis of their expertise and with a goal of balancing, not eliminating, bias. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE December 18, 1995 Page 2-24 lua oO Oe ND nN 10 1] 12 14 15 16 ‘@ RATIONALE EPA’s peer-review policy specifies the need for independent scientific experts and the importance of avoiding bias. When selecting members for its committees, the National Research Council policy is to focus on balancing bias rather than an sminan it, dee eee et wabeseats the most knowledgeable committee members often have strong opinions in A ee LS areas of expertise. The Commission prefers the National Research Council’s * spproach to that of EPA, and believes that expertise should be the primary criterion for selection. Diversity of scientific expertise plays a very valuable role in peer review. Efforts should be made also to achieve a culturally diverse membership, to draw upon younger scientists, and to provide training or guidance in good peer-review practices. The individual or individuals responsible for selecting peer-review panel membership can have a great deal of influence on the nature of the bias of the membership, the areas of expertise represented, and by extension, on the outcome of the review. That “gatekeeper” role should be structured carefully to ensure that a small number of individuals does not have undue influence on panel characteristics or decisions. Newdl&ar sey, feu Reacts ple rf Lis {ot gate pepe PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE December 18, 1995 Page 2-25 2 oO mom nN ND NH fF 10 12 13 j4 17 18 19 20 21 22 23 24 2.1 C omplex Mixtures 2.7.1 ‘@ ISSUE: Humans are exposed to many chemicals simultaneously from the environment, but regulations focus on single chemicals and seldom take other exposures into account. Risk assessments generally assume that the risks from multiple agents can be added together to obtain total risk, and do not take into account potential synergistic or antagonistic interactions that could lead to under- or over-estimation of human risk. € RECOMMENDATION The Commission supports continued reliance on the assumption that either doses or effects, as appropriate, can be added together for the purpose of risk assessment when exposure to chemical mixtures occurs at low, environmental doses, and when these chemicals have similar toxic effects or affect the same organ. The components of mixtures with independent effects should be considered independently, not additively. @ RATIONALE Estimating the potential human toxicity of chemical mixtures is difficult because of inadequate chemical and toxicological characterization. For the purpose of human health risk assessment, the practice has been to assume either response additivity or dose additivity for similar components of a mixture. The additivity assumption has caused some concern because of the PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 30, 1996 Page 2-1 ba to Oo oOo SN HD \&N 10 ll 12 14 15 16 17 18 19 20 21 22 possibility that synergistic interactions’ among mixture components or their effects could occur, leading to a toxic response greater than that predicted on the basis of additivity and consequently to underestimation of the risk of human toxicity. Interactive effects (either synergistic or antagonistic) are usually highly dose-dependent, however (Filov et al. 1979); as a result, characterizing interactions that occur at one set of dose levels is likely to provide very little information about interactions at another set of dose levels. "High" dose levels for combined effects are defined as the exposure levels at which statistically significant increases in cancer risk, for example, are observed in either laboratory or epidemiologic studies, or as levels that are close to their NOAELs. For the most part, exposure to chemical mixtures in the environment occurs at "low" levels, however—at least three orders of magnitude below those at which a toxic response is observable in rodent bioassays. As a result, evaluating interactions that are observed in bioassays gives little insight into the effects of chemical mixtures at environmental levels of exposure. The combined effect of exposure to a chemical mixture is determined by the way in which individual components of the mixture affect the biological processes involved in toxicity. The components of a mixture can affect those biological processes in a variety of ways—any event that affects the absorption, distribution, metabolism, or elimination of a compound will affect the level of that compound that is available to react with DNA, for example, or other cellular target. Because all chemical-biological interactions are the result of reactions at many cellular sites with multiple molecules of agents, any mathematical dose-response model of a response that depends on such mechanisms would have to be non-linear at low doses. For example, if "Interaction" is a general term that has been applied to toxicity-test results that deviate from dose- or response-additive behavior expected on the basis of dose-response curves obtained from individual agents. "Synergism" is any result that is greater than would be expected from simple addition of doses or responses. In epidemiology, synergism is a result that is greater than would be predicted on the basis of multiplication of the individual relative risks. "Antagonism" is a situation in which the response is less than would be predicted on the basis of simple addition of doses or responses, or on the basis of multiplication of relative risks. Such classifications are thus dose-response model-dependent. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 30, 1996 Page 2-2 ivy) fo Oo fo I DH MN 10 11 12 13 14 17 18 19 20 21 22 23 24 25 26 27 28 two chemicals combined to form a carcinogenic agent, the rate of formation would be proportional to the product of the concentrations of the two chemicals. A linear reduction in the concentrations of the chemicals would thus result in a quadratic reduction in the formation of the carcinogenic agent and in its consequent risk. The nonlinearity of the typical chemical- biological interaction strongly suggests that mechanisms of any disease process that depends on such interactions are only marginally important at environmental levels of exposure. At high doses of one or more mixture components (such as cigarette smoke and some occupational exposures), the multiplicative effect term can dominate the toxic response, and the combined effect can be much greater than the sum of the individual effects. However, if exposures are reduced by several orders of magnitude, the combined effect would be, to a very close approximation, equal to the sum of the individual effects. Whether one or hundreds of mixture components are included, deviation from additivity would not be an appreciable relative amount. The NRC report Complex Mixtures (NRC 1988) supports that conclusion, stating, “On the basis of theoretical considerations and its examination of some epidemiologic studies, the committee noted that effects of exposures to agents with low response rates usually appear to be additive. The only examples of interaction that were considered greater than additive occurred in humans exposed to agents, such as cigarette smoke, that alone produced a high incidence of effects. Current quantitative models used to assess cancer risks support these results.” The additivity assumption should be confined to mixtures of agents that have similar toxicity or that affect the same organ, however. Exposure to agents with different targets and different effects will lead to risks of each effect that are independent of each other. The components of such mixtures should be considered independently. Experimental evidence appears to support the low-dose additivity or independence assumptions. For example, when eight or nine arbitrarily chosen noncarcinogens with unrelated mechanisms of action and target organs were administered to rats for four weeks, no PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 30, 1996 Page 2-3 nN Oo won KH DH S&S YW adverse effects were seen when the concentrations of each agent were one-third to one-tenth of their respective NOAELs. When the concentrations approximated their NOAELs, some minor toxicologic effects were observed. When the agents were administered at their LOAELs, however, a range of interactive effects was observed, both synergistic and antagonistic, in addition to additive effects (Jonker et al. 1990, Groten et al. 1994). In experiments using agents with the same target organ but different mechanisms of action, administration of four nephrotoxicants to male rats resulted in no effects at doses one-fourth of their respective NOAELs, in minor effects when doses were equal to their NOAELs, and in greater toxicity than that induced by each compound alone at doses equal to their LOAELs (Jonker et al. 1993) [need to review study—is this antagonism?]. Administration of four nephrotoxicants with the same mechanism of action to rats at doses equal to one-half their respective LOAELs resulted in clear nephrotoxicity, while doses equal to their NOAELs produced only a slight increase in kidney weight (no lower doses were tested) (Jonker et al. 1994). Overall, fewer than 3% of the 331 studies in the EPA Database on Toxic Interactions show clear evidence of synergism at bioassay dose levels. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 30, 1996 Page 2-4 N Oo Oo YN HH Fe W 10 1] 12 13 14 16 17 18 19 20 21 22 23 24 25 26 3.0 Risk Management and Regulatory Decision-Making Risk assessment can provide a valuable framework for setting environmental, health, and safety regulatory priorities and for allocating resources within regulatory agencies. Technical risk assessments seldom set the regulatory agenda, however, because of the different ways in which the non-technical public perceives risks. Risk assessment provides only part of the information that risk managers use, along with information about public values, statutory requirements, and cost-effectiveness, to make decisions about the need for and methods of risk reduction. Different regulatory goals have engendered different risk-assessment methods, different definitions of negligible and unacceptable risk, and different roles for risk assessment to play in decision-making. This chapter examines some of the issues that have arisen as the use of risk assessment in regulatory decision-making has evolved and matured. The use of bright lines, or benchmarks to distinguish negligible from unacceptable risk, has led to questions about what those lines should be, who decides what they should be, and to which situations they should be applied. Communicating decisions about whether a risk is or is not unacceptable to parties affected by those decisions has become a complex and confusing undertaking. Making decisions about how to allocate resources towards risk reduction can be made partly on the basis of risk, and methods to do so are developing. Making decisions that include information on the costs of implementing or failing to implement a risk-reducing activity, which can include consideration of the results of risk assessments, has become increasingly important in this era of resource constraints. Examining the legality of risk-related decisions and the process by which they were made can either assure reasonable, supportable decisions or hopelessly impede the PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 10, 1996 Page 3-1 & W WN regulatory process. And finally, striving for consistency among decisions made by different agencies can improve regulatory predictability but hinder regulatory flexibility. Recommendations on each of those issues are made that are hoped might contribute to the further evolution and improvement of risk-based decision-making. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 10, 1996 Page 3-2 Oo fo YD DH NH FF W NW 3.1 Bright Lines 3.1.1 & ISSUE: Should risk managers have clearly demarcated bright lines' defining boundaries between unacceptable and negligible risks to guide their decisions? € RECOMMENDATION The Commission supports the use of bright lines as guideposts or goals for decision-making. Using a range between bright lines as a goal (such as between incremental cancer risks of 10° to 10“), where decisions about protective action are negotiable, is consistent with the flexibility needed to account for uncertain and variable risks, differences in the size of populations potentially at risk, and differences in local factors such as community values. @ RATIONALE Bright lines are chosen to provide a pragmatic definition of “safe” and “unsafe”. A bright line is a single numerical value between unacceptable and negligible levels of risk. Regulated parties are expected to demonstrate that risk estimates are below the bright line in order to operate a manufacturing facility, introduce a new product to the market, or sell foods with low levels of contaminants. ‘An example of a bright line is 10° excess cancer risk, which means that if a risk assessment predicts that out of a population of 100,000 people exposed to a substance more than one case of cancer is likely to occur as a result of exposure, then that risk is unacceptable and protective action is required. Conversely, if the predicted risk is less than 10°, that risk is negligible and no protective action is required. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 3-3 0 Oo DA HW BF WN & wm NO HN NO NN NY N NY NO RB Be Ree RP Re ESE lL oN OKO AW BR BO NY KH Oo 0 me ND OKO OH Se WY KK © Risk managers are accustomed to the clear guidance provided by bright lines for implementing and determining compliance with risk-based standards or guidelines. Measurable contaminant concentrations—like permissible exposure limits (PELs) or threshold limit values (TLVs) in the workplace, action levels for food contaminants like aflatoxin on peanuts or mercury in swordfish, and National Ambient Air Quality Standards (NAAQS) for carbon monoxide or ozone levels in air—provide assurance that risks should be negligible so long as contaminant exposure concentrations are below the bright line of those values. If risks or contaminant concentrations are found to exceed their bright lines, action is expected to be taken to protect workers, consumers, or the community. Small quantitative differences in contaminant concentrations above or below those lines can make a big difference in whether protective actions are taken. Nonetheless, bright lines provide a basis for consistent decision-making. Bright lines expressed as contaminant concentrations are easier to implement than bright lines expressed as risks. Although concentration-based bright lines are derived from some judgment about what exposure level constitutes negligible risk, risk managers or compliance officers can easily determine whether or not they are being met because they can actually be measured. When bright lines are expressed as risks, uncertain and variable risk estimates must be compared to determine compliance. Comparing risk levels will become even more difficult as distributional approaches to risk estimation are implemented. Ranges of bright lines have often been adopted by regulatory policy. For example, under Superfund, a pair of bright lines has been used to define a potentially acceptable risk range for carcinogens. A contaminated site is considered to pose a negligible risk if a multi- pathway risk assessment of the site produces an upper-bound lifetime incremental cancer risk estimate not exceeding 10°. The site is considered an unacceptable risk, requiring remediation, if the risk estimate is 10% or higher. Between 10° and 10°, remedial actions, if any, are determined on a case-by-case basis. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 3-4 Oo oN DH NH BP WD NS pee et nA BP WwW NY | © 17 18 19 20 21 22 23 24 25 26 27 There are several potential problems with using specified bright lines. Bright lines are burdened by all of the uncertainty, variability, and assumptions inherent in risk estimation; thus, the all-or-nothing nature of a bright line could be misunderstood and construed to imply that an exact boundary exists between safety and risk. Risk assessments themselves could be manipulated so that their results occur above or below the bright line according to a risk manager’s particular policy preferences. Bright lines have the potential to be applied inflexibly, leading to decisions that do not reflect the unique characteristics of particular populations. Regulators and stakeholders have little or no experience using bright lines for decisions based on cost-effectiveness or cost-benefit analyses. wa wed acon peas LS pemuudrcls ye, (Le tare @ RECOMMENDATION. */<~3 «ital eal gry co ba amuse jr Hee wettest" ord tye ¥% eget In addition to ranges between bright lines intended io protect the general population, additional bright lines should be established to protect especially susceptible subpopulations, such as young children. pregnant women, or adults with lung disease. @ RATIONALE Section 2.4 of this report discusses sensitive subpopulations and the need to consider such populations in risk assessments. The results of risk assessments that include consideration of sensitive subpopulations might be expressed in terms of an estimated risk for the general population, and a different estimated risk for a sensitive subpopulation. Those risk estimates could be used to establish a bright line for the general population and a different bright line Se, for the sensitive subpopulation. Decisions about appropriate levels of risk reduction could then be made with the benefit of the knowledge of those differences. her Heed TA puagher suite. Tm furl Mave 15 Cane f af Me Sreeny oT Ue Acer | PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 3-5 0 Oo ND HW BP WN NM Nw HO WN NY NY NY NO KR KR KK FF SF BF Pe YI DA Wn BP WN KH CO DO ONAN BD Un FP W NY YF CO 3.1.2 @ ISSUE: Should bright lines be specified by Congressional legislation, promulgated as a part of normal agency rulemaking, or established by individual precedents? € RECOMMENDATION The Commission recommends that Congress leave the establishment of specific bright lines or ranges of bright lines to regulatory agencies. Congress should continue to provide broad guidance, using such qualitative language as “substantially reducing risk”, “achieving exposure levels associated with negligible risks”, or “assuring reasonable certainty of negligible risks” with regard to risks, and “benefits justify and are reasonably related to costs” with regard to economic analysis. @ RATIONALE Congress has included bright line risk provisions in several legislative bills proposed in recent years. Only in the 1990 Clean Air Act Amendments, however, did Congress pass legislation specifying a quantitative risk level for the first time, when it mandated the development of a strategy for controlling residual risks after Maximum Available Control Technology implementation based on an incremental lifetime cancer risk level of 10°. Bright lines have been well established by regulatory policy despite their absence in legislation. For example, the Food and Drug Administration regulates intentional and unintentional additives in food by calculating an “estimated daily intake” and comparing that value to a previously established “acceptable daily intake”. When the ratio exceeds 1.0, the agency considers the exposure unacceptable (Flamm & Lorentzen, 1988). Noncancer health effects are evaluated similarly under Superfund; contaminant doses are compared to values called Reference Doses. If the ratio is less than a bright line of 1.0, adverse effects are PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 3-6 oOo C6 SN DW Wn BP WW KB — o considered unlikely and no action is required. In practice, legislated bright lines may do little to constrain agency decisions, because the agencies (either centrally or regionally) will exercise considerable discretion in the conduct and evaluation of risk assessments (such as choosing and justifying assumptions and selecting the most relevant data sets), even if procedural guidance such as that proposed by the 104th Congress is enacted. For similar reasons, the absolute value of a risk judged to be negligible is of less importance than the size of that risk compared with similar risks, or with dissimilar but familiar risks. Even more important should be evidence that exposures and risks judged to be too high are, in fact, being reduced. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 3-7 Oo Oo nN KH NH FH W 10 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 3.2 Communicating Risle 3.2.1 &@ ISSUE: Risk communication is a critical component of the risk-management process, but it has received too little funding and too little attention by both risk assessors and risk managers. Effective risk communication greatly influences the acceptability of a risk assessment and risk-management decision to stakeholders. @ RECOMMENDATION The Commission urges the adoption of comprehensive risk communication programs within regulatory agencies that provide for research on risk communication messages, training of risk managers and others engaged in communicating risk to the public, and the inclusion of risk communication funding. objectives. and evaluation in risk management plans. € RATIONALE Since the process of risk assessment has been used by the federal government to support decision-making. there has been a need for risk communication. The National Research Council has defined risk communication as “an interactive process of exchange of information and opinions among individuals. groups, and institutions. It involves multiple messages about the nature of risk and other messages not strictly about risk, that express concerns, opinions, or reaction to risk messages or to legal and institutional arrangements for risk management” (NRC 1989). Inan effort to improve risk communication and thereby improve the understanding of risk, Congress has made various proposals to increase the transparency of risk assessments and PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 5, 1996 Page 3-8 Oo fo ND UW FP WN oO considered unlikely and no action is required. In practice, legislated bright lines may do little to constrain agency decisions, because the agencies (either centrally or regionally) will exercise considerable discretion in the conduct and evaluation of risk assessments (such as choosing and justifying assumptions and selecting the most relevant data sets), even if procedural guidance such as that proposed by the 104th Congress is enacted. For similar reasons, the absolute value of a risk judged to be negligible is of less importance than the size of that risk compared with similar risks, or with dissimilar but familiar risks. Even more important should be evidence that exposures and risks judged to be too high are, in fact, being reduced. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 3-7 oOo oN A vw Ff WY YN — meme UM F&F WwW NO — © 17 18 19 20 21 22 23 24 25 26 27 28 to require the use of risk comparisons. Transparency is generally equated with revealing and characterizing the assumptions, uncertainties, default factors, and methods used to estimate risks. Legislation has also been proposed that would require agencies to compare the risk to be regulated to other risks regulated by the agency and to other risks experienced by the public. However, risk communication is not a straightforward process. One of many examples where risk communication has gone awry is the case where a pesticide residue was compared to the risks associated with aflatoxin in peanut butter. Mothers responded angrily because the communicator was perceived as trying to trivialize their concerns and, moreover, was calling into question their abilities as mothers by pointing out another risk that was unknown to them. Both risks were not controllable at the individual level without giving up “me of value, generating great frustration. fbn een ares foget tage Ec, a fod seran seen bes et re dy of research HB prov es some me guidance Oa g risk, peek fi bate tty, effectively an& on using tisk co comparisons stot commuficate risk. Some researchers have suggested that people’s perception’s of risk must be considered, because they will influence how m anew activity, product, or situation is evaluated and accepted or rejected. Paul Slovic has ‘Dn identified seven psychological dimensions that influence people’s perceptions of risk: “peaks . voluntariness, exposed individual’s knowledge of risk, dread, severity of consequences, control, equity, and novelty (Holtgrave 1993). Another model of risk perception considers probability of gain, probability of loss, probability of status quo, and expected benefit and harm (Holtgrave 1993). he mental models approach suggests that people process new information within the context of their existing beliefs. The three main tenets of the mental models approach are: the recipient of any communication needs a basic understanding of the exposure, effects, and mitigation processes relevant to making decisions about a hazardous process; recipients’ existing beliefs affect how they interpret and use any new information; and risk information should be presented PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 5, 1996 Page 3-9 oO won nA Nn & W NY BO HN BH HN KN KB RO mmm RR ee nN nO B® W NY KK OD Oo ON DBD HA BPW NY | © with appropriate text structure and enforced with textual aids. Those researchers have said that “one should no more release an untested communication than an untested product” (Holtgtave 1993). Experimental research shows that people avoid unfamiliar risks more than known risks, even when objective probabilities are similar. Attempts to fully disclose uncertainties in risk analysis may thus generate public concern, suggesting not that the public be protectedfrom knowledge of scientific facts but that such information should be communicated carefully a8 With the growing use of risk assessments and risk estimates by regulatory agencies, there is a need to increase public understanding and credibility of that information. In general, agencies and Congress have emphasized the importance of improving the quality of risk assessments, while paying less attention to the need for training and educating risk assessors and risk managers on how best to communicate information about risk. Comprehensive risk communication programs need to be established within regulatory agencies. Funding for training risk assessors and risk managers in risk communication and for testing risk communication messages should be part of each risk management agency’s budget. In addition, communication should be a specific component of risk management plans. Specific communication objectives, such as awareness and involvement of stakeholders, should be identified in the plans, along with appropriate methods for evaluating the effectiveness of a communication. The state of the art of risk communication has moved from trying to explain risk information to a non-technical audience, to a highly evolved stage of building partnerships between plant managers and nearby residents, companies and consumers, and agency risk managers and the public. To make this transition successfully, an investment of time and resources is needed. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 5, 1996 Page 3-10 to oo en Dn UH S& W 10 11 12 13 14 16 17 18 19 20 21 22 23 24 25 26 3.3 Comparative Risk Assessment 3.3.1 & ZSSUE: Government agencies responsible for protecting human health and the environment are confronted with many statutory mandates but have limited time and resources to implement them. @ RECOMMENDATION The Commission recommends that agencies use the comparative risk-ranking paradigm to make resource allocation decisions. That paradigm includes organizing teams of analysts or stakeholders, such as business and environmental representatives; making a comprehensive list of environmental problems; assembling the best information possible about the sources of the problems and the risks they pose to human health, ecosystems, and the quality of life; ranking the problems in order of the seriousness of the risks they pose; and using the rankings to guide strategic planning and budgeting. € RATIONALE Priority setting by comparing risks is one way to confront and weigh choices when money, time, and staff are in limited supply. The call for greater use of this tool has come from many sources, including Supreme Court Justice Stephen Breyer, the Carnegie Commission on Science, Technology and Government, the National Academy of Public Administration, and many members of Congress. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 5, 1996 Page 3-11 Oo Oo NY WO Wn F&F WY YN BO NO NHN CN =| =| =| SF HP FO Rr OES lS Ww NY & OG 0 en DA nH fF WY NY | O&O N NHN NY WN YH NA Comparative risk assessment for priority setting is a process that brings together elements of risk assessment, cost-benefit analysis, strategic planning, and public involvement. Combining those analytic tools with questions of ethics, values, and principals of democratic governance leads to a very high level of complexity that requires commitment of technical and human resources. Although the Environmental Protection Agency and the Department of Energy have had some experience with comparative risk ranking for priority setting, the paradigm has developed primarily from the 34 state, 10 local, and 2 tribal projects fostered by EPA. To begin the process, a planning team is assembled to define the problems to be addressed and initially set project goals. The planning team writes a work plan that includes the project’s structure, budget, and methods. In addition, the team identifies the individuals needed to achieve the project goals, who then become the comparative risk team. Potential stakeholders include representatives from the highest political officeholder sponsoring the project, such as the governor or mayor; agencies, such as environmental protection, health department, natural resources, agricultural department, and land use commission; and legislators, academics, business interests, environmentalists, farmers, fishers, and ethnic and racial representatives. The organizational units include a project manager, who supervises all aspects of the project, and a steering committee that provides overall direction for the project. A public advisory committee ensures public participation in the process and that the project’s work remains understandable, relevant, and credible to the public. Finally, technical work groups perform data collection, data analysis, and preliminary rankings. The technical work groups may be arranged by medium, by risk type, or by combining them into one large work group (EPA 1993). While each federal agency will need to adapt the fundamental elements of the comparative risk- ranking paradigm to its mission, statutory mandates, and current and emerging responsibilities, it is easily translated to the federal level by substituting Congressional staff from authorizing PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 5, 1996 Page 3-12 aN HD WW FF W WN of 10 11 12 13 14 15 17 18 19 20 21 22 23 24 25 26 27 committees of the Congress for gubernatorial, mayoral, and state legislative representatives and identifying stakeholders based on the programs and projects of the specific agency. Depending on the agency, it will be important to include representatives from state, local, and other federal agencies with shared responsibility. The participants in each comparative risk project must decide whether or how to address such issues aS environmental equity, future risks, and effects across jurisdictional boundaries. Another area of early decision-making is agreeing on risk ranking methods and processes. Most comparative risk projects look at three criteria when ranking risks: effect on human health, effect on ecosystems, and effect on quality of life, including economic well-being. Ranking methods have ranged from voting by participants, formulae which rely more heavily on quantitative data, matrix-based discussions that employ graphics in a shared decision-making process, decision- seeking consensus, and bargaining or tradeoffs among stakeholders. Typically, a comparative risk ranking project may take two to three years to complete. Keeping participants, the press, and the political leadership enthusiasti motivated throughout the process can be a challenge. ed od" While priority setting may be the primary goal of comparative risk projects, there are often a number of other benefits that make the time and effort valuable (Minard 1993). . Comprehensive catalog of problems. Most comparative risk projects produce a catalog of a state’s environmental problems. The analysis is an important foundation for the project, yet can be a resource for the public and managers separate from the priority- setting goal. ° Increased knowledge among public and government decision-makers about a variety of issues. Participants in a comparative risk project learn about problems that are not part PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 5, 1996 Page 3-13 Oo Oo JN DB WA F&F YW NHN = Nm Bw BO RO RO my El kh WwW NO |= CGF Oo OB yD Nn A HBR WD NY KF O&O of their daily interests or responsibilities. The interdisciplinary activity improves understanding and appreciation of competing priorities and provides potential new insight into solutions. Teamwork and trust. As a result of increased communication between different institutions and interest groups, new avenues of cooperation can be established across agencies and with new interest groups. While adversarial relationships among interest groups may not be eliminated or turf conflicts between agencies may not disappear, comparative risk projects can reveal unexpected agreement among parties and understanding of differences in perspectives. Consensus for change. The process itself helps build coalitions that favor shifting priorities to higher-risk endeavors. In turn, the broader public support for a common agenda allows agencies and legislatures to move money and staff into priority areas with less litigation, less controversy, and léss second-guessing of each other. Increased public involvement has increased project success. Making significant changes in governmental activities takes public understanding and support. In a comparative risk- ranking process, where ranking includes value-laden choices, the group making the ranking should have a clear understanding of how the public’s values relate to the choices. The comparative risk ranking paradigm emerging from the state, local, and tribal projects supported by EPA provides a useful starting point for federal agencies to use in ranking priorities and making resource allocation decisions. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 5, 1996 Page 3-14 i 0 Om nN Hn nH &S& W 10 11 12 13 14 16 17 18 19 20 21 22 23 24 25 26 3.4 Economic Analysis 3.4.1 @ ISSUE: The results of risk assessments are frequently based on assumptions that are inconsistent with the needs of cost-benefit analysis. @ RECOMMENDATION The Commission recommends greater collaboration between risk assessors and economists who must rely on the results of risk assessments, to minimize the inconsistencies between scientific and economic approaches to characterizing risks. Where inconsistencies exist, they should be revealed explicitly as sources of analytic uncertainty. @ RATIONALE The results of a risk assessment are used in cost-benefit analysis to estimate benefits, but risk characterization end-points are often inconsistent with economic valuation start-points. The traditional methods of evaluating health effects for the purpose of health risk assessment can conflict with the needs of the economist who is asked, at least implicitly, to provide information on individual preferences for avoiding health risks. For example, a 10% improvement in lung function is not meaningful to most individuals. They do not demand greater lung function, they want fewer sick days. Health risk assessments seldom evaluate risks in terms of sick days, and there are no economic studies available that can be used to value a 10% improvement in lung function. Closer collaboration between economists who are familiar with the valuation literature and scientists who are estimating concentration-response PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 3-15 oO Oe nN Dn Nr Ff W YH mw NHN NY NY WN NY NY NN NN YF KF | SY KR FP PK PE le eo NWA Ama BON SF SF CM AKA DA BRB wWN HE DS functions can help avoid such mismatches by seeking end points that can be meaningfully evaluated in terms of both their risk and their economic value. Another conflict between the needs of the economist and the results of risk assessments is that health risk assessments generally focus on individual rather than population risk. There are two reasons why economic analysis focusses on estimating benefits for the population at large. First, if costs are to be compared with benefits, it would make no sense to compare total costs to the benefits experienced by only one (hypothetical “maximally reasonably exposed”) individual. Second, even if one were performing a cost-effectiveness analysis in which abatement costs per risk to the maximally exposed individual were being estimated, the resulting estimates could be very misleading. Suppose that two abatement strategies were equally costly, but one had a very high individual risk and low population risk (because few people were exposed to the pollutant of concern), while another strategy exposed many more people but the individual risk was small. A cost-effectiveness analysis based on individual risk would lead to adoption of the first strategy instead of the one based on the population risk, which could be considered the more relevant measure. Another inconsistency results from the traditional risk-assessment practice of building uncertainty about risk characterization into the assumptions used to estimate risks. This tradition purposely skews risk estimates upwards to build in a margin of error that is intended to protect a population from health risks (estimating average risk reductions instead might result in protection of only part of a population), and thus provides only one point at the upper end of a risk distribution. According to economic tradition, the analyst attempts to describe the distribution of risks (or the distribution of risk improvements) in the population and leaves it to the decision-maker to decide what is an acceptable level of protection and which strategies deliver that level of protection. Current trends towards moving away from expressing risk-assessment results in terms of upper-bound point estimates and using distributions of risks instead may overcome this inconsistency. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 4, 1996 Page 3-16 L L T OnsS> i ” HF Moats 1 TTERA- TIVE PROCE SS TrSocnahon srhering, Qutcorne Evaluation Oo Oe HN KD nH Fe W 10 L1 12 14 15 ‘6 18 19 20 21 22 23 24 25 26 27 28 greatest risks to public health or the environment; and, it does not provide a means to develop or identify the most cost-effective strategies to control hazards. Risk management is the process that should incorporate those considerations into decision-making, but currently there is no consistent, comprehensive strategy for managing, controlling, or reducing risks to public health or the environment. In the absence of a consistent, comprehensive approach to risk management, the Commission proposes the risk-management framework shown in Figure 4.1. Our framework puts a decision- analysis framework in an environmental and public-health risk context. The framework has five steps: problem, risks, options, decision, and action. Each step involves different sets of questions. Answers to those questions form the basis of the systematic and comprehensive nature of the risk-management framework. Use of a collaborative process and an iterative process guides how the answers are obtained. The following is a description of the five steps and the iterative and collaborative processes that occur throughout the five steps. 1. Problem: What is the problem? A problem might be identified on the basis of environmental monitoring, emissions inventories, disease surveillance, epidemiologic observation, or public concern. The problem should be examined in not just a medium- and pollutant-specific manner, but also in a comprehensive and multimedia context. Potential inter-relationships among different problems should also be considered. After the problem is characterized, goals and objectives of problem intervention are identified. 2. Risks: What risks does this problem pose to public health or the environment? Risk is considered to be the likelihood of an occurrence of an adverse effect on human health, the environment, or public welfare. The goal is to articulate the factual and scientific basis of the problem and to identify any subjective perceptions of the problem by characterizing its risks to PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 4-2 Oo fe HN DB Wn Ff WY NO NO NO NH HH VN KH NO RO RR et et NA ON FBP W NO KY DT OO WOH HD WH FP W NY OS human and environmental health, cultural and societal values, quality of life, and environmental equity. Cumulative risks from related problems should also be identified, and where appropriate, comparative risk analysis should be performed. 3. Options: What should be done about the problem and what are the potential consequences of intervention? Solutions to the problem are identified by stakeholders, regulators, and scientists, as appropriate, and might include both regulatory alternatives such as permits, regulations, and enforcement actions, and non-regulatory solutions such as pollution prevention, recycling, market incentives, voluntary reductions, or education. Institutional, financial, and other arrangements for implementing the solutions are identified. The extent of risk reduction and the relationships between the costs and benefits of each solution are determined and compared. Potential impacts of the solution, including ethical considerations, are characterized. 4. Decision: What is the best solution to the problem and how should that decision be made? The goals and objectives of problem intervention are reviewed and the most feasible and acceptable solution to the problem is identified, with involvement of affected parties. The criteria for feasible and acceptable might be that which is the most reasonable and cost-effective, or that which minimizes risks in the most cost-effective manner. A mechanism for conflict resolution, or for reaching closure in the absence of consensus, is identified and implemented. 5. Action: How effective is the decision? The solution is implemented and the outcomes of the solution are evaluated. The impact that the solution has on the problem is characterized, for example, through environmental monitoring or through analysis of relationships between inverventions and trends in health and environmental indicators. The original problem is redefined and the five steps repeated, if appropriate. The framework is implemented iteratively; that is, the process is refined based on continuing PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 4-3 Oo fo HN DH NT FS WG 10 11 12 13 14 15 7 18 19 20 21 22 23 24 information acquisition, verification, and monitoring. This process is similar to the one used in scientific investigations—conclusions can be changed based on new data. Iteration could apply to a rule that has already been promulgated and is found to be irrelevant or inappropriate in light of new information; or, iteration could occur as a new rule or approach to a problem is being developed, as public comment, negotiation, or analysis redefine that problem or other issues of concern. It is possible that exploring a problem more deeply in the analysis stages may lead to a better understanding of how a problem should have been defined and scoped at the outset. Using an iterative process to scope a problem may actually speed up the process, as goals and issues are clarified, possibly leading to a quicker resolution than expected initially if it becomes apparent that proceeding with the entire process is no longer necessary. Of course, iteration must not be allowed to become a device for indefinite delay. The framework is also implemented collaboratively; that is, the process is conducted with full participation of stakeholders or affected parties.’ Such partnerships facilitate the exchange of information and ideas that all parties need to make informed decisions about reducing risks. A number of studies have shown that the success of a regulatory action or decision depends on the involvement of affected parties in the scoping and decision-making process (Richards 1993). While risk assessors and risk managers may tend to base their responses primarily on technical and scientific information, non-technical stakeholders are likely to base their responses on very different, more value-laden perceptions and concerns. Both must play a role in decision-making if the outcome is to succeed—effective collaboration plays a central role in effective implementation, especially if the general public is expected to change its view of environmental protection as being solely a government-industry responsibility and to participate in both the choice and implementation of risk-management strategies (McCallum ‘Stakeholders are people or organizations that are likely to be affected by the outcome, and might include the community, elected officials, industries or businesses, and regulatory agencies. The identity of the stakeholders will depend on the characteristics of the particular problem to be addressed. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 4-4 Oo Oo SN DN FF WW NY NO BQO KN RQ RO BW NO FSF © OO wo nH DH FF WY NY —- & and Santos 1995). Meaningful stakeholder involvement in regulatory decisions will require a shift in attitudes of agency decision-makers as well, however, so that the affected public is seen as part of the problem-solving process rather than as an obstacle to it (Van Horn 1988, Chess et al. 1995). It is clear that “public comment” and “public meetings” are not substitutes for collaborative approaches to problem-solving (although they may be appropriate in some cases). A potential disadvantage of our framework may be the investments of both time and money required to implement a collaborative and systematic process. While the process may lead to considerable long-term savings, the up-front cost of implementation may be an obstacle.’ In addition, while assessing impacts on human and environmental health involve fairly well- established, if controversial and evolving procedures, evaluating impacts on public welfare, which includes considerations of costs, benefits, values, ethics, and perceptions, is considerably less straightforward. Different mechanisms for integrating those considerations into risk management must be explored. Thus there are three critical advantages of our risk-management framework, which represents a major shift in the role that risk assessment plays in risk management decision-making. First, an integrated, holistic, top-down approach to a public health or environmental problem is used instead of a chemical-by-chemical, medium-by-medium, bottom-up approach to characterizing individual risks. Second, communication, collaboration, and negotiation among stakeholders are emphasized in an open and inclusive process so that public values can be included in the shaping of risk-management strategies. The result is decisions that are more pragmatic and more easily implemented than those made in the absence of consensus, and solutions that no single “It is unlikely that performing every step of a complete analysis will be required for every decision-making problem, however. Different levels of decisions require different levels of analysis. The framework described here is meant to provide a guideline for a thought process that might be pursued when decision-making issues arise. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 4-5 A Nn -& WG participant could have devised because of the diversity of interests, knowledge, and technical expertise represented. And finally, like the scientific process, the risk-management process is iterative. At any stage of the process, conclusions and decisions can change on the basis of new information, and the problem can be reformulated and reevaluated as more information is acquired. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE February 6, 1996 Page 4-6 5.0 Recommendations for Specific Regulatory Agencies and Programs [as yet to be provided] PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE January 10, 1996 Page 5-1 References [not vet revised to retlect new dratt] Agency for Toxic Substances Disease Registry (ATSDR). 1988. The Nature and Extent of lead Poisoning in Children in the United States: A Report to Congress. Centers tor Disease Control. Atlanta, Georgia. Agency for Toxic Substances Disease Registry (ATSDR). 1995. Testimony before the Senate Appropriations, Labor, Health and Human Services, Education Subcommittee. May 5. Beranek, W., Jr., Davies, C.J., and Moses, S. 1985. Issues in Peer Review of the Scientific Basis for Regulatory Decisions. American Chemical Society. Bernstein, L., Gold, L.S., Ames, B.N., Pike. M.C.. and Hoel. D.G. 1985. Some tautologous aspects of the comparison of carcinogenic potency in rats and mice. Fund. Appl. Toxicol. 5:79-86 Berry, B.J.L. 1977. The Social Burdens of Environment Pollution: A Comparative Metropolitan Data Source. Ballinger Publishing Co. Cambridge, MA. Breyer, S.G. 1993. Breaking the Vicious Circle: Toward Effective Risk Regulation. Harvard University Press. Cambridge, MA Bullard, R.D. and Wright, B.H. 1992. The Quest for Environmental Equity: Mobilizing the African-American Community for Social Change. In: R.E. Dunlap and A.G. Mertig, eds. American Environmentalism. Taylor and Francis: Philadelphia, PA. Burns, J.E. 1974. Organochlorine pesticide and polychlorinated biphenyl residues in biopsied human adipose tissue—Texas 1969-72. Pesticide Monitoring 7:122-124. Byrd, D. and Lave, L.B. 1987. Narrowing the Range: A framework for risk regulators. Issues in Science and Technology. Summer. Pages 92-100. California Comparative Risk Project. 1994. Towards the 21st Century: Planning for the Protection of California’s Environment. Final Report. May. Chess, C., Salomone, K.L., Hance. B.J., and Saville, A. 1995. Results of a national symposium on risk communication: next steps for gevernment agencies. Risk Analysis 15:115-125 PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE November 3, 1995 Page R-1 Chess, C. and Hance. B.J. 1989. Opening Doors: Making Risk Communication Agency Reality. Environment: Vol. 31. Congressional Research Service (CRS). 1994. Risk Analysis and Cost-Benefit Analysis of Environmental Regulations. The Library of Congress. Costner, P. and Thornton. J. 1990. Playing with Fire: Hazardous Waste Incineration. Greenpeace USA. Washington. D.C. Crouch, E.A.C., Wilson, R., and Zeise, L. 1987. Tautology or not tautology? J. Toxicol. Environ. Health 20:1-10 Davies, J.C., Covello, V.T.. and Allen, F.W., eds. 1987. Proceedings of the National Conference on Risk Communication. The Conservation Foundation. Washington, D.C. Davies, J.E. et al. 1972. The role of social class in human pesticide pollution. Am. J. Epidemiol. 96:334-338. Deisler, P.F. 1988. [title] Environ. Sci. Tech. 22:15-19 Flamm, W.G. and Lorentzen, R.J. 1988. Quantitative Risk Assessment (QRA): A Special Problem in the Approval of New Products. In: C.R. Cothern, M.A. Mehiman, and W.L. Marcus, eds. Risk Assessment and Risk Management of Industrial and Environmental Chemicals. Princeton Scientific Publishing Co., Inc.: Princeton, NJ Gaylor, D.W. 1989. Preliminary estimates of the virtually safe dose for tumors obtained from the maximum tolerated dose. Reg. Toxicol. Pharmacol. 9:101-108 Gelobter, M. 1987. The distribution of outdoor air pollution by income and races: 1970-1984. Master’s thesis for the University of California Energy and Resources Group. Berkeley, CA. Gelobter, M. 1992. Toward a model of environmental discrimination. In: B. Bryant and P. Mohai, eds. Race and the Incidence of Environmental Hazards. Westview Press. Boulder, CO. Gold, L.S., Sawyer, C.B., Magaw, R., Backman, G.M., de Veciana, M., Levinson, R., Hooper, N.K., Havender, W.R., Bernstein, L., Peto, R., Pike, M.C., and Ames, B.N. 1984. A carcinogenic potency database of the standardized results of animal bioassays. Environ. Health Perspect. 58:9-21 Gold, L.S., de Veciana, M., Backman, G.M., Magaw, R., Lopipero, P., Smith, M., Blumenthal, M., Levinson, R., Bernstein, L., and Ames, B.N. 1986. Chronological PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE November 3, 1995 Page R-2 supplement to the carcinogenic potency database: Standardized results of animal bioassays published through December 1982. Environ. Health Perspect. 67:161-200 Gold. L.S.. Slone, T.H.. Backman, G.M.. Magaw, R.. Da Costa. M.. Lopipero. P.. Blumenthal, M., and Ames, B.N. 1987. Second chronological supplement to the carcinogenic potency database: Standardized results of animal bioassays published through December 1984 and by the National Toxicology Program through May 1986. Environ. Health Perspect. 74:237-329 Gold, L.S.. Slone, T.H., Backman, G.M.. Eisenberg, S.. Da Costa, M.. Wong, M.. Manley, N.B.. Rohrbach, L.. and Ames, B.N. 1990. Third chronological supplement to the carcingoenic potency database: Standardized results of animal bioassays published through December 1986 and by the National Toxicology Program through June 1987. Environ. Health Perspect. 84:215-286 Goldman, B.A. 1993. Not Just Prosperity. National Wildlife Federation. Goldman, B.A. 1995. An Environmental Justice Paradigm for Risk Assessment. Environmental Equity Seminar Series. (Sponsored by Duke University, North Carolina Central University, U.S. Environmental Protection Agency) Raleigh, North Carolina. April 17. Goldstein, B. 1989. The maximally exposed individual: an inappropriate basis for public health decisionmaking. Environ. Forum. Nov.-Dec. Gould, J.M. 1986. Quality of Life in American Neighborhoods: Levels of Affluence, Toxic Waste, and Cancer Mortality in Residential Zip Code Areas. Westview Press. Boulder, CO. Harvard Group on Risk Management Reform. 1995. Reform of Risk Regulation: Achieving More Protection at Less Cost. Center for Risk Analysis. Harvard School of Public Health Hoffman, W.S., Adler, H., Fishbein, W.L, and Bauer, F.C. 1967. Relation of pesticide concentration in fat to pathological changes in tissues. Arch. Environ. Health. 15:758-762. Krewski, D., Gaylor, D.W., Soms, A.P., and Szyszkowicz, M. 1993. Correlation Between Carcinogenic Potency and the Maximum Tolerated Dose: Implications for Risk Assessment. Appendix F in NRC (1993) Kutz, F.W., Yobs, A.R., and Strassman, S.C. 1977. Racial stratification of organochlorine insecticide residues in human adipose tissue. J. Occup. Med. 19:619-622. McCallum, D.B. and Santos, S. 1995. Participation and Persuasion. A Communications PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE November 3, 1995 Page R-3 Perspective on Risk Management. In: (Need) Minard. R.. Jr. 1994. A Practitioner's Guide to Comparative Risk--and how we got here. Prepared for the Relative Risk Ranking Roundtable. Resources for the Future. National Academy of Public Administration (NAPA). 1995. Setting Priorities, Getting Results: A New Direction for EPA. Washington, D.C. National Research Council (NRC). 1983. Risk Assessment in the Federal Government: Managing the Process. National Academy Press. Washington. D.C. National Research Council (NRC). 1989. Improving Risk Communication. National Academy Press. Washington, D.C. National Research Council (NRC). 1993. Issues in Risk Assessment. National Academy Press. Washington, DC National Research Council (NRC). 1994. Science and Judgment in Risk Assessment. National Academy Press. Washington, D.C. Northeast Center for Comparative Risk. 1993. State Comparative Risk Projects: A Force for Change. Vermont Law School. South Royalton, VT. Office of Science and Technology Policy (OSTP). 1995. [incomplete] Peto, R., Pike, M.C., Bernstein, L., Gold, L.S., and Ames, B.N. 1984. The TD: A proposed general convention for the numerical description of the carcinogenic potency of chemicals in chronic-exposure animal experiments. Environ. Health Perspect. 58:1-8 President’s Council on Sustainable Development (PCSD). 1995a. January President’s Council on Sustainable Development (PCSD). 1995b. Public comment survey Richards, M. 1993. Siting Industrial Facilities. Lessons from the Social Science Literature. Presented at the Fifth Annual International Conference of the Society for the Advancement of Socio-Economics. Environmental Decision-Making. New York City. March 26-28. Rieth, J.P. and Starr, T.B. 1989a. Chronic bioassays: Relevance to quantitative risk assessment of carcinogens. Reg. Toxicol. Pharmacol. 10:160-173 Rieth, J.P. and Starr, T.B. 1989b. Experimental design constraints on carcinogenic potency PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE November 3, 1995 Page R-4 estimates. J. Toxicol. Environ. Health 27:287-296 Rosenthal. A.. Gray, G.M., and Graham. J.D. 1992. Legislating Acceptable Cancer Risk from Exposure to Toxic Chemicals. Ecology Law Quarterly 19:269-362 Slovic, P., Fishchotf. B.. and Lichtenstein, S. 1980. Facts and fears: Understanding perceived risk. In: R.C. Schwing and W.A. Albers, Jr.. eds. Societal Risk Assessment: How Safe Is Sate Enough? Plenum Press. New York. Spitzer, H. 1995. Peer Review Practices in the Federal Government. Prepared for the American Industrial Health Council. The National Law Journal. 1992. Unequal Protection: The Racial Divide in Environmental Law. Special Supplement. September 21. United Church of Christ Commission for Racial Justice and Public Data Access, Inc. 1987. Toxic Wastes and Race in the United States: A National Report on the Racial and Socio- Economic Characteristics of Communities with Hazardous Waste Sites. New York, New York. United Nations Commission on Environment and Development (UNCED). 1987. Our Common Future. Chaired by Dr. Gro Harlem Bruntdland. U.S. Environmental Protection Agency (EPA). 1987. Unfinished Business: A Comparative Assessment of Environmental Problems. Office of Policy Analysis. Washington, D.C. U.S. Environmental Protection Agency (EPA). 1990. Reducing Risk: Setting Priorities and Strategies For Environmental Protection. Science Advisory Board. SAB-EC-90-0021. Washington, D.C. U.S. Environmental Protection Agency (EPA). 1991. Guidance for Risk Assessment. [incomplete] U.S. Environmental Protection Agency (EPA). 1992. Environmental Equity: Reducing Risk For All Communities. EPA230-R-008. Office of Policy, Planning and Evaluation. Washington, D.C. U.S. Environmental Protection Agency (EPA). 1992a. Framework for Ecological Risk Assessment. EPA/630/R-92/001. Risk Assessment Forum. Washington, DC U.S. Environmental Protection Agency (EPA). 1993. A Review of Ecological Assessment Case Studies from a Risk Assessment Perspective. EPA/630/R-92/003. Risk Assessment PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE November 3, 1995 Page R-5 Forum. Washington, DC U.S. General Accounting Office (GAO). 1983. Siting of Hazardous Waste Landfills and Their Correlation with Racial and Economic Status of Surrounding Communities. Washington, D.C. Government Printing Office. Van Horn, C.E. 1988. Breaking the Environmental Gridlock. The Eagleton Institute of Politics. Rutgers University. New Brunswick, N.J. Vermont Agency of Natural Resources. 1991. Environment 1991: Risks to Vermont and Vermonters. West, P., Fly, J.M., Larkin, F.. and Marans. R.W. 1992. Minority anglers and toxic fish consumption: evidence from a statewide survey of Michigan. In: B. Bryant and P. Mohai, eds. Race and the Incidence of Environmental Hazards. Westview Press. Boulder, CO. Yankelovich, D. 1991. Coming to Public Judgment: Making Democracy Work in a Complex World. Syracuse University Press. Zeise, L., Wilson, R., and Crouch, E. 1984. Use of acute toxicity to estimate carcinogenic risk. Risk Anal. 4:187-199 Zeise, L., Crouch, E.A.C., and Wilson, R. 1985. Reply to comments: On the relationship of toxicity and carcinogenicity. Risk Anal. 5:265-270 Zeise, L., Crouch, E.A.C., and Wilson, R. 1986. A possible relationship between toxicity and carcinogenicity. J. Am. Coll. Toxicol. 5:137-151 Zimmerman, R. 1993. Social equity and environmental risk. Risk Anal. 13:649-666. PRIVILEGED DRAFT—DO NOT QUOTE, CITE, OR DUPLICATE November 3, 1995 Page R-6 Appendix A.1 Mandate of the Commission on Risk Assessment and Risk Management UNITED STATES ENVIRONMENTAL PROTECTION AGENCY ADVISORY COMMITTEE CHARTER RISK ASSESSMENT AND MANAGEMENT COMMISSION 1. PURPOSE. This charter renews the Risk Assessment and Management Commission in accordance with requirements of the Federal Advisory Committee Act, 5 U.S.C. App. 2, §9(c). 2. AUTHORITY. The Commission was specifically directed under Section 303 of the Clean Air Act, as amended on November 15, 1990. 3. OBJECTIVE AND SCOPE OF ACTIVITY. As required by the Clean Air Act Amendments of 1990, the Risk Assessment and Management Commission shall make a full investigation of the policy implications and appropriate uses of risk assessment and risk management in regulatory programs under various Federal laws to prevent cancer and other chronic human health effects which may result from exposure to hazardous substances. The Commission shall consider: (a) The report of the National Academy of Sciences authorized by section 112(0) of the Clean air Act, the use and limitations of risk assessment in establishing emissions and effluent standards, ambient standards, exposure standards, acceptable concentration levels, tolerances or other environmental criteria for hazardous substances that present a risk of carcinogenic effects or other chronic health effects and reductions in the number of persons exposed at various levels of risk, the incidence of cancer, and other public health factors; (b) The most appropriate methods for measuring and describing cancer risks or risks of other chronic health effects from exposure to hazardous substances considering such alternative approaches as the lifetime risk of cancer or other effects to the individual or individuals most exposed to emissions from a source or sources on both an actual and worst case basis, the range of such risks, the total number of health effects avoided by exposures standards, acceptable concentration levels, tolerances and other environmental criteria, reductions in the number of persons exposed at various levels of risk, the incidence of cancer, and other public health factors; (c) Methods to reflect uncertainties in measurement and estimation techniques, the existence of synergistic or antagonistic effects among hazardous substances, the accuracy of extrapolating human health risks from animal exposure data, and the Al-1 ADVISORY COMMITTEE CHARTER existence of unquantified direct or indirect effects on human health in risk assessment studies; (da) Risk management policy issues including the use of lifetime cancer risks to individuals most exposed, incidence of cancer, the cost and technical feasibility of exposure reduction measures and the use of site specific actual exposure information in setting emissions standards and other limitations applicable to sources of exposure to hazardous substances; and (e) Comment on the degree to which it is possible or desirable to develop a consistent standard of acceptable risk, among various Federal programs. 4. FUNCTIONS. (a) In the conduct of the studies required by this section, the Commission is authorized to contract (in accordance with Federal contract law) with nongovernmental entities that are competent to perform research or investigations within the Commission's mandate, and to hold public hearings, forums, and workshops to enable full public participation. (b) The Commission may appoint and fix the pay of such staff as it deems necessary in accordance with the provisions of title 5, United States code. The Commission may request the temporary assignment of personnel from the Environmental Protection Agency or other Federal agencies. (c) The members of the Commission who are not officers or employees of the United States, while attending conferences or meetings of the Commission or while otherwise serving at the request of the Chair, shall be entitled to receive compensation at a rate not in excess of the maximum rate of pay for Grade GS 18, as provided in the General Schedule under section 5332 of title 5 of the United States Code, including travel time, and while away from their homes or regular places of business they may be allowed travel expenses, including per diem in lieu of subsistence as authorized by law for persons in the Government service employed intermittently. (d) A report containing the results of all Commission studies and investigations under this section, together with any appropriate legislative recommendations or administrative recommendations, shall be made available to the public for comment not later than 42 months after the date of enactment of the Clean Air Act Amendments of 1990 and shall be submitted to the President and to the Congress not later than 48 months after such date of enactment. In the report, the Commission shall make recommendations with respect to the appropriate use of risk assessment and risk management in Federal regulatory programs to prevent cancer or other chronic health effects which may result from exposure to hazardous substances. Al-2 ADVISORY COMMITTEE CHARTER 5. COMPOSITION AND MEETINGS. The Commission shall be composed of ten members who shall have knowledge or experience in fields of risk assessment or risk management, including three members to be appointed by the President, two members to be appointed by the Speaker of the House of Representatives, one member to be appointed by the minority Leader of the House of Representatives, two members to be appointed by the Majority Leader of the Senate, one member to be appointed by the Minority leader of the Senate, and one member to be appointed by the President of the National Academy of Sciences. Meetings will be held as necessary. A full-time employee of the Environmental Protection Agency has been assigned as the Designated Federal cfficer, who will be present at all meetings and is authorized to adjourn any meeting whenever it is determined to be in the public interest. The estimated annual operating cost of the Commission for FY94 was approximately $48,976.38, which includes .35 FTE work year of staff support. This figure will increase in FY95 once the Commission hires it's staff, meets on a monthly basis for a year, obtains office space, etc. The Office of the Administrator oversees and executes the budget assigned to the Commission and the Office of Air provides administrative support as provided by the Clean Air Act Amendments of 1990. 6. DURATION. The Commission shall cease to exist upon the date determined by the Commission, but not later than 9 months after the submission of such report. NOV 14 1994 ved) aan Agency Approval Date Deputy Administrator Nov | 5 1994 Date Filed with Congress Al-3