Science and Decisions: Advancing Risk Assessment (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

1 Introduction In response to a request from the Environmental Protection Agency (EPA) National Center for Environmental Assessment (NCEA), the National Research Council established the Committee on Improving Risk Analysis Approaches Used by the EPA. The committee was charged with developing recommendations that, if implemented, could assist the agency in developing risk assessments that are both consistent with current and evolving scientific understanding and relevant to the many risk-management missions of the agency. Recom- mendations were to focus on both short- and long-term objectives. The importance of risk assessment to the mission of EPA—indeed to the mission of many other federal agencies and to their state counterparts—is attested to by a long series of major efforts by the National Academies and other expert bodies to strengthen the tech- nical content and utility of risk assessment and to ensure its scientific integrity. As EPA has attempted to respond to the recommendations that have resulted from the various efforts, both the science underlying risk assessment and the decision contexts in which risk assess- ments are used have been increasingly complex. As will be revealed later in this report, the committee perceives that risk assessment is now at a crossroads and its value and relevance are increasingly questioned (Silbergeld 1993; Montague 2004). Nonetheless, the commit- tee believes strongly that risk assessment remains the most appropriate available method for measuring the relative benefits of the many possible interventions available to improve human health and the environment and that its absence or its inappropriate application will result in seriously flawed decisions. The committee believes that implementation of the recommendations set forth in this report will do much to enhance the power and usefulness of risk assessment and will be the appropriate road forward.   EPA’s charge to the committee used the phrase risk analysis. The latter is sometimes used synonymously with risk assessment but sometimes used more broadly. The committee will use risk assessment to describe the process leading to a characterization of risk. Risk as defined by NRC (2007a) can be a hazard, a probability, a consequence, or a combination of probability and severity of consequence. 15 

16 SCIENCE AND DECISIONS: ADVANCING RISK ASSESSMENT BACKGROUND Since the 1983 publication of the National Research Council’s report Risk Assessment in the Federal Government: Managing the Process (the so-called Red Book), EPA has made efforts to advance risk assessment with the generation of risk-assessment guidelines, the establishment of intra-agency and cross-agency science-policy panels, and improvements in peer-review standards for agency risk assessments. The Red Book committee demonstrated how risk assessment could fill the gap between results emerging from the research setting and their use in risk management. A framework for systematically carrying out the process of risk assessment was established, and the Red Book’s risk-assessment framework remains in place today. The Red Book also revealed how the development of what were called infer- ence guidelines (see below) was necessary to ensure the scientific integrity of the process by which risk assessments were conducted and of the product of that process. Various closely related forms of the risk-assessment framework have been widely used by international organizations and other federal agencies, including the Consumer Product Safe- ty Commission, the Nuclear Regulatory Commission, the Food and Drug Administration, the Occupational Safety and Health Administration, the U.S. Department of Agriculture, the Department of Defense, and the Department of Energy. OSTP (50 Fed. Reg. 10371[1985]) adopted the Red Book framework for carcinogen analysis and provided agencies a basis for developing the guidelines recommended by NRC (1983). Publication of the Red Book was followed by an intensification of risk-assessment activ- ity in EPA. EPA endorsed the Red Book in the publication, Risk Assessment and Manage- ment: Framework for Decision Making (EPA 1984). The agency established in 1984 what is now called the Risk Assessment Forum and in 1993 added a Science Policy Council (see Appendix C for a timeline of selected risk-assessment activities)—evidence that the Red Book and EPA’s efforts to advance risk assessment fell on fertile ground (Goldman 2003). William Ruckelshaus, during his second tour as administrator of EPA (1983-1985), used the Red Book as the basis of a main theme of his tenure: strengthening risk assessment as a tool to inform decision-making. EPA initially focused on human health risk assessment with the Guidelines for Carcinogen Risk Assessment (EPA 1986) and the agency’s Unfinished Busi- ness: A Comparative Assessment of Environmental Problems (EPA 1987), which compared the magnitude of environmental risks with EPA’s resource allocations to programs that ad- dress them. The agency’s Science Advisory Board evaluated the latter document in another key report, Reducing Risk: Setting Priorities and Strategies for Environmental Protection (EPA SAB 1990), and EPA was involved in a 1992 conference that evaluated the risk-based model for setting national priorities against several alternatives that incorporated informa- tion about solutions, environmental justice, and other factors (Finkel and Golding 1994). In the 1990s, the four-step approach outlined in the Red Book was adapted to ecologic risk assessment to address evaluations in which human health is not the primary focus (EPA 2004). Ecologic risk assessors pioneered new approaches to complex risk problems by delin- eating the need for “planning and problem formulation” to address technically challenging assessments of ecosystems, chemical mixtures, and cumulative risk. In the planning step, the risk managers—in consultation with risk assessors and other interested parties—frame management goals, management options, and the scope and necessary level of complexity for the risk assessment. Problem formulation is the phase in which the risk managers’ charge to the assessors is converted into an actionable plan for performing the assessment (EPA 1998; Suter 2007). Several National Research Council and other expert panels expanded on the risk- assessment principles presented in the Red Book with the publication of reports that included 

INTRODUCTION 17 Pesticides in the Diets of Infants and Children (NRC 1993), Science and Judgment in Risk Assessment (NRC 1994), and Understanding Risk: Informing Decisions in a Democratic Society (NRC 1996). In 1997, another expert panel issued its report, Presidential/Congres- sional Commission on Risk Assessment and Risk Management (PCCRARM 1997). EPA has also recently upgraded its standards for peer review of technical documents with the Science Policy Council’s Peer Review Handbook (EPA 2000) and guidance (EPA 2002) to conform with the Office of Management and Budget’s Final Information Quality Bulletin for Peer Review (OMB 2004). CHALLENGES As risk assessment has come to be widely used in a fairly consistent framework, EPA practices have continued to draw scrutiny in that competing pressures are pushing the agency to improve the timeliness and quality of its risk assessments. It is now evident that many risk assessments are taking 10-20 years to complete including assessments on chemicals such as dioxin, formaldehyde, and trichloroethylene (GAO 2008). There are a myriad of reasons for delays in the completion of risk assessments including controversy surrounding the sci- ence, uncertainties in the data, regulatory requirements, political priorities, and economic factors. In the absence of completed risk assessments, risk management decisions continue to be made by state and federal agencies; however it is not known whether the decisions being made are health protective. To the extent that this practice continues, the value of risk assessment will erode. For example, trichloroethylene, the most common organic contaminant in groundwater, which has been linked to cancer, does not have a completed EPA toxicity assessment. The EPA assessment has been under development since the 1980s and subjected to multiple independent reviews including EPA’s Science Advisory Board. Key issues were evaluated by the National Research Council in 2006. NRC (2006) urged that the toxicity assessment be finalized with currently available data, but the assessment is not anticipated to be final- ized until 2010 (GAO 2008). Another example is formaldehyde, which the World Health Organization classified as a known human carcinogen and whose assessment was begun by EPA in 1997 but is not expected to be completed until 2010 (IARC 2006). The lack of an updated toxicity assessment for formaldehyde has impacted EPA’s regulatory decisions (GAO 2008). In recent years, a number of federal agencies have raised concerns about EPA risk as- sessments of contaminants and are now playing a more formal role in risk policy-making at the federal level. Some of the agencies are also potentially responsible parties facing cleanup responsibilities and are seeking more input as EPA moves toward final reviews. Those other agencies and other public and private stakeholders often assert that they are inadequately involved in EPA processes (for example, Risk Policy Report 2005, 2007). The Integrated Risk Information System (IRIS) is an important compendium of chemical toxicity values in which new EPA science policies are often implemented for the first time. However, IRIS has been criticized because of limitations, including a lack of funding and delays in updating toxicity values. EPA is now seeking greater science-policy input on its chemical reviews earlier in the process so that critical issues can be identified and adjustments made in response to new scientific and science-policy information.   AO G (2008) acknowledges that because there was no updated EPA cancer risk estimate, EPA’s Office of Air and Radiation used an alternative estimate in establishing a National Emissions Standard for Hazardous Air Pollutants covering facilities in the plywood and composite wood industries. 

18 SCIENCE AND DECISIONS: ADVANCING RISK ASSESSMENT Those types of problems are exacerbated by the fact that the scientific issues underly- ing risk assessments and the decisions that risk assessments are developed to support are increasingly complex, as a result of a greater quantity and diversity of data stemming from advancements such as in genomics and biomarkers. This report is intended to assist EPA as it attempts to deal with those and other challenges. TRADITIONAL AND EMERGING VIEWS OF THE ROLES OF RISK ASSESSMENT A large community of public-health research scientists in many disciplines is involved in the development of knowledge about how agents in the environment—whether chemi- cal, biologic, radiologic, or physical and whether of natural origin or resulting from human activity—can harm human health and about the conditions under which they may do so. As this type of knowledge emerges from research, policy-makers in government and many other institutions concerned with public health begin to focus on whether some type of action is needed to protect public health and, if so, whether some courses of action yield better results than others. Societal support for action is found in the many laws that guide regulatory and public-health agencies. This support is evident in the relationship between the research community concerned with understanding threats to ecosystems and people responsible for protecting them. It is clear that research findings are rarely directly suitable for decision-making. Results of different studies of the same phenomena often conflict, uncertainties can be large, and the conditions under which health and ecosystem threats are studied (or can be studied) usually do not match the conditions of interest for public-health or ecosystem protection. Research findings need to be interpreted. In matters related to public and ecosystem health, the interpretive process is called risk assessment. Risk assessment has come to be seen as an essential component of regulatory and related types of decision-making, and its scientific underpinnings and its roles in decision-making are the central subjects of this report. Much scholarly work that has appeared since the publication of the Red Book has been devoted to countering a tendency to view risk assessment, in its practical applications, both as the sole source of information on the problems to be managed and as providing the management choice. To the extent that that tendency exists, we urge that it be resisted. Risk assessment, we propose, should certainly continue to capture and accurately describe what various bodies of research findings do and do not tell us about various threats to hu- man health and to the environment, but it should do so only after the questions that risk assessment is supposed to address have been posed, through careful evaluation of the options available to manage the environmental problem at hand, similar to what is done in ecologic risk assessment. In this context, risk assessment is seen as a method for evaluating the rela- tive merits of various options (or interventions) for managing risk. Risk assessment, in that decision-making context, is an essential tool for understanding what public-health and environmental goals can be achieved or have been achieved by the actions taken. As will be seen later in this report, early emphasis on identifying risk-man- agement options and on seeking, through risk assessment, analyses that are most useful for evaluating the options is somewhat at variance with the risk-assessment–risk-management model first proposed in the Red Book in that the management options are no longer driven by whatever risk-assessment findings happen to emerge. The new model does not alter the technical content of risk assessment from that set out in the Red Book, and, if appropriate precautions are taken, it does not lead to inappropriate intrusions by risk managers into the risk-assessment process (an issue of much concern to the Red Book authors; see Chapter 2). But it has great potential to increase the influence of risk assessment on ultimate decisions 

INTRODUCTION 19 because it is asked to cast light on a wider range of decision options than has traditionally been the case. We see this as a necessary and worthwhile extension of the Red Book model, one better suited to today’s challenges. Its full scope is elucidated in Chapters 3 and 8, which focus on increasing the utility of risk assessments. Regulatory decision-makers, including those in EPA, do not routinely approach public- health and environmental problems by arraying a wide range of options for dealing with them and then setting into motion the various technical analyses (risk assessments, control- technology analyses, analyses of resource costs, and so on) that are necessary to achieve the optimal outcome. The various laws administered by EPA and other regulatory agencies appear to constrain, or have traditionally been interpreted as constraining, the options to be considered for risk management. The broader decision context that we propose (discussed in Chapters 3 and 8) recommends the consideration of other tools now being used or under development (such as life-cycle analysis [LCA] and sustainability evaluation) that are directed at environment-related problems of broader scope than those traditionally considered by EPA and related institutions. The integration of the scientific power of risk assessment with the broader reach of LCA, for example, should enlarge the influence of risk assessment and increase its utility for managing the most urgent and far-reaching problems—those having both human and environmental health components. Whether operating in a broad or more narrowly constrained decision context, risk as- sessment is essential for the reasons described above. Whatever the decision context, the goal of risk assessment is to describe the probability that adverse health or ecosystem effects of specific types will occur under specified conditions of exposure to an activity or an agent (chemical, biologic, radiologic, or physical), to describe the uncertainty in the probability estimate, and to describe how risk varies among populations. To be most useful in decision- making, risk assessment would consider the risks associated with existing conditions (that is, the probability of harm under the “take no action” alternative) and the risks that would remain if each of various possible actions were taken to alter the conditions. There would also be a need for some commonality in the uncertainty analysis goals and assumptions that are applied to each of the analyses so that the different policy options can be compared. The conduct of risk assessment in the broadest practicable risk-management context brings to light the fullest possible picture of net public-health and environmental benefits. That does not mean that other options cannot surface during the conduct of a risk assessment; in fact, improved stakeholder engagement in the process may make this possible. Achieving such results requires the use of the framework for the conduct of risk as- sessment set forth in the 1983 Red Book, which has been adopted by numerous expert committees, regulatory agencies, and public-health institutions and which this committee sees no reason to alter. The framework includes three well-known analytic steps—hazard identification, dose-response assessment, and exposure assessment—and a fourth step, risk characterization, in which results of the first three steps are integrated to yield information on the probability that the adverse effects described in hazard identification will occur under the conditions described in exposure assessment. Uncertainty findings from the first three steps are also integrated into risk characterization. Many other types of review of human-health or ecologic data emerge from regulatory and public-health institutions, but only those which in some way incorporate all four of the above steps can properly be called risk assessments. Although all risk assessments include the four steps, it is critical to recognize that risk assessments can be undertaken at various levels of technical detail. Given a sufficiently rich database, highly quantitative estimates of risk can be developed, sometimes involving probabilistic modeling and substantial biologic data. In other cases, risk assessments may be semiquantitative. Similarly, descriptions of the uncertainties inherent in all risk assessments 

20 SCIENCE AND DECISIONS: ADVANCING RISK ASSESSMENT may be complex or relatively simple. Because risk assessments can vary in detail and com- plexity, it is important to know how a risk assessment will be used before it is undertaken so that it can be designed and carried out at the level of technical detail appropriate to the problem at hand. Risk-assessment design is the subject of Chapter 3. Decisions regarding risks and risk changes expected under various risk-management options are informed by the availability of risk assessments. The goal of achieving accurate, highly quantitative estimates of risk, however, is hampered by limitations in scientific under- standing and the availability of relevant data, which can be overcome only by the advance of relevant research. Decisions to protect public health and the environment cannot await “perfection” in scientific knowledge (an unachievable goal in any case); in the absence of the understanding that risk assessments, however imperfect, can bring, it will not be possible to know the public-health or environmental value of whatever decisions are ultimately made. It is therefore important that risk assessments incorporate the best available scientific informa- tion in scientifically rigorous ways and that they capture and describe the uncertainties in the information in ways that are useful for decision-makers. Moreover, the goal of timeliness is as important as (sometimes more important than) the goal of a precise risk estimate. The need to seek improvements in EPA’s regulatory decision-making by improving the quality and utility of risk assessment is the impetus for the current study. TECHNICAL IMPEDIMENTS TO RISK ASSESSMENT It is useful to describe some of the types of obstacles that hamper the risk-assessment process and that limit the utility of its results. It should be kept in mind that risk assess- ments should not be blamed for a lack of relevant scientific data and knowledge; such a lack reflects inadequate support for research. But inadequacies in the use of whatever data and knowledge are available clearly are a problem for risk assessment. The following questions will receive much attention in this report because they reflect identifiable impediments to risk assessment and its most important use—for informing decision-making. 1. Are the decision contexts in which risk assessments are to be developed well defined in advance? It is important to understand the context in which a risk assessment will be used, so that the appropriate options for addressing a problem can be considered. It seems that current regulatory thinking on this matter may be overconstrained and often fails even to begin to incorporate a full range of decision options, perhaps because of limitations, or perceived limitations, embodied in laws. In any event, the utility of risk assessments may be less than ideal because of a failure to achieve clarity regarding the options for decision- making in advance of identifying the types of risk assessments that will be of value. 2. What is the right level of detail for a risk assessment? Early delineation of problems and options for managing them allows—through the necessary interactions among risk man- agers, risk assessors and other technical analysts, and other stakeholders—the development of risk assessments whose level of detail and scientific completeness match the decision-mak- ing requirements and so can maximize the efficiency of the process. 3. Are the criteria for selecting the “defaults” necessary to complete risk assessments and for departing from them fully specified and set forth in agency guidelines? Because of the need for a variety of inferences in risk assessment and because the rationales for draw- ing the inferences are not always distinguishable on purely scientific grounds, the choice of   The committee recognizes that the current EPA policy on defaults uses the term “invokes” rather than “de- parts.” EPA’s current policy on defaults is presented in Chapter 6. 

INTRODUCTION 21 default options to be used involves an element of policy (see the discussion of the Red Book in Chapter 2). The inferences selected, which are commonly referred to as defaults, can have substantial effects on the results of risk assessments. Their selection and the criteria for judging when, in a specific case, a default can be replaced with an alternative inference based on chemical-specific information are among the most contentious elements of the risk-assessment process and a cause of sometimes great delays in their completion. 4. Are the best available scientific information and defaults used to deal with the prob- lem of variability? Variability in exposures to hazardous agents and in biologic responses to them is a fact of nature. Scientific knowledge of variability is highly limited, and current risk-assessment approaches to the problem rely heavily on uncertainty factors and other as- sumptions. It is important for the advance of risk assessment to consider the types of scientific knowledge now available and their use for improving the quantitative characterization of variability. 5. What methods should be used to describe and express the uncertainties that accom- pany all risk assessments? Failure to deal adequately with this matter is a source of much contention and hampers the goals of decision-making. An issue of central concern is the rela- tive utility for decision-makers of the various methods available to express uncertainties. 6. Is information about the hazardous properties of chemicals and other agents given adequate attention in risk assessment? The toxic or carcinogenic properties of substances under assessment are now typically described in qualitative terms (a weight-of-evidence evaluation), and without quantitative expressions of the probability that the adverse effect is relevant to the human population that is the subject of the risk assessment. The possible importance of this limitation in risk assessment has been little discussed. 7. Are current methods for dealing with substances thought to act through threshold mechanisms (for example, the development of toxicity reference doses) yielding the most useful information for decision-making? Current “bright-line” approaches, while valuable in certain public-health decision-making contexts, clearly lack utility in other contexts. 8. Do current methods for integrating and weighing evidence from different sources (for example, from epidemiology and experimental studies) ensure that subjective influences are minimized and transparency maximized? 9. What are the appropriate scientific and policy approaches for dealing with sub- stances on which very little health-effects or exposure information is available so that the risks they pose are not ignored relative to those posed by better-studied substances? 10. What approaches should be pursued for defining the risk assessments necessary to address broad questions of communitywide and cumulative risks (which may involve many exposure sources and pathways)? Given an ability to formulate appropriate risk questions in such broad contexts, how can risk information best serve decisions needed to reduce burdens on public health and the environment? IMPROVING RISK ANALYSIS Based on the above questions, improvements in risk analysis can be considered at two broad levels. First, consideration can be given to improvements in the utility of risk assess- ments for decision-making. Second, improvements in the technical analysis supporting one or more of the steps of risk assessment can also be feasible, as new scientific knowledge becomes available. The committee understands its charge to encompass both types of improvements. Improved utility can be achieved in several ways. As has been noted, there are op- portunities to improve the processes through which risk-related problems and options for 

22 SCIENCE AND DECISIONS: ADVANCING RISK ASSESSMENT intervention are identified and formulated prior to the development of risk assessments. Similar opportunities arise for improvements in the interactions among risk managers and other stakeholders and risk assessors during the development of assessments. Utility might also be enhanced by improvements in the ways risks are characterized and uncertainties expressed, to ensure they are adequately understood by decision-makers. Can the public health be better served in certain circumstances, for example, by probabilistic expressions of risk and uncertainties, for toxicity information, than they are by “bright line” estimates such as toxicity reference doses and concentrations? Can assessments in which the results of applying different default options are presented, each with a description of its scientific strengths and weaknesses, better serve decision-makers than those that rely primarily upon pre-assigned defaults? These types of questions pertain to improvements that might increase the utility of risk assessments for decision-making. Improving the technical analysis involved in each of the steps of risk assessment gener- ally refers to the development and use of scientific knowledge and information that, for a number of reasons, might lead to more accurate characterizations of risk. Because there are generally no means empirically to verify the results of most risk assessments, it is difficult to assess whether “accuracy” has been improved. But there nevertheless seems to be a basis for believing that greater understanding of the biological processes underlying the production of toxicity or other types of adverse health effects can, if properly applied, increase confidence in risk-assessment results. Indeed, much of the current research in toxicology is directed at gaining that understanding, and with that understanding can come reduced reliance upon de- faults. In addition, the development of databases of empirical observations relevant to specific uncertainty factors can be used to replace single-point uncertainty factors with distributions. Increased confidence in risk assessments might also arise from increased development and use of human data—both epidemiology and in vitro data (NRC 2007b). It should be noted that, while improvements in the utility of risk analysis are always desirable, the quest for improvements in scientific accuracy may not always be necessary or desirable in the context of specific risk assessments. The latter usually requires investment in significant research, and so will necessarily be limited to substances of significant social or economic importance. Default-based risk assessments will continue to have significant roles because decisions must be efficiently made on large numbers of hazards for which resources will not be available to corroborate the validity of each default, or to explore specific alternatives, and because as experience accrues, many of the defaults are viewed as a culmination of scientific understanding about general phenomena (for which exceptions may apply in particular cases). It is, of course, possible that, as new scientific understanding becomes available, certain alternatives to established defaults may prove to be supportable on a general basis, and this would increase confidence in risk assessments based on them. But default-based risk assessments will remain necessary for many substances and situations. Much of what follows in the remaining chapters of the report derives from the com- mittee’s view of these two broad ways in which improvements in risk analysis might be achieved. THE NATIONAL RESEARCH COUNCIL COMMITTEE In response to the study request from EPA, the NRC established the Committee on Im- proving Risk Analysis Approaches Used by EPA. Committee members were selected for their expertise in biostatistics, dose-response modeling, ecotoxicology, environmental transport and fate modeling, environmental health, environmental regulation, epidemiology, exposure assessment, risk assessment, toxicology, and uncertainty analysis. Members come from uni- 

INTRODUCTION 23 versities and other organizations and serve pro bono. Committee members were asked to serve as individual experts, not as representatives of any organization. The committee was charged with developing scientific and technical recommendations for improving risk analysis approaches used by EPA, including providing practical improve- ments that EPA could make in the near term (2-5 years) and in the longer term (10-20 years). The committee focused primarily on human health risk assessment, but considered the implications of its findings and recommendations to ecological risk analysis. In reviewing EPA’s risk analysis concepts and practices, the committee considered past evaluations and ongoing studies by NRC and others, and risk analyses involving different exposure pathways and environmental media. In its evaluation, the committee was asked to consider a number of topics relating to uncertainty, variability, modeling, and mode of action (see Appendix B for complete statement of task). To address its task, the committee held five public sessions in which it heard presenta- tions from officials from EPA’s Office of Research and Development, its policy, program and regional offices; the Centers for Disease Control and Prevention; representatives from industry and environmental organizations; consultants; and academia. In addressing its charge, the committee considered carefully the concerns expressed by the presenters regarding the challenges and limitations of risk assessment (Callahan 2007; Kyle 2007). Peter Preuss, the director of NCEA urged the committee to consider three specific questions (Preuss 2006): 1) What improvements can be made to risk assessment in the present? 2) What improvements can be made to risk assessment in the longer term? 3) What alternative risk paradigms should be considered? Although the charge is focused on risk assessment at EPA, it is the committee’s hope that the recommendations have influence over risk assessment wherever it is practiced and used. ORGANIZATION OF THE REPORT The body of this report is organized into nine chapters. Chapter 2 presents an evolution of risk assessment and its applications since the 1980s. Chapter 3 addresses the design of risk assessment, emphasizing the role of planning and scoping and problem formulation in the process. Chapter 4 considers uncertainty and variability in risk assessment, addressing both EPA’s methodologies and needs for improvement. Chapter 5 presents a unified approach for non-cancer and cancer dose-response modeling that explicitly incorporates uncertainty and variability into the process. Chapter 6 addresses an important area of uncertainty, selection and use of defaults. Chapter 7 discusses the need and methods for considering a broader range of factors in risk assessment, that is cumulative risk assessment, including chemical and non-chemical stressors, vulnerability of the exposed population, and the impact of actions on stakeholders, in particular communities. Chapter 8 presents a framework for risk-based decision-making that is intended to improve the utility of risk assessment. Chapter 9 presents the committee’s conclusions and recommendations along with a strategy for implementing them.   description of observable key events or processes from interaction of an agent with a cell or tissue through A operational and anatomical changes to the disease state (EPA 2005). 

24 SCIENCE AND DECISIONS: ADVANCING RISK ASSESSMENT REFERENCES Callahan, M.A. 2007. Improving Risk Assessment: A Regional Perspective. Presentation at the Third Meeting of Improving Risk Analysis Approaches Used by EPA, February 26, 2007, Washington, DC. EPA (U.S. Environmental Protection Agency). 1984. Risk Assessment and Risk Management: Framework for Decision Making. EPA 600/9-85-002. Office of the Administrator, U.S. Environmental Protection Agency, Washington, DC. December 1984. EPA (U.S. Environmental Protection Agency). 1986. Guidelines for Carcinogen Risk Assessment. EPA/630/R- 00/004. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. September 1986 [online]. Available: http://www.epa.gov/ncea/raf/car2sab/guidelines_1986.pdf [accessed Jan. 7, 2007]. EPA (U.S. Environmental Protection Agency). 1987. Unfinished Business: A Comparative Assessment of Envi- ronmental Problems Overview Report. EPA 230287025a. Office of Policy, Planning and Evaluation, U.S. Environmental Protection Agency, Washington, DC. February 1987. EPA (U.S. Environmental Protection Agency). 1998. Guidelines for Ecological Risk Assessment. EPA/630/R- 95/002F. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. April 1998 [online]. Available: http://oaspub.epa.gov/eims/eimscomm.getfile?p_download_id=36512 [accessed Feb. 9, 2007]. EPA (U.S. Environmental Protection Agency). 2000. Science Policy Peer Review Handbook, 2nd Ed. EPA 100-B- 00-001. Office of Science Policy, Office of Research and Development, U.S. Environmental Protection Agency, Washington, DC. December 2000 [online]. Available: http://www.epa.gov/osa/spc/pdfs/prhandbk.pdf [accessed Feb. 9, 2007]. EPA (U.S. Environmental Protection Agency). 2002. Guidelines for Ensuring and Maximizing the Quality, Utility and Integrity of Information Disseminated by the Environmental Protection Agency. EPA/260R-02-008. Office of Environmental Information, U.S. Environmental Protection Agency, Washington, DC. October 2002 [on- line]. Available: http://www.epa.gov/QUALITY/informationguidelines/documents/EPA_InfoQualityGuidelines. pdf [accessed Feb. 9, 2007]. EPA (U.S. Environmental Protection Agency). 2004. Risk Assessment Principles and Practices: Staff Paper. EPA/100/ B-04/001. Office of the Science Advisor, U.S. Environmental Protection Agency, Washington, DC. March 2004 [online]. Available: http://www.epa.gov/osa/pdfs/ratf-final.pdf [accessed Feb. 9, 2007]. EPA (U.S. Environmental Protection Agency). 2005. Guidelines for Carcinogen Risk Assessment. EPA/630/P- 03/001F. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington, DC. March 2005 [online]. Available: http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=116283 [accessed Feb. 7, 2007]. EPA SAB (U.S. Environmental Protection Agency Science Advisory Board). 1990. Risk Assessment: Setting Priori- ties and Strategies for Environmental Protection. EPA/SAB-EC-90-021. U.S. Environmental Protection Agency Science Advisory Board, Washington, DC. Finkel, A.M. and D. Golding, eds. 1994. Worst Things First? The Debate Over Risk-Based National Environmental Priorities. Washington, DC: Resources for the Future Press. GAO (U.S. General Accountability Office). 2008. Chemical Assessments: Low Productivity and New Interagency Review Process Limit the Usefulness and Credibility of EPA’s Integrated Risk Information System. GAO-08- 440. U.S. General Accountability Office, Washington, DC. March 2008 [online]. Available: http://www.gao. gov/new.items/d08440.pdf [accessed June 11, 2008]. Goldman, L.R. 2003. The Red Book: A reassessment of risk assessment. Hum. Ecol. Risk Assess. 9(5):1273-1281. IARC (International Agency for Research on Cancer). 2006. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. Volume 88. Formaldehyde, 2-Butoxyehtanol and 1-tert-Butoxypropan-2-ol. Lyon: Interna- tional Agency for Research on Cancer Press. Kyle, A. 2007. Community Needs for Assessment of Environmental Problems. Presentation at the Fourth Meeting of Improving Risk Analysis Approaches Used by EPA, April 17, 2007, Washington, DC. Montague, P. 2004. Reducing the harms associated with risk assessment. Environ. Impact. Asses. Rev. 24:733-748. NRC (National Research Council). 1983. Risk Assessment in the Federal Government: Managing the Process. Washington, DC: National Academy Press. NRC (National Research Council). 1993. Pesticides in the Diets of Infants and Children. Washington, DC: National Academy Press. NRC (National Research Council). 1994. Science and Judgment in Risk Assessment. Washington, DC: National Academy Press. NRC (National Research Council). 1996. Understanding Risk: Informing Decisions in a Democratic Society. Wash- ington, DC: National Academy Press. 

INTRODUCTION 25 NRC (National Research Council). 2006. Assessing the Human Health Risks of Trichloroethylene: Key Scientific Issues. Washington, DC: National Academies Press. NRC (National Research Council). 2007a. Scientific Review of the Proposed Risk Assessment Bulletin from the Office of Management and Budget. Washington, DC: National Academies Press. NRC (National Research Council). 2007b. Toxicity Testing in the Twenty-First Century: A Vision and a Strategy. Washington, DC: National Academies Press. OMB (Office of Management and Budget). 2004. Final Information Quality Bulletin for Peer Review. December 15, 2004 [online]. Available: http://cio.energy.gov/documents/OMB_Final_Info_Quality_Bulletin_for_peer_ bulletin(2).pdf [accessed Jan. 4, 2007]. PCCRARM (Presidential/Congressional Commission on Risk Assessment and Risk Management). 1997. Frame- work for Environmental Health Management-Final Report, Vol. 1 [online]. Available: http://www.riskworld. com/nreports/1997/risk-rpt/pdf/EPAJAN.PDF [accessed Jan. 7, 2008]. Preuss, P. 2006. Human Health Risk Assessment at EPA: Background, Current Practice, Future Directions. Pre- sentation at the First Meeting of Improving Risk Analysis Approaches Used By the U.S. EPA, November 20, 2006, Washington, DC. Risk Policy Report. 2005. EPA Plan for Expanded Risk Reviews Draws Staff Criticism, DOD Backing. Inside EPA’s Risk Policy Report 12(17):1, 8. May 3, 2005. Risk Policy Report. 2007. GAO Inquiry Prepares to Focus on EPA Delay of New Risk Review Plan. Inside EPA’s Risk Policy Report 14(38):1, 6. September 18, 2007. Silbergeld, E.K. 1993. Risk assessment: The perspective and experience of U.S. environmentalists. Environ. Health Perspect. 101(2):100-104. Suter, G.W. 2007. Ecological Risk Assessment, 2nd Ed. Boca Raton, FL: CRC Press.