SRA / Society for Risk Analysis / SRA 2007 Annual Meeting

M5
For Your Eyes Only: Poster Reception

Room: Marriott Riverwalk Ballroom 6:00 - 8:00 PM

M5.1 Comparison of derivation methods for chronic oral toxicity values. Weinrich A*, Lin I, Crawford D, Reid J; US Environmental Protection Agency - National Center for Environmental Assessment, Cincinnati OH weinrich.alan@epa.gov

Abstract: This presentation describes and compares methods used to derive chronic oral toxicity values for the EPA Integrated Risk Information System (IRIS) and Provisional Peer-Reviewed Toxicity Values (PPRTV) with those used to derive US Agency for Toxic Substances and Disease Registry (ATSDR) oral minimal risk levels (MRL), and California EPA drinking water public health goals (PHG). The presentation discusses EPA Health Effects Assessment Summary Tables (HEAST) only briefly, because they are being replaced by PPRTVs. Only minor differences in values for sample chemicals resulted from differences in derivation methods used by EPA, ATSDR, and CalEPA. Most notable differences between derivations for sample chemicals involved data available and methods used at the time of the assessment, as well as differences in scientific judgment including selection of uncertainty factors and choice of key studies and critical toxicity endpoints.

M5.2 Risk assessment variability across state programs: Implications for risk management. Frantzen KA*, Meyer AK, Hadley P; Consulting; Federal Government, State Government kfrantzen@kleinfelder.com

Abstract: The Interstate Technology and Regulatory Council (ITRC) is a state-led coalition working with industry and stakeholders to achieve regulatory acceptance of environmental technologies and technology-related initiatives. Since 2003, the ITRC’s Risk Assessment Resources (Risk) Team has investigated the use of risk assessment in remediation decision-making. In 2005, the Risk Team published a report that documented the similarities and differences among state approaches to the development of risk-based soil screening criteria. Currently, the Risk Team is investigating the variability among states and state programs concerning how risk assessments are conducted and how this relates to the application of risk-based clean-up criteria. The resulting documentation demonstrates a range of potential risk management (remediation) decision outcomes resulting from this variability. This analysis is following a case study approach using actual site cleanups, surveys of state regulators on key questions related to site characterization and risk assessment, and through hypothetical site case studies that compare the range of answers and approaches to the same data set. The results of the analysis document that variability among states and state programs can lead to a significant range of risk management (remediation) outcomes.

M5.4 Lead in Drinking Water, Washington DC: A Case Study in Environmental Health Risk Management. Guidotti TL*, Moses M, Goldsmith DF; George Washington University eohtlg@gwumc.edu

Abstract: In 2001, following a change in disinfection agent in anticipation of the EPA Disinfection Byproduct Rule, lead levels began rising in drinking water in Washington, DC, and in 2002 the DC Water and Sewer Authority was found to have exceeded the EPA lead action level, requiring compliance with a series of measures under the Lead and Copper Rule. In 2004, the issue became a public concern, drawing considerable media attention. The problem was eventually resolved through the application of orthophosphate but while it played out the utility was forced to respond to a novel public health issue with few risk management options. In order to evaluate the public health implications of the exceedance, the DC Department of Health expanded the scope of its monitoring programs for blood lead levels in children. Children from 6 months to 6 years of age constituted 2,342 of those tested; 65 had blood lead levels above 10 mg/dL (the “level of concern” defined by the Centers for Disease Control and Prevention), the highest with a level of 68 mg/dL. Most of the children with elevated blood lead levels (46, or 70.8%) lived in homes without lead drinking water service lines. Two correlation analyses found no association between matched first-draw tap water lead concentrations and blood lead level for a child at the same address. This study cannot be used to correlate lead in drinking water with blood lead levels directly because it is based on an ecological rather than individualized exposure assessment, the protocol for measuring lead is based on regulatory requirements rather than intake, numerous interventions were introduced to mitigate the effect, and the period of exposure spanned approximately two years rather than lifetimes for adults. We examine lessons learned, including simultaneous compliance.

M5.5 Innovative Risk Management Approaches to Reduce Sodium Consumption. Forshee RA*; University of Maryland--College Park rforshee@umd.edu

Abstract: Many regulatory and professional organizations are calling for strict new regulations to reduce sodium consumption on a population basis. New regulations on sodium may have significant public health and economic impacts. The new regulations could also affect the range of food products that are available to consumers. This brief explores three regulatory approaches to reduce sodium consumption: 1) risk communication and information, 2) revoking GRAS status of sodium and regulating its use in food products, and 3) establishing a market-based, cap-and-trade system for sodium that builds on similar approaches in environmental regulatory policy. Risk communication and information is a promising risk management strategy for sodium consumption, but current approaches have not reduced sodium consumption to recommended levels. A market-based, cap-and-trade system for sodium would require significant development, but it could produce more economically efficient ways to reduce sodium consumption than traditional regulatory approaches, such as revoking GRAS status.

M5.6 An integration of the salient value similarity model and the traditional perspective on trust. Nakayachi K*; Tezukayama University nakayachi@tezukayama-u.ac.jp

Abstract: The degree of trust in risk managers heavily influences the public's risk perception and their acceptance of hazardous technologies. The present study examined factors affecting public trust in risk managers, integrating the salient value similarity (SVS) model and the traditional view of trust. A thousand residents in the Japanese metropolitan area participated in a survey that investigated public trust in the government offices responsible for the decision to approve a genetically modified rice that claims to inhibit pollen allergies. In the high interest group, giving moral meaning to the outcomes of the decision was found to be a powerful predictor of trust. Perceived agreement in values between individuals and the regulators was also found to influence trust in decisions, as predicted by the SVS model. On the other hand, as expected with the traditional perspective on trust, a perception of procedural fairness and a belief in the competency of the risk managers were found to be powerful predictors of trust in the low interest group. Implications of these results for the practice of risk management were discussed.

M5.7 Food Safety Risk Assessment and the Risk Communication Related to Methyl Mercury in Seafood in Japan. Koizumi N*, Hara M; Commissioner, Food Safety Commission, Japanese Government megumi.hara@cao.go.jp

Abstract: Japan is culturally one of highly fish consumed countries in the world. While the Ministry of Health, Labour and Welfare had called attention to the public that the consumption of specific fishes would affect for the fetal risk through the pregnant mothers in 2003. Subsequently it invoked media attention and social confusion on the consumption of fish. The driver of public fear might be explained from 2 deadly past disasters, MINAMATA disease, which were methyl mercury poisoning caused by industrial effluent. In July, 2003, the Food Safety Commission was established under the Cabinet Office, independent from risk management organizations. It was established for the purpose of implementing risk assessments and the risk communication in a neutral and fair manner on the basis of scientific knowledge. In 2005, the Food Safety Commission published the assessment report of methyl mercury in seafood which clearly determined the tolerable level intake and pregnant women as the high risk group. The result was accepted successfully to the public without severe confusion through effective risk communication activities which were press releases, understandable references on web, public meeting and so on. We would like to present the assessment result and the risk communication activities. In addition we would show the DVD “How about methyl mercury in fish?” which we made for one of the tool on the risk communication activities to lay people and we would like to introduce some DVDs about clinical cases of MINAMATA disease as valuable experiences in Japan.

M5.8 The Avaluator – A New Support Tool for Decision-Making in Avalanche Terrain. Haegeli P*, McCammon I; Simon Fraser University - Burnaby BC, Canada; Snowpit Technologies - Salt Lake City UT pascal@avisualanche.ca

Abstract: Over the past decade, recreational winter backcountry activities, such as backcountry skiing and snowmobile riding have experienced a tremendous increase in popularity. However, the winter mountain wilderness is a hazardous place, where even little mistakes can lead to injury or death. Snow avalanches pose one of the most significant threats to backcountry travellers. Worldwide, approximately 150 backcountry travellers die from avalanches every winter. Travelling safely in avalanche terrain requires an in-depth understanding and considerable practical experience in the field. Most amateur recreationists, however, are not able to spend enough time in the backcountry to get the necessary practical experience. In addition, travelling in avalanche terrain is a voluntary risk that can provided participants with very powerful positive experiences. It is well known that this configuration results in a decision environment where people tend to be overconfident in their skills to control their exposure to the hazard and generally underestimate the likelihood of an accident or negative outcome. These characteristics result a challenging environment for decision-making and the development of effective awareness initiatives. In this presentation, we will introduce the Avaluator, a new decision support tool for travelling in avalanche terrain, which aims to address some of these major challenges. The tool uses a combination of expert expertise and avalanche accident statistics to provide users with appropriate travel recommendations during various stages of backcountry trips. In addition, the Avaluator presents a general decision framework that can be expanded and refined as the user gains more experience over time. The presentation will primarily focus on the reasoning and research that led to the development of the Avaluator, but we will also touch on the first results of our efforts to evaluate the effectiveness of the new tool on decision-making in avalanche terrain.

M5.9 Panic: Dimensions of its use and meaning in print journalism. Laskin A, Robinson JA, Alt A*, Chamoff M, Tilley A, Edwards L, Flowers G, Johnson L, Weikman K; University of Florida jrobinson@jou.ufl.edu

Abstract: In an emergency situation, risk communicators want the public to proceed with an orderly evacuation or take preventive actions rather than ensuing in panic. The current understanding of panic from psychology includes an acute emotional response or unreasoning terror often associated with mass flight, as opposed to LaPierre’s sociological definition as the opposite of regimental behavior. This distinction between social and individual psychological response is important for understanding the role of panic in the public’s response to risky situations and whether communicators should avoid mentioning it. Fischoff contends that group panic is actually relatively uncommon in response to a dangerous situation, yet photographs and descriptions often linger on the first panic response, neglecting the orderly evacuation that follows (e.g. Chess, 2006). A content analysis was conducted for 340 instances of the word panic used to describe a situation in 299 articles sampled over one year from influential local/regional US newspapers and national newspapers from three western countries. The results suggest that all newspapers do not use the term panic with the same frequency (X2 = 100; p<.000), panic is most likely to be in a news or feature story, and negative uses appear nearer the front page. Interestingly, individuals rather than groups were the focus of most panic references, with groups significantly higher only for financial panic, whereas, individuals were dominant for panic originating from transportation, health-related, personal and unintentional causes. The most frequent cause of panic was anxiety (41%), followed by financial panic, sports, intentional acts (e.g. terror attack), time-constraints, and personal health issues other than anxiety. Panic was most frequently used in the journalist’s own writing rather than a direct quote. International, gender, valence and source differences, as well as implications for risk communicators are also discussed.

M5.10 Requirements for information for the public on health risks from benzene by applying the ‘Mental Models Approach’. Kubota H*, Kosugi M, Yokoyama T, Tsuchiya T; Environmental Science Research Laboratory, Central Research Institute of Electric Power Industry hiromi-n@criepi.denken.or.jp

Abstract: The objective of this study was to clarify what specific information should be provided for the public to progress its understanding and support decision making of chemical health risks and to make the risk messages for risk communication in Japan. As a method, we applied the ‘Mental Models Approach’ to a case of health risks from benzene, which is one of the volatile organic compounds and a carcinogen. First we designed an ‘expert mental model’ of the risks from benzene based on the current scientific knowledge. Next, a ‘public mental model’ was drawn from data elicited from open-ended interviews with 30 participants and web survey of questionnaires of 211 participants. The comparison of the two mental models enabled us to identify key knowledge gaps and misunderstandings of exposure, the health effects and self-protection. It was suggested that the following information on benzene should be provided for risk communication: 1) the information on major emission sources; 2) the proper information of the target organs associated with the exposure pathways, including the fading of benzene in body excretion process and the environment; 3) benzene and benzine are different chemicals. With these results, we produced a concise leaflet on benzene through a technical review by experts. Focus group interviews of 30 target people were conducted to evaluate the validity of the risk messages. This study, while demonstrating the viability of the applied methodology, also derived notable lessons in making risk messages on volatile organic compounds.

M5.11 Linking Risk Communication and Biomedical Ethics: The Case of PGD . Longstaff H.A.*; The W. Maurice Young Centre for Applied Ethics at the University of British Columbia longstaf@interchange.ubc.ca

Abstract: Risk communication is a field driven largely by case study research and the results of communication experiments (Scherer and Grelner, 2005). The evolution of risk communication theory (as described by expert accounts such as Fischhoff, 1995; Covello & Sandman, 2001; and Leiss, 1996) illustrates how practitioners have employed various theories to explain research findings in order to grapple with the changing nature of risk problems. Until approximately ten years ago, communicators were primarily responsible for helping communities cope with regional environmental hazards (i.e., nuclear power plant explosions, radon gas, natural disasters) controversial events (i.e., the Love Canal, Times Beach, Union Carbide Bhopal accident, Centralia mine fire, 60 Minutes Alar saga), controversial products (i.e., IUD, hormone replacement therapy), or encouraging the public to reconsider dangerous behaviours (i.e., smoking, drinking and driving) (Slovic, 1992). Now, in addition to these efforts, practitioners must realize new ways of approaching systemic multi scalar risk issues that affect the health of national (i.e., pre-implantation genetic diagnosis) or even global communities (i.e., global climate change) among other things. Yet despite the significant impact that risk communication is intended to have on the behaviour of individuals and groups, only a small number of papers have been written on ethics and risk communication. Articles that have been published typically focus primarily on procedural ethics, or ethically constructed risk communication. One of the main assumptions of this paper is that one must pay close attention to both procedural and substantive ethics (including generally accepted bioethical principles) when crafting health risk communication protocols as both influence our ability to achieve just outcomes (Sherwin, 2001).

M5.12 Exploring the Notion of Public Meetings as Social Stations of Risk Amplification. McComas KA*, Foster MW, Trumbo CW, Besley JC; Cornell University, Cornell University, University of Vermont, University of South Carolina kam19@cornell.edu

Abstract: Public meetings are often held to communicate with community residents during investigations of possible cancer clusters. One lingering question is the extent to which attending public meetings influences residents’ risk perceptions. To answer this question, previous research has used survey research to examine pre- and post-meeting attitudes to evaluate changes in risk perceptions. Research has also simply asked attendees whether they were more or less concerned after attending public meetings. This study goes one step further to pose the question “why?” to meeting attendees. Specifically, as part of a larger study of cancer clusters, we obtained written responses from 125 individuals across six U.S. communities who had attended a public meeting in their community about suspect cancer rates. In the survey, we asked respondents to elaborate on why attending the public meeting made them feel more or less concerned. Based on previous research, we developed a coding sheet to examine the degree to which respondents mentioned informational, relational, and individual factors. Two coders analyzed the open-ended responses using quantitative content analysis. Overall, 62% of respondents indicated that attending the meeting made them more concerned. We found significant correlations between being more concerned and citing family health/history of cancer and acquaintance history of cancer. In addition, people who were more concerned were also more likely to mention that information they received from neighbors or other citizens at the meeting was the reason for this heightened concern. In comparison, those who were more concerned were also less likely to state they relied on mass media or other, unspecified sources of information. Few respondents mentioned the results of the official investigations as a rationale for their concern. The results suggest that meeting attendees may give more weight to unofficial sources of risk information than to official information presented at public meetings.

M5.13 Nano Literacy in Japan: Hopes and Limitations. Nishizawa M*; Litera Japan Co. nishizawa@literajapan.com

Abstract: There is an increasing awareness of potential risks of nano-particles to human health. It is known that the building of belief and opinions about negative health effects of science and technology largely relies on selective, sensationalised information distributed by the mass media such as newspaper articles and television programmes (e.g. O'Neill, 2002, Nishizaw a and Renn, 2006). Drawn from case studies in Japan, this presentation will address the role that the media play in influencing the construction of risk perception about nanotechnology, and argue how,for example, government-led risk communication programmes could deliver more accurate images of the technology to the wider society. References: O”Neill, O. 2002：Autonomy and Trust in Bioethics, Cambridge University Press, Cambridge. Nishizawa, M and Renn, O. 2006: Responding Public Demand for Assurance of GM crops: Case from Japan, Journal of Risk Research, (9)1, 41-56.

M5.14 A systematic review of methods for studying risk perception in risk communication of chemicals: An attempt at categorization and classification. Clauberg M*, Schuetz H, Wiedemann PM; Univ of Tenn. Knoxville; Forschungszentrum Juelich, Germany claubergm@mindspring.com

Abstract: Risk communication (RC) is a fundamental component of the risk analysis (RA) paradigm. Not only is it “unavoidable” to engage in RC amongst even risk experts from various disciplines in order to assess the risk, but it is also an essential tool for bi-directional sharing of information, values, and preferences between experts, decision-makers and stakeholders. However, non-experts are known to perceive risks differently and this insight has spawned the field of risk perception (RP) research. How RP studies are to be undertaken, what outcomes are to be expected and how this may/should influence risk communication strategies is still being actively researched. Our project supported the Physical and Chemical Exposure Unit of the Institute for Health and Consumer Protection in the European Commission’s Joint Research Centre (JRC) in their goal of developing an expert knowledge system on risk perception and risk communication and enhancing the resource capability for risk communication for chemicals released from consumer products/articles. Our first step was to perform an in-depth literature review of the state-of-the-art, which was followed by expert workshops and questionnaire surveys to elucidate a categorization and classification scheme allowing for a systematic review of addressing RP methods and studies for chemicals released from consumer products/articles. The development of an expert knowledge system on the basis of this review will be discussed.

M5.16 Use of Guided Interactive Statistical Decision Tools for Hanford Site Risk Assessments. Stockton T, Houghteling W, Bronson D, Linville J, Swanson W*; Neptune and Company, Inc. rryti@neptuneinc.org

Abstract: The present-day mission of the Hanford Site, a former nuclear fuel production facility near Richland, Washington, is to clean up the radioactive and hazardous wastes generated during previous years. As final Records of Decision are being developed for Hanford waste sites, risk assessments are being performed utilizing newly-collected and historic analytical data to determine residual risks to human health and the environment. Accessibility and consolidation of data from various sources are integral pieces in the performance of a concise, defensible, and transparent assessment of risk to human health and the environment. The Guided Interactive Statistical Decision Tool (GiSdT) is a user-friendly web-based interface for assembly, query, and analysis of data relevant to risk assessment efforts at the Hanford Site. The user interface promotes transparency of project activities by offering a repository of analytical data, project documentation, and data analysis tools that are equally available to clients, state and Federal agencies, stakeholders, and the interested public. The centralized database version control system provides a measure of quality assurance and ensures that all project members have access to the most current information. The GiSdT interface uses open-source programming languages (R, SQL, and XML) to assemble and normalize data from multiple sources into a single, universally available, cross-platform, queryable interface. GiSdT’s library of statistical and graphical analysis tools provide resources for individuals and organizations desiring real-time calculations and graphical analyses.

M5.18 Unofficial Carriers: Accidental Risk Communicators and the Public. Rickard LN*; Cornell University lnr3@cornell.edu

Abstract: Within the risk communication literature, considerable attention has focused on intentional messages delivered by “official” messengers. This research pursues an alternative course, examining unintentional risk messages and their “unofficial” carriers. Referred to as “accidental risk communicators” (ARCs), these individuals routinely discuss health and environmental risks with the public as a secondary, not formally recognized job. From the tanning salon owner explaining the “safety” of UV radiation, to the auto mechanic performing an inspection, to the commercial painter handling lead-based paint, ARCs span a variety of occupations. To some extent, journalists are ARCs when they translate risk issues into newsworthy articles. Since ARCs often represent for-profit businesses, the public may be suspicious of their message, perhaps assuming that the information reflects corporate self-interest rather than “impartial” scientific information. Yet often blue-collar laborers, ARCs may appear more similar to some sectors of the public, and thus more credible than “remote” government officials. Moreover, the pressure to sell a product may lead the ARC to deemphasize risks, whereas an occupational injury may make certain risk factors more salient. More systematic study of ARCs could link current areas of risk communication (e.g., trust and credibility; affect), and build a foundation for developing theory. To begin explicating this concept, this research examined turf and lawn-care workers who are trained and certified to apply pesticides, and often expected to discuss risks with curious onlookers. How these workers perceive, position, and evaluate risks could play a critical role in what and how they communicate with the public. This presentation offers a preliminary analysis of ongoing in-depth interviews with a purposive sample of this population.

M5.19 Canada’s critical infrastructure protection at the interface: Building information-sharing capacity across sectors and jurisdictions. Quigley K*, Pelot R; Dalhousie University kevin.quigley@dal.ca

Abstract: Critical Infrastructure Protection (CIP)—activities that enhance the physical and cyber-security of key public and private assets—is the focus of urgent attention among Western governments in light of recent power failures, computer viruses, natural disasters, epidemics and terrorist attacks, both threatened and realized. The human and technological systems that underpin the critical infrastructure are complex and interdependent, and the number of entities involved leaves authority and responsibility diffuse. While there is an emerging consensus that industry and government at all levels must cooperate on issues regarding securing the critical infrastructure, few fora or mechanisms have emerged in Canada to facilitate reliable information-exchange for strategic planning and research. The School of Public Administration and the RBC Centre for Risk Management at Dalhousie University have recently been awarded a contribution from the Canada School of Public Service’s Innovative Public Management Fund to help initiate appropriate cross-sectoral and cross-jurisdictional fora in which participants can access and share diverse and expert perspectives on CIP, exploring technical as well as social, legal, political, economic and managerial issues. Four projects support the goal: an interactive webpage; an on-line expert directory; a newsletter; and a workshop. The implementation phase of the initiative will take eighteen months to complete (March 2007-August 2008). The medium term goal is to ensure motivated networks will emerge better informed about the complexity of the infrastructure and the holistic approaches necessary to make the infrastructure more secure and resilient for the benefit of all stakeholders. The purpose of the poster session is to highlight and promote the four initiatives and seek opportunities for collaboration with like-minded academics and practitioners.

M5.21 Open Collaboration in Risk Assessment. Pohjola MV*, Tuomisto JT; National Public Health Institute of Finland, University of Kuopio mikko.pohjola@ktl.fi

Abstract: Successful assessment of complex real-world phenomena, such as environmental health risks, requires contributions from various individuals and groups with varying roles and perspectives to the assessed phenomena; scientists, decision-makers, other stakeholders, and public at large. Broad and diverse groups of contributors should thus be involved in risk assessments in order to increase the quality of the information content and improve the acceptability of assessment outputs. Within the context of risk assessment, open collaboration means creation of shared understanding among a diverse group of contributors about the objects of the assessment, which become explicated in the form of the assessment output. Open collaboration is currently being applied and tested in some environmental health risk assessment cases, where the National Public Health Institute of Finland is involved, representing varying levels of openness and diversity of contributors. Open collaboration is enabled in the cases with a wiki-type assessment workspace which provides a common interface for diverse, geographically dispersed groups to participate in the assessments. The preliminary experiences from the case studies show that enabling and managing collaborative work in risk assessments among broad and diverse groups is feasible. It also appears that in order to ensure the efficiency of open collaboration in risk assessment, openness may need to be managed according to its different dimensions; participation, access, influence and temporal presence. The assessment workspace developed on Mediawiki software-platform suits well for facilitating collaboration of diverse groups and managing openness of contributions. Open collaboration, involving diverse groups of contributors, appears to be a promising means to improve the effectiveness of risk assessments. The currently existing technology makes it possible to carry out risk assessments in a more open and collaborative way.

M5.22 Melding Science and Stakeholders to Achieve Solutions at Amchitka Island: Breaking the Log-jam in the Department of Energy's Nuclear Weapons Complex. Burger J*, Powers C.W., Gochfeld M, Kosson D; Rutgers, The State University burger@biology.rutgers.edu

Abstract: The Department of Energy (DOE) is faced with massive cleanup and closure of its nuclear weapons facilities. Despite multibillion dollar expenditures, very few sites have been cleaned up, closed down, or returned to other uses. Partly this is a result of failure to meld science, technology, and stakeholder needs. Traditional scientific research proceeds from development of a hypothesis, through data gathering to final conclusions without much input from stakeholders. Scientists wonder why policy makers ignore scientific data, while policy makers wonder why scientific research has little relevance to policy challenges. We propose that the melding of scientists and stakeholders throughout the process can reduce conflicts and lead to acceptable solutions for problems that are complex and have eluded resolution. We use the closure of the DOEs Amchitka Island, where three underground nuclear tests were conducted, as a case study to illustrate how stakeholders can be included throughout the process, leading to acceptance and incorporation of the science. There was intense stakeholder controversy, frustration, and anger among federal and state agencies, and Native Alaskan about Amchitka, obscuring any path forward to a solution that would allow closure of Amchitka. The process was dynamic, and acrimonious at some stages, but at every step new mechanisms had to be developed to address the specific concerns of a range of stakeholders. Without such inclusion, well-intended policies and practices may be ineffective and may not lead to a solution, particularly to such difficult problems as closure of hazardous waste sites, and the handling of civilian and military nuclear wastes in the future. This impasse influences the future of nuclear energy in the United States.

M5.23 Modeling the Structure of Risk: Path Analysis Revisited. Donelson AC*; University of California, Berkeley adonelson@berkeley.edu

Abstract: Motor vehicle crashes have both a temporal and a causal structure. This paper presents findings from a study of structural models for estimating crash and injury risk. Earlier attempts to develop models with two or more dependent variables that were categorical (binary, ordinal) suffered from a dearth of modeling software with estimators appropriate to the form of data. For example, LISREL, PROC CALIS [in SAS], and EzPath (in SYSTAT) required that endogenous variables be random continuous variables. The evolution of statistical software that supports analysis of categorical data (e.g., Mplus, ℓEM) provided an opportunity to revisit “path analysis” as a viable way to model the “structure of risk”. As before in this line of research, vehicle rollover provided the substantive problem for purposes of demonstration. Structural models developed with Mplus and ℓEM are compared to those based on an alternative approach – “tetrad analysis” – applied both as a confirmatory and as an exploratory method for structural equation modeling.

M5.24 The Western Canada Study: Management of a High-Profile Study of Risk. Guidotti T*, Waldner C; George Washington University eohtlg@gwumc.edu

Abstract: In 2000, the Government of Alberta took the initiative to organize The Western Canada Study on Animal Health Effects Associated with Exposure to Emissions from Oil and Natural Gas Field Facilities (henceforth called “the Study”) to investigate allegations that reproduction in cattle and productivity in cow-calf operations had been adversely affected by emissions from oil and gas facilities. The Western Interprovincial Scientific Studies Association (WISSA) was set up to manage the study professionally but at arm’s length. A Scientific Advisory Panel provided oversight to develop a strategy for the Study, to guide and monitor its implementation, for quality assurance, to ensure scientific integrity and to review data and assist in the interpretation of findings. The Study took six years, cost Can$17 million (provided by the provincial governments, the Canadian Petroleum Producers Association and the Alberta Beef Producers). The Study was designed to accomplish two important goals: 1) to determine if emissions downwind from oil and gas facilities influence the reproduction and health of cattle and wildlife in western Canada and 2) to provide new and economically useful information for cow-calf producers on factors affecting herd productivity and herd health. A tandem wildlife study was implemented using European starlings as an indicator species. The overall pattern of findings demonstrated no adverse effect of downwind exposure on animal health, although the Study found some associations that need to be interpreted carefully and better understood. The Study is the largest and most comprehensive study undertaken to date of beef cattle productivity in North America. Important design features of this Study included the innovative management structure, a secondary goal of improving herd management, a hypothesis-driven study protocol, the unusually active role of science advisors.

M5.25 A Weight-of-Evidence Framework for Reaching Scientific Consensus on Causality . Bradley A*, Durda J, Brzuzy L; Integral Consulting abradley@integral-corp.com

Abstract: A site-specific, weight-of-evidence (WOE) decision framework was developed and applied by a multi-stakeholder group to assess the potential contribution of groundwater discharges to sediment toxicity in a Texas bayou. The framework was developed because existing published WOE frameworks were too narrow in focus to address the types of information being considered in the assessment, did not consider causality, and/or used decision criteria that were not explicitly defined a priori. This framework incorporates analyses of four independent lines of evidence: chemical correspondence between groundwater and sediment, chemical loading from groundwater to sediment, risk analysis, and spatial analysis of toxicity data. It requires stakeholders to weight types of data and lines of evidence on a relative scale that reflects data quality and certainty as well as overall relevance to the questions being addressed. Analyses that are tied most closely to the assessment question are given a higher weight, and ultimately have the greatest influence on the conclusions. The framework is being used to reach consensus on whether groundwater has contributed to measured sediment toxicity in the bayou. The scientific answer is then linked to a pre-defined management decision for the site. This paper presents the results of this specific application and discusses how this WOE framework, or the process used to develop it, can be applied to similar cases to synthesize, reconcile, and integrate multifaceted data to understand causality and inform management decisions.

M5.26 Ecological risk assessment in the Alaskan tundra - a portfolio approach. Pfingst AJ*, Stelljes ME; SLR International Corp apfingst@slrcorp.com

Abstract: Ecological risk assessments (ERAs) were conducted at multiple, similar Sites as part of closure activities under the RCRA process in the Alaskan tundra. Rather than conducting individual ERAs at dozens of sites, a portfolio approach was implemented that grouped Sites based on similar characteristics. These characteristics include (1) habitat, (2) drainage features, and (3) nature of contamination. Three primary habitat types were observed across the range of investigated Sites: dry, wet, and mixed tundra. Surface water features represent the primary drainage features; these varied in amount and proximity to individual source areas. Nature of contamination ranged from metals to PAHs and volatile organic compounds. Sites with all three characteristics in common were evaluated in a single ERA, with a single conceptual site model that incorporated all sites in each group. This approach greatly reduced the number of individual ERAs needed in support of the closure activities, saving both time and money in the process. Details of the process, and of the challenges facing those conducting ERAs in the unique environment of the Alaskan tundra, will be discussed.

M5.27 Multi-disciplinary approach to assessing the risk of P. ramorum, Sudden Oak Death, to US ecosystems. McDowell RM*; USDA Animal and Plant Health Inspection Service robert.m.mcdowell@usda.gov

Abstract: A phenomenon called Sudden Oak Death began occurring in California in the 1990’s and in 2001, a fungal-like organism, Phytophthera ramorum was identified as the causative agent. P. ramorum has been shown to affect a wide range of woody and herbaceous plants, and has caused significant morality to the oak forests in California. It has also been found plant nurseries in Oregon and Washington State. APHIS conducted a multi-disciplinary risk assessment to evaluate the threat posed by P. ramorum to the plant resources in the USA, the possible spread by nursery stock, and potential mitigations. This analysis integrated in many levels the epidemiologic characteristics of the pathogen and the host species. This analysis involved GIS approaches including climate-matching and host distribution analysis to determine areas where (1) climatic conditions would permit P. ramorum survival and (2) hosts were present. This analysis incorporated normal weather variability in predicting where and how frequently appropriate climate conditions would prevail. In addition, several reliability engineering techniques were used to model the risk of specific pathways. Quantitative epidemiological approaches were used to develop epidemiology-informed sampling approaches that take into account the nature of the disease diagnostic test uncertainty to develop statistically valid conclusions from sampling data. The consequence assessment included consideration of objective economic measures of various outcomes and non-economic considerations which included the overall ecological impact of species loss in woodland ecosystems.

M5.28 Public acceptance and commercialization of nanotechnology products in Japan. Kishimoto A*; National Institute of Advanced Industrial Science and Technology kishimoto-atsuo@aist.go.jp

Abstract: Public perception and acceptance of nanotechnology in Japan is investigated via Internet surveys and market research. Public awareness has been enhanced mainly through TV commercials announcing the applications of nanotechnology. Product sellers consider it an attractive proposition to inform consumers about their products containing nanomaterials. We conducted a survey for three consecutive years on the public perception of nanotechnologies and their application to consumer products. These surveys, conducted from 2005 to 2007, used Internet monitors as a subject. In the most recent survey, first, participants were asked to offer their perceptions of nanotechnology in the abstract sense, without providing them any information about nanotechnology. After they received general information about nanotechnology, including its advantages and disadvantages, the participants were asked how they perceived the risks and benefits of nanotechnology and other various technologies and substances, and their applications. In May 2006, we also launched the world’s second online inventory called “Nanotechnology Consumer Products Inventory in Japan,” which contains more than 300 products. The cosmetics category accounts for more than one third of these products. The kind of nanomaterials used and their self-reported new properties are classified and compared to those in other countries.

M5.29 Risk communication in nanotechnology: Our scope. Sekiya M*, Ishizu S, Ishibashi K, Ata M; National institute of advanced industrial science and technology mizuki-sekiya@aist.go.jp

Abstract: Products using nano materials are increasing and visibility in store shelves and in advertisement are also increasing. But actual numbers of products using nano materials are higher than we can find in store shelves. Some companies become uncomfortable using word “nano” on their products and avoid to state. Last year a Japanese consumer group appealed moratorium on nanotechnology R&D to the government. These trends seem to be originated from lack of information in public about nanotechnology R&D and risks associated with them. I introduce our research activities which includes finding a better way to reach public and help nanotechnology accepted in society. We cooperated with other research institutes and conducted a research project “Facilitation of Public Acceptance of Nanotechnology” in 2006 and issued policy recommendations. Our recommendations were reflected in the government policy and a new cross ministries research framework for social acceptance of nanotechnology which includes research on better risk communication with public has initiated. Through our activities, the importance of communication with public is thrown light on, and disclosure of information on nanotechnology R&D to public is expected as a role of researcher. Considering the fact that many people rely on the media when they get information, researchers, policy makers and other nanotechnology related persons should understand the importance of learning how to respond the media. Encourage public into nanotechnology R&D process is also a key to gain public trust. We keep working on establishment of the suitable method to communicate with public. As one of the efforts, we organized “nanotech debate - Let’s talk about precautionary principle” on February 1st 2007. The meeting aimed to help public to understand current situation in the nanotechnology R&D. I believe our activities contribute to risk communication in nanotechnology R&D and help to win the public’s trust.

M5.30 Risk Communication: The Spinach Outbreak and the Media . Todd E C D*, Harris CK, Knight AJ, Worosz MR; Michigan State University toddewen@cvm.msu.edu

Abstract: The media was the main source of information as the investigation of E. coli O157:H7 spinach outbreak of 2006 unfolded. The final total was 205 cases in 26 states. Numerous experts representing different groups speculated as to the cause of the outbreak, its impact, and possible preventative and control measures for the future. On a daily basis, the public was informed through short news items in which the opinions of spokespersons and self-declared experts were stated but sometimes differed. Comments included trust in the spinach industry, organic spinach production, environmental contamination sources, processing decontamination strategies, tracking leafy green products, improvements to foodborne surveillance and the government oversight system, and the impact of this information on consumers’ attitudes to spinach. Two issues arising from these media statements in retrospect are (1) what qualifies an expert to speak on these issues, and (2) how do journalists extract critical statements that are newsworthy and still convey the experts’ main messages. In major outbreaks such as this one, official government reports are not going to be completed until well after the heightened awareness of the event. Thus, the role of industry spokespersons, appropriate government officials, and the media in communicating information to the public is critical to help consumers make informed decisions for themselves and their families without demonizing or exonerating those most immediately involved. Unfortunately, the exact means of E. coli transmission to the spinach was not determined and spinach sales are still down. The industry has agreed to create a uniform approach to good agricultural practices to anticipate and prevent future contamination of leafy greens, and try to win back public acceptance. However, spinach and lettuce growers may have to deal with stricter state and federal governments standards, which will be an additional financial burden.

M5.31 Foundations and Frameworks of Microbial Risk Assessment. Parkin RT*; The George Washington University parkinr@gwu.edu

Abstract: Although a variety of microbial risk assessment (MRA) approaches are in use, limited analyses of MRA frameworks or their underlying principles and concepts exist. Organizations’ desires to develop a unified MRA approach have been tempered by the realization that flexibility is essential for addressing various legislated mandates and regulations and for meeting diverse field application needs. This EPA-sponsored project included a review of recent MRA policies and guidelines and a comparative analysis of 13 MRAs conducted or sponsored by governments worldwide. The most commonly cited underlying principles for MRA were: make public health protection the priority, base MRA on sound science, ensure transparency, use a structured and consistent approach, and allow for iterations. Two forms of risk assessment—chemical risk assessment and ecological risk assessment—provide the foundations for MRA. The National Academy’s widely used four-step risk assessment paradigm is the prevailing context from which many modelers have approached MRA. The dynamic aspects of pathogens, environmental processes, human populations, and the interrelationships of these entities are increasingly being noted as important MRA modeling issues. The four fundamental types of MRA frameworks found will be described in the presentation. Organizations use MRA paradigms for different purposes and work through the components in a variety of sequences and depths. Few MRAs include problem formulation, but most include a problem statement or definition. Some organizations have developed modules (self-contained mathematical models) to represent parts of the source-to-health effects chain. Recommendations will be made for advancing MRA through systems methodologies and communication strategies.

M5.32 Risk Evaluation of Invasive Species Transport Across the U.S. – Canada Border in Washington State. Sellens LS*, Jewell AT, Markiewicz AJ, Landis WG; Western Washington University laurasellens@gmail.com

Abstract: Impacts of non-indigenous invasive species (NIS) are a multi-million dollar cost to society every year. Incongruities exist between the Canadian and American policies that regulate the introduction and control of invasive species. Policies may ineffectively overlap across terrestrial, freshwater, and marine jurisdictions. NIS are easily overlooked at border crossings due to the small size of propagules and because immigration, customs, and security concerns take precedence. The Relative Risk Model was used to conduct an ecological risk assessment to evaluate the risks posed by human mediated transport of NIS across the Washington State-British Columbia border. This project investigated the terrestrial, freshwater and marine transport pathways of invasive plants and animal species to the major habitat types in Washington, using Canada as the hypothesized source. Risks of impacts by NIS were calculated for economic, cultural, and ecological endpoints of importance to stakeholders. Quantitative analyses were performed using spatially explicit data and a geographic information system (GIS). Uncertainty and sensitivity analyses were performed using Monte Carlo simulation software. Overall risk of NIS introduction was highest for freshwater boats compared with terrestrial and marine pathways. The results of this assessment will be used to inform policy makers and managers on both sides of the border of the most important transboundary transport pathways for NIS as well as major data gaps and sources of uncertainty.

M5.33 Quantitative risk assessment of pathogenic Vibrio parahaemolyticus for raw horse mackerel consumption in Japan. Iwahori J*, Yamamoto A, Suzuki H, Yamamoto T, Tsutsui T, Motoyama K, Sawada M, Matsushita T, Hasegawa A, Kasuga F; Kochi University, Hyogo Prefectural Institute of Public Health and Environmental Sciences,National Institute of Health Sciences, National Institute of Animal Health, Hitachi East Japan Solutions, Ltd., Mitsubishi Research Institute, Inc. iwahorij@kochi-u.ac.jp

Abstract: Quantitative risk assessment of Vibrio parahaemolyticus(V.p.) in horse mackerels (Trachurus japonicus) consumed raw in Japan was conducted on the basis of the data available in 2006. In Japan, raw horse mackerels and other raw fish are far more important than raw oysters as source of V.p. infection because of their popularity as sashimi, sushi, tataki, etc.　The present model aims to assess effects of control measures to reduce illness by V.p. in raw seafood choosing horse mackerel as typical seafood eaten raw in Japan. In the model, growth of V.p. during all stages from landing to consumption, decrease or increase in washing/storage water, transfer of the bacteria from fish surface, gill, intestine to fillet are modeled by using experimental data. Parameters of the beta-Poisson dose-response model were newly determined from the human feeding test data that were used by US FDA, but with two different definitions of illness: detection of V.p. in feces and diarrhea or more than three times of diarrhea.　These definitions gave, respectively, twice and 7% of the probability of illness that was estimated by the model using US FDA parameters. Choice of the parameters is not critical for comparison of effectiveness of prevention measures since they give consistent results. Probability of illness caused by V.p. was estimated for each selected pathway of scenario alternatives: washing whole fish at landing, usage of contaminated water for storage, experiencing high temperature period in transportation, and washing during preparation. The last was found to be most effective for prevention of illness reducing the risk by about a factor of 10. High temperature period increases the risk depending on its temperature and duration; with our assumed distributions the risk increased by 50%. The other alternatives had insignificant effects on the risk.

M5.34 Simulating the Nutrient Distribution of Dietary Patterns for the Revision of Canada’s Food Guide. Junkins B, Vigneault M*, Johnston JL; Health Products and Food Branch, Health Canada michel_vigneault@hc-sc.gc.ca

Abstract: As part of the recent revision of Canada’s Food Guide, an innovative method was developed to assess the distributions of nutrient and energy content of various food intake patterns. The patterns were defined by specific numbers of servings from the food groups, food subgroups, or foods from these groups with particular characteristics such as low fat. Nutrient distributions were developed for the different food intake patterns by simulating a large number of diets consistent with the proposed patterns, but chosen to reflect the relative food preferences in the population. The simulation model was driven by actual food choices within each food group and subgroup, taken from reported diets, instead of food disappearance aggregates or composites. As the specificity of the food choices behind the patterns were increasingly focused to try to improve nutrient results, the simulated diets and their projected impact remained grounded in the popularity of food choices by Canadian consumers. The nutrient distributions of the simulated diets were estimated for different age-sex groups, and compared against the Dietary Reference Intake benchmarks such as Estimated Average Requirements (EAR), Adequate Intakes (AI), Tolerable Upper Intake Levels (UL) and Acceptable Macronutrient Distribution Ranges (AMDR). The method allowed examination of the differential nutritional impact of food patterns in subpopulations (such as varying age, sex and income groups) where there may be differences in food preferences or access. The presentation will describe this approach and its application in the revision of Canada’s Food Guide. It will also address some of the relevant methodological and interpretation challenges.

M5.35 Classification of risk assessment approaches by risk assessment questions in food safety risk analysis. Yamaguchi H*; Kyoto University h.yamaguchi@kw4.ecs.kyoto-u.ac.jp

Abstract: Approaches to risk assessment have been shown to vary among regulatory agencies and become more complex and diverse for the following two reasons. First is that the question to be asked of risk assessment have become diverse, dependent on different issues required to use cumulative risk assessment, alternative risk assessment considerations or risk-benefit analysis. Second is that different assumptions and justifications are used by each risk assessor, because of the limitations of data or the specific characteristics of the technological sector. This poster analyzes current approaches to risk assessment and classifies its approaches based on the view point of risk assessment questions, with the case of a major chemical or biological risk in food. Furthermore, it analyzes how the answer by risk assessment is informed decision making in risk management and describes the necessary requirement for coordination among risk assessment and risk management to implement the appropriate approach to risk assessment.

M5.36 Evaluation of a PM2.5 Standard in Chile. Cifuentes Luis*, Strappa V, Gonzalez-Caballero G; P. Universidad Catolica de Chile lac@ing.puc.cl

Abstract: Many Chilean cities suffer from air pollution problems, especially from particulate matter. The current Chilean primary standard are similar to the older US PM10 NAAQS: 50 ug/m3 for the annual mean and 150 ug/m3 for the 98th percentile for 24 hour average. Due to increasing evidence of differential impacts depending on the size of the particles, the Chilean environmental authority is considering the implementation of a PM2.5 based standard. This work presents an analysis of the net social benefits of implementing such an standard in four cities with different emissions and demographic characteristics: Santiago, the country’s capital, a large city with a complex mix of emission sources, Tocopilla, a small city dominated by power plant emissions, Concepcion, a middle size city with high industrial activity, and Temuco, another middle size city whose emissions are dominated by wood burning, especially in the winter. Implementation of the standards is assessed in terms of the costs of compliance and the health benefits of assessment. The health risks to the population are discussed in the context of other environmental and public risks. The results show that the convenience of implementing a fine particle standards depend on the city analyzed and on the background concentrations of both fine and coarse fractions of particulate matter.

M5.37 Finite Difference scheme for finding Option Price Using Long memory volatility. Sarangi p, padhy s*; Sr .Lecturer ,professer prakash_sarangi89@yahoo.com

Abstract: The aim of this paper is to present a stochastic model that accounts for the effects of a long-memory in volatility on option pricing. The starting point is the stochastic Black-Scholes equation involving volatility with long-range dependence. We define the stochastic option price as a sum of classical Black-Scholes price and random deviation describing the risk from the random volatility. By using the fact that the option price and random volatility change on different time scales, we derive the set of linear equations for this deviation by using finite difference scheme. Then solving all linear equations by using Gauss Seidal numerical method, the solution to this equation allows us to find the option pricing .

M5.38 Cost effectiveness of reducing heavy metal emissions from municipal solid waste incinerators in Japan. Ono K*, Gamo Y; National Institute of Advanced Industrial Science and Technology (AIST) kyoko.ono@aist.go.jp

Abstract: In Japan, since 1997, most of municipal solid waste incinerators (MSWIs) were modified and newly constructed due to meet the new emission standard for dioxin. Kishimoto et al. (2001) called that modification and new construction of MSWIs as an emergency countermeasurement and a long-term countermeasurement, respectively, and they estimated the annual costs of these countermeasurements. The estimated annualized total costs are 1.46 billion yen for the emergency countermeasurement and 41.2 billion yen for the long-term countermeasurement, respectively. These countermeasurements result in a reduction of not only dioxin emissions but heavy metals emissions from MSWIs. This is because a bag filter was introduced for many MSWIs by the countermeasurements, by which smaller radius of particle matter is captured, and particle matter contains most of heavy metals. In this study, we focused the annual decrease of heavy metal emissions after the countermeasurement and estimated cancer cases avoided by that. Reduction of ambient air concentration was calculated by an atmospheric dispersion model and weighted by population, and then life-time cancer cases avoided were calculated using a unit risk which was obtained by IRIS. Finally life-time cancer cases avoided were converted into annualized life-years saved. Targeted heavy metals in this study are hexavalent chromium (Cr(VI)) and cadmium and so on, which we were able to estimate incinerated amounts in MSWIs and for which reasonable unit risks were obtained. As for Cr(VI), the results showed that annualized life-years saved was 7.8 years, thus cost per life-year saved was 187 million yen for the emergency countermeasurement and 5,282 million yen for the long-term countermeasurement, respectively.

M5.39 Cost Benefit Analysis of Promoting Photovoltaic System. Arakawa C*, Sakai M, Yoshida K; Advanced Industrial Science and Technology c-arakawa@aist.go.jp

Abstract: Photovoltaic (PV) power generation has increased exponentially since the 1990s worldwide. This generation system is attracting attention as an alternative of thermal power generation system which release CO2 and hazardous air pollutants (HAPs) such as NO2 and particulate matter (PM). Therefore, PV power generation has benefits reducing CO2 release and health risk of HAPs from thermal power plant. In contrast, the external costs for promoting the PV system increases due to health risks of chemicals in PV modules and also higher generating cost of the PV power generation causes increase of the cost for introducing the system. The objective of this study was to analyze quantitatively costs and benefits of promoting PV power generation in Japan. For the analysis, we assumed two scenarios for introducing PV power generation from 2003 to 2030, and assumed reduction of health risk due to reduction of NO2 and PM emission from the thermal power plant by introducing the PV. Based on the scenarios of introducing PV, the population who exposed exceeding NO2 or PM air quality standards decrease about 120,000 and 60,000 people, respectively, in 2030. The reduction of the cost for CO2 emissions was estimated based on emission trading costs. It was estimated to be about 220 billion Japanese yen by 2030. Change in power generation costs were calculated from published prediction data of thermal power generation cost and PV introduction target price. We estimated that 143 billion Japanese yen is increased by 2030. Human health issues of hazardous lead included in PV modules were not identified during the entire lifecycle of the module. When the one of scenario, considering that the introducing PV may bring a benefit of about 140 billion Japanese yen by 2030 compared to 2003. The result of the cost-benefit analyses indicated that the further reduction of the generating costs of the PV system is required to accelerate its introduction.

M5.41 Learning from previous risk assessments: Methodological development needs. Tuomisto JT*; National Public Health Institute (KTL) jouni.tuomisto@ktl.fi

Abstract: INTRODUCTION. Collecting necessary data and models, and synthesising information into a useful form for a risk assessment is a laborous task. This cost is unavoidable when the scientific quality of a risk assessment is kept high. However, if risk assessments are looked at as a group, we can see that they often share identical or very similar pieces of information. There are already repositories for this kind of information, such as the IRIS database. But there are good reasons to believe that there are also a lot of other kinds of critical and useful information that could be collected and synthesised. This paper examines the theoretical properties of such a format. OBJECTIVE. The objective is to define a generalisable data format that makes it possible to directly use a piece of information in any risk assessment that has a right scope. Argumentation is used as a primary method in this work. RESULTS. Risk assessments can be divided into pieces in a number of different ways. However, describing risks as causal networks of real-world variables and causal connections between them (as in Bayesian belief networks), is arguably an efficient and useful method and has little philosophical resistance. Therefore, we use this structure as the starting point. Each variable needs the same basic structure, and we suggest seven attributes: name, scope, description, unit, definition, result, and discussion. Based on the contents of these attributes, it is possible to evaluate the usability of a variable in a particular risk assessment. Each variable should pass the so called clairvoyant test, i.e. a putative clairvoyant should be able to give a clear and unambiguous result to the question presented in the scope. A major need is to clarify the relations between general and case-specific variables. CONCLUSION. The data format presented here has worked well in preliminary assessments and offers a promising tool to incerase efficiency in the risk assessment work.

M5.42 Fault tree analysis of patient preparation incidents in radiation treatment for cancer . Lee RC*, Ekaette E, Cooke DL, Iftody S, Craighead P; University of Calgary, Alberta Cancer Board rclee@ucalgary.ca

Abstract: Inappropriate administration of radiation for cancer treatment can result in severe consequences such as premature death or appreciably impaired quality of life. There has been little study of vulnerable treatment process components and their contribution to the risk of radiation treatment (RT). In this paper, we describe the application of probabilistic fault tree methods to assess the probability of radiation misadministration to patients at a large cancer treatment centre. We conducted a systematic analysis of the RT process that identified four process domains: Assessment, Preparation, Treatment and Follow up. For the Preparation domain, we analyzed possible incident scenarios via fault trees. For each task, we also identified existing quality control measures. To populate the fault trees we used subjective probabilities from experts and compared results with incident report data. Both the fault tree and the incident report analysis revealed simulation tasks to be most prone to incidents, and the treatment prescription task to be least prone to incidents. The probability of a Preparation domain incident was estimated to be in the range of 0.1-0.7% based on incident reports, which is comparable to the mean value of 0.4% from the fault tree analysis using probabilities from the expert elicitation exercise. In conclusion, an analysis of a portion of the RT system using a fault tree populated with subjective probabilities from experts was useful in identifying vulnerable components of the system, and provided quantitative data for risk management.

M5.43 DECERNS WebSDSS: Multicriteria Spatial Decision Support System for Risk Based Land Management. Yatsalo B*, Didenko V, Tkachuk A, Mirzeabasov O, Grebenkov A, Sullivan T, Linkov I; Obninsk State Technical University of Nuclear Power Engineering (IATE) yatsalo@prana.obninsk.org

Abstract: Spatial data processing, and analysis and presentation within investigation of environmental problems are based on implementation of GIS-technologies along with the use of math models for examination of different scenarios and assessment of corresponding risk values and criteria under consideration. Decision-making support for selecting remediation options and alternatives on landuse management should take into account not only risk and cost criteria, but also different technological, ecological, geological and social objectives along with the stakeholders interests. Original WebSDSS DECERNS is developed for assistance in decision-making process when solving multicriteria problems with spatially distributed information/alternatives. This Web-system is an MCDA-oriented SDSS with implementation of MAVT, AHP, PROMETHEE and some other MCDA models, and with realization of the basic and advanced GIS-functions. The structure of DECERNS WebSDSS, features of integrating different MCDA methods, GIS-technologies and math models, and implementation within the case studies will be presented. This work is carried out within the DECERNS (IPP/ISTC #3549) project.

M5.44 Risk path finder 2: A tool for profiling documents about risks. Maeda Y*; Shizuoka University maeda@sys.eng.shizuoka.ac.jp

Abstract: Risk Path Finder is an information system for supporting discovery of scenarios of new risk events, that is, causal paths from risk sources to consequences, by displaying profiles of documents related to risks. So far Risk Path Finder had been developed as a system consists with GETA, the generic engine for transposable association, and DualNAVI, a user-interface for text retrieval by using GETA. However because DualNAVI is a product of a company, the users have to pay for the license. So the new version of Risk Path Finder, in which TouchGraph, which is an open source program, is utilized instead of DualNAVI, has been developed. This Risk Path Finder version 2 shows profile of documents related to risks by displaying simultaneously the relationship between the documents and the relationship between characteristic words in the documents. Currently Risk Path Finder 2 is running on a trial basis with two databases, documents related to food safety risk gathered from the WWW, and articles appeared in the period from 2002 to 2004 in the Asahi Shimbun, the biggest Japanese newspaper.

M5.46 Factors associated with sensitivity of input parameters in the assessment of health risk from dietary intakes of carbendazim. Nieh C, Wu KY*; National Health Research Institutes kywu@nhri.org.tw

Abstract: Sensitivity analysis is an important tool to evaluate the relative impact of each input parameter on its own model outputs and has been frequently used in the decision making process to achieve optimal management efficiency. Although various methods have been proposed to conduct sensitivity analysis, the influences of factors on its outcomes have been not studied. The objective of this paper was to evaluate how model input parameters, such as intrinsic property, mean, coefficient of variance, and distributional shape of a parameter of a model, affect the outcomes of sensitivity analysis. The assessment of health risk from dietary intakes of carbendazim residues in fruits for consumers in Taiwan was used as an example. Tangerine intake rate is the most sensitive input parameter than the intake rates of other fruit groups. In order to elucidate why tangerine intake is the most sensitive factor, means, coefficient of variation, and shapes distribution of tangerine intake rate were varied for Monte Carlo simulation. Sensitivity analysis was also conducted simultaneously. Our results showed that sensitivity of tangerine intake rate logarithmically increases with coefficient of variation (R2 = 0.97) and is also highly correlated with the mean of tangerine intake rate. Results from this clearly demonstrate the importance that correct specification of the distributions of input parameters and could have impacts on the decision-making.

M5.47 Department of Homeland Security Chemical Terrorism Risk Assessment. Paulus HI, Hawkins B*, Bowen SE, Kooistra S, Morton DW; DHS heidi.paulus@us.army.mil

Abstract: The Chemical Security Analysis Center (CSAC), one of five laboratories under the Science and Technology Directorate of the Department of Homeland Security (DHS), was established to provide chemical threat awareness and assessment through the characterization of current and emerging chemical threat materials. As part of this mission, the CSAC prioritized the potentially most dangerous chemicals according to risk to the United States (US) homeland population. Due to the complexities involved with chemical threats (thousands of possible threat chemicals and dissemination scenarios), it is difficult for decision makers to adequately allocate resources for research and development that promote the protection of the United States. These resources are divided among various threats: chemical, biological, radiological and nuclear. Therefore, there is the need for a validated, robust tool to quantify the risk associated with each of these threat areas. The CSAC used a two phase approach to perform a Screening Assessment and a Quantitative Risk Assessment (QRA). The Screening Assessment identified chemicals to be assessed in the QRA based on their potential risk to the US homeland population. The QRA, an event tree-based approach, will employ the National Biodefense Analysis and Countermeasures Center (NBACC) of DHS probabilistic approach and will determine risk, defined as a function of the likelihood of the unwanted event and its consequences, by identifying the possible scenarios and estimating the likelihoods and consequences for each scenario.

M5.49 Developing a Generic Risk Assessment Tool to Estimate the Likelihood of Foot and Mouth Disease Introduction into the USA via Fresh Beef Importation. Gebremedhin B*, Abebe M; Auburn University gebrebw@auburn.edu

Abstract: The uses of statistical methods and models that can predict risk have a long history. But using it in global trade in animal product is new (Miller et al., 1993). This paper shows a statistical approach for estimating the presence of Foot and Mouth Disease virus (risk) during a global trade in Fresh Beef (product). A well-known statistical approach—Monte Carlo Simulation— was used to develop the risk estimation (risk assessment) that incorporates the uncertainties inherent in the global trade. Monte Carlo methods are useful for modeling phenomena with significant uncertainty in inputs (Caflisch, 1998). In animal product importation many variability or uncertainty are incorporated such as measuring accurately the failure rate of the mitigation for Foot and Mouth Disease virus occurrence in Fresh Beef. In this study a generic risk assessment tool developed to do the Monte Carlo Simulation analysis by using @RISK Developer’s Kit (Palisade Corporation, Newfield, NY.USA) as add-in on Visual Basic 6 (Microsoft Corporation, Redmond, Washington, USA). It is possible to develop Monte Carlo Simulation models by using @RISK on MsExcel (Microsoft Corporation, Redmond, Washington, USA) that capture the major features of such risk assessment model. However, this is difficult if a large number of iteration needed. In contrast, the approach is simple, and provides the flexibility to study the risk assessment variations with simple changes to the model. The model attempts to accurately put the uncertainties and variability’s levels, its resultant impact, and predict the range of the risk (contaminated fresh beef imported per year) which creates a transparent and scientifically based approach for the decision-maker.

M5.50 Scanning the Horizon to Identify and Monitor Emerging Contaminants. Rak A*, Hutchens S, Bodden M; 1 and 3 Noblis Falls Church, VA and 2) Army Center for Health Promotion and Preventative Medicine, Aberdeen Proving Ground, MD andrew.rak@noblis.org

Abstract: Emerging contaminants are materials or chemicals with an evolving or increasing regulatory interest. They are characterized as chemicals (or classes of chemicals) that are currently unregulated or currently regulated but subject to change. Typically, ECs lack regulatory approved health benchmarks or standards, or the benchmarks or standards are being reviewed for potential change. ECs have a perceived or real threat to human health or environment, are of current interest from a regulatory standpoint, and have either no peer reviewed health standard or a standard that is subject to change. Also, the human health or environmental data may also be insufficient, or there may be new analytical detection limits or new exposure pathways may have been identified. Of specific interest to the Depart of Defense (DoD), are those ECs that may have a significant impact on one or more key functional areas, including training and readiness, operation assets, and acquisition of new systems . The main purpose of the information awareness task is to identify key events and information that suggests possible changes in scientific, statutory, regulatory, or guidance/policy areas that may result in changes of specific interest to DoD. The information identified can be either contaminant-specific (e.g., dioxin) or issue-specific (e.g., toxicogenomics). In addition to identification, a parallel effort is conducted to monitor developments for previously identified ECs to maintain high level of awareness regarding news coverage and ongoing regulatory, scientific, and/or policy developments. The outcome of the information awareness task is to communicate selected information to the DoD staff and the DoD’s Materials of Evolving Regulatory Interest Team (MERIT). The goal is to identify targeted relevant information that is tied to a pre-defined set of DoD functional areas and present it in a timely and concise manner for decision makers.

M5.51 Aggregating interval judgments of dependent experts -- a comparison study. Lin SW*; Yuzn Ze University shiwoei@saturn.yzu.edu.tw

Abstract: In risk analysis, people usually need to rely on subjective probabilities provided by experts. However, experts are usually overconfident about the accuracy of their judgment. Furthermore, the positive correlation between experts’ opinion can make the issue of overconfidence even worse. To assess the extent of dependence between experts’ probability interval estimates and how the dependence affects the quality of aggregated judgments, we analyzed a large data set on expert opinion compiled by Cooke and colleagues at the Technical University of Delft and elsewhere. We first used Kendall’s tau for measuring the pair-wise correlation between experts and found that the average correlation falls between 0.17 and 0.80 across different studies. This finding empirically confirms that the interval estimates provided by different experts are positively correlated. We then partitioned the data set (for each study) into different subsets, and used cross-validation to compare the performance of various aggregation methods (including Cooke’s performance-based method, a simple average, and the Bayesian approach). We found that although Cooke’s method does not explicitly account for the dependence between experts, the aggregated probability interval can achieve good calibration. Cooke’s method produces much more robust calibration scores across different studies than does the simple average, and Cooke’s method works as effectively as the more complicated Bayesian methods, even when the correlation between experts is high. Our results suggest that the use of simpler axiomatic models to combine expert opinions with dependence may still be justified.

M5.53 Study of dioxin and fine particle health risks in Helsinki. Leino O*, Tainio M, Tuomisto JT; National Public Health Institute olli.leino@ktl.fi

Abstract: Dioxins and airborne fine particles both are environmental health problems that have been subject to active public debate. Knowledge on fine particles has increased substantially during the last ten years, and even the current, lowered levels in Europe and the United States appear to be a larger public health problem than previously thought. On the other hand, dioxins are ubiquitous persistent contaminants and animal carcinogens at high doses, and therefore of a great concern. Our aim was to quantitatively analyze these two health risks and compare whether there are differences in given risks scenarios. Scenarios were chosen to match current and forthcoming EU regulations and standards for these two pollutants. We performed a comparative risk assessment for both pollutants in the Helsinki metropolitan area (Finland), and estimated the health effects for several scenarios: For primary fine particles: a comparison between the present emission situation of heavy-duty vehicles to particle emission standards set by the EU, For dioxins: an EU-directive that regulates for commercial fishing of Baltic salmon and herring that exceed the dioxin concentration limit, and a derogation from the directive for these two species. Both of these two decisions are very topical issues. We found the risk of fine particles emitted by heavy-duty vehicles clearly outweighing the risk of dioxin in Finnish fish. Substantial improvement to public health could be achieved by advancing in emission standards from present situation to stricter emission standards, about 30 avoided premature deaths annually in Helsinki. In addition, we found that benefits of fish consumption due to omega-3 exposure were hundreds of times larger than the potential dioxin cancer risk.

M5.56 The Amber Dilemma Task. Miller JC*; Air Force Research Laboratory james.miller@brooks.af.mil

Abstract: A driving simulator is used to assess risk estimation and risk acceptance in a realistic task that is dynamic and uncertain, with time constraints. The design of the task is based heavily upon the original work of Allen and Stein (ca. 1981). A 7-minute, simulated drive is accomplished in an urban area on a 6-lane street. The driver repeatedly encounters a traffic light at an intersection. Just before the driver reaches the intersection, the light turns to amber (yellow). The driver now faces a dilemma that needs a quick response: continue through the intersection or stop? Once the driver makes a decision and then stops or passes the intersection, the driver states to the investigator their pre-intersection estimate of failure, i.e., their estimate that they were or were not going to run a red light. Each drive contains one practice intersection with a long headway time when the light turns yellow and then six test intersections with randomized, shorter headways. Post-drive software provides data reduction, a display of a reward/penalty results for the run, manual data entry for the subjective estimates, and data file management. Results and test-retest reliability information will be presented for a small set of subjects who performed the task hourly for 12 hours.

M5.57 Quantitative uncertainty associated with modeling indoor vapor intrusion. Kountzman JA*; Black & Veatch Special Projects Corp. kountzmanja@bv.com

Abstract: Indoor vapor intrusion of volatile organics from contaminated groundwater is the focus of increased interest as a component of total exposure from hazardous waste sites. This can be an important pathway; however, measured indoor air concentration data is often not available. Risk assessment guidance for the evaluation of the vapor intrusion pathway, as developed by the USEPA and numerous state agencies; generally recommend that predictive modeling methods be used when measured indoor air data is not available. Due to the complex nature of volatilization from contaminated groundwater and soil, upward gaseous diffusion through the subsurface, and migration into buildings with varied structural characteristics, a considerable level of uncertainty may be associated with the available modeling methods. The purpose of this poster is to demonstrate the magnitude of uncertainty that may be associated with predictive modeling methods for the indoor vapor intrusion pathway. Quantitative estimates of uncertainty will be presented based on a comparison of risks generated from model estimates with risks generated from measured soil gas and indoor air sampling. The Johnson and Ettinger (J&E) Model was used to develop estimates of indoor air concentrations for a site contaminated with volatile organics in groundwater. Since the site does not have contaminated soils, no contribution to volatile indoor air concentrations from contaminated soil was considered. Measured groundwater concentrations of volatile organics are used as input to the J&E Model to predict indoor air concentrations. Measured sub-slab gas data are also used as input to the J&E Model to predict indoor air concentrations. Measured indoor air data from the same site is compared to the predicted results to quantify the magnitude of uncertainty associated with the model runs. Quantitative estimates of uncertainty are presented and discussed. Recommendations are presented.

M5.58 Risk assessment for industrial uses of contaminated groundwater. Stineman CH*, VanKerkhove PJ; Ecology and Environment, Inc. cstineman@ene.com

Abstract: An industrial facility in the Midwest was using groundwater found to be contaminated with trichloroethene (800 to 1,000 µg/l) and other volatile organics for several industrial purposes that had the potential to result in human exposure to the contaminants. Approximately 80% of the water was used as feed water for the facility’s 4 cooling towers to provide evaporative cooling; after being degassed to remove CO2, 15% was used to make ultra high purity water by reverse osmosis (RO) for cleaning applications; and 5% was converted to steam and fed into the facility’s air handling system for humidity control. All three applications result in transfer of the volatile contaminants to indoor or outdoor air. Since the cooling towers were located close to the facility with their exhausts at about roof level, the air around the facility itself could be impacted and drawn into facility air intakes under certain wind conditions. Other nearby industrial facilities and residences also could be affected by cooling tower emissions. The quantity and behavior (plume rise) of emissions from the cooling towers depended on the number of units in operation and whether they were operating in a forced draft or natural convection mode. EPA’s SCREEN3 air model was used to estimate air concentrations close to the facility in eddies that can result from building downwash effects; ISCST3 was used to estimate concentrations at receptor locations greater distances from the facility. The degassing system was located in a machine room and exhausted to machine room air. Air concentrations in the machine room and in the in the remainder of the facility were estimated from the volumes of water used as RO feed and for humidity control and air exchange rates determined from the operating characteristics of the air handling equipment. The methods used to evaluate these unusual exposure scenarios will be described and exposure and risk estimates presented.

M5.59 Estimating population exposure to benzene in Japan. Higashino H*, Shinozaki H, Yoshikado H; National Institute of Advanced Industrial Science and Technology haru@ni.aist.go.jp

Abstract: Exposure assessment of benzene has been carried out in Japan, taking into account the spatial distribution of the concentration and the population by using the dispersion model we have developed. The version 2.0 of AIST-ADMER (AIST’s Atmospheric Dispersion Model for Exposure and Risk assessment) has been applied to estimate the regional distribution of concentration. The ADMER 2.0 includes the newly-developed function “sub grid module” which can estimate a distribution of concentrations at the spatial resolutions of the maximum of 100 m x 100 m within the basic ADMER grid (5 km x 5 km). The emission data, derived from the Pollutant Release and Transfer Register (PRTR) system, include estimates of annual emissions from point, non-point, and mobile sources. The largest source of benzene in Japan was automobiles. Meteorological data used for the simulation were collected from national meteorological observatories and from Automated Meteorological Data Acquisition System (AMeDAS). These data were collected in 2004 throughout the country. The simulation was performed at 364 of selected basic ADMER grids throughout Japan. Every selected grid includes one or more monitoring stations for benzene. Benzene was monitored at 418 stations throughout Japan every month in 2004. Total population of selected 364 grids was estimated to be 33 million, about a quarter of Japanese population. Estimated concentrations by the ADMER 2.0 were generally consistent with measured data. The results showed the reasonable agreements within a factor of 2 were performed at almost all monitoring stations. In case of considering background concentrations (0.5 micro g/m3, derived from monitoring data analysis), 18 thousand grids of 100m x 100m throughout Japan were exceeded the Japanese environmental quality standard (3 micro g/m3), and 1.8 million people were exposed to this level.

M5.60 Derivation Of A Lead Wipe Sample Clearance Criterion For Commercial Buildings. Magee BH*; AMEC Earth & Environmental brian.magee@amec.com

Abstract: Although the U.S. EPA and HUD have derived surficial dust criteria for performing hazard assessment and clearance of residential buildings based on the protection of young children under the age of six, these criteria are applicable only to residential structures. Lead-containing dust in commercial buildings is specifically excluded from the regulations that promulgate these criteria. This paper derives a criterion for lead wipe samples in commercial buildings using de novo risk assessment techniques in accordance with standard governmental risk assessment approaches. The methodology includes the following steps: (1) Define the most sensitive adult receptor, (2) Define the target blood lead level that is protective of the most sensitive adult receptor, (3) Execute the U.S. EPA model for calculating blood lead levels in adults to define the average daily dose that leads to the target blood level, (4) Perform de novo exposure assessment calculations to estimate the reasonable maximum exposure that the target receptor would have to dust on surfaces in commercial buildings and (5) Calculate the target lead wipe level from the above. The resulting lead wipe criterion for accessible surfaces in commercial buildings that is protective of the fetuses of a population of pregnant women workers is a range from 1.11 ug/cm2 to 2.21 ug/cm2, which is equivalent to a range of 111 ug/100 cm2 to 221 ug/100 cm2 (1,032 ug/ft2 to 2,054 ug/ft2 ) For ease of implementation, the criterion range can be rounded to: 1,000 ug/ft2 to 2,000 ug/ft2 or 100 ug/100 cm2 to 200 ug/100 cm2.

M5.61 Exposure to Di(2-ethylhexyl)phthalate(DEHP)from PVC Medical Devices in Korea. Min CS, Ryu H-Y*, Lee YJ, Lee JH, Kim H-K, Hong Y-P, Jang J, Kim JC, Jang DD, Yoon H-S; National Institute of Toxicological Research, Korea Food and Drug Administration, Seoul, Korea hryu0308@kfda.go.kr

Abstract: Exposure to Di(2-ethylhexyl)phthalate(DEHP) from PVC Medical Devices in Korea C. S. Min, H..-Y. Ryu*, Y. J. Lee, J. H. Lee, H.-K. Kim, Y.-P. Hong1, J. Jang, J. C. Kim, D. D. Jang, H.-S. Yoon National Institute of Toxicological Research, Korea Food and Drug Administration, Seoul, Korea, 1School of Medicine, Chung-Ang University, Seoul, Korea DEHP is used as plasticizers of polyvinyl chloride (PVC) in the manufacture (contained 5-80% by weight) of various consumer products, such as PVC medical devices, child products, food packaging, and cosmetics. Recently, adverse health effects caused by DEHP exposure through PVC medical devices have been issued and the policy decision on the necessity of the continuous usage has been considered in Korea. The purpose of this study is to investigate the exposure level of DEHP among the patients who might have high risk of DEHP exposure through the PVC medical devices and to provide information on the additional exposure factors contributing to the DEHP exposure. From 778 subjects(healthy control(HC) n=319, outpatient control(OC) n=171, patients using infusion solution(ISUP) n=128, and hemodialysis patients(HP) n=160), blood samples were collected and epidemiological surveys were conducted with questionnaires including demographic and life-style related factors. The mean values(±S.E.) of DEHP levels(ng/ml) in plasma were 1.43 ± 0.11, 3.57 ± 0.24, 5.77 ± 0.37, 11.33 ± 0.49 in HC, OC, ISUP, and HP groups, respectively. The DEHP levels in HP and ISUP groups were significantly higher compared to those in OC and HC controls(p<0.0001). In addition, the statistical significance remained after adjustment of the confounders to the DEHP exposure(p<0.0001). These results suggest that patients exposed to the PVC medical devices have higher risk of DEHP exposure than general population. Therefore, risk assessment for DEHP from PVC medical devices is warranted for the decision on making of public health policy in Korea.

M5.62 Evaluation of copper and nickel exposure and risk of adverse health effects in people living near a Cu-Ni smelter in SW Finland. Kollanus V*, Komulainen H, Tuomisto JT; National Public Health Institute, University of Kuopio virpi.kollanus@ktl.fi

Abstract: Metals are natural and ubiquitous constituents of the environment and many of them function as essential trace elements in humans. However, at high level of exposure all metals can cause toxicity and unacceptable health risks. It has been demonstrated in previous studies, that the people most at risk from environmental exposure perspective are often those living near metal industry sites. The objective of this study is to determine the daily intake and uptake of copper and nickel via environmental media and to evaluate the consequential risk of adverse health effects in people living in the surroundings of a copper and nickel smelter located in South-West Finland. In order to define the degree and quintessential sources of uncertainty and variability in the assessment, exposure is calculated probabilistically by utilising Monte Carlo simulation. The average daily exposure in adults and children is determined indirectly based on a synthesis of environmental data collected from the area in previous studies, as well as on parameter values and probability densities retrieved from the literature. All relevant exposure pathways and routes are included in the assessment. The risk of adverse health effects is evaluated by comparing the calculated daily intake to the reference values considered to represent a safe level of intake. Preliminary worst case results indicate that in the most polluted areas the average daily intake of copper may come close to or even exceed the level of tolerable daily intake (TDI, 140 µg/kg/d), especially for children. The determinant sources of exposure appear to be the ingestion of soil and home grown food plants. For nickel, the estimated daily intake is unlikely to surpass TDI (50 µg/kg/d). However, nickel is a known respiratory tract carcinogen in occupational settings, and cancer risk via inhalation of Ni containing particles in ambient air needs also to be assessed.

M5.63 Follow Up Study of Air Pollution in Residential Area and School Environments Using GIS Related to Lung Function among School Children. Lin MH*, Ho WC, Shiao GM, Hartley WR, Lin RS; CHINA MEDICAL UNIVERSITY linmh911@yahoo.com

Abstract: Air pollution is an important risk factor related to lung function. Exposure assessment is one of the key steps in assessing human health risk. Based on the high density of air monitoring stations in Taiwan, this study was using Geographical Information Systems (GIS) to assess the air pollution exposure in household and school environments. The health risk would focus on lung function among school children with one year follow-up. Geostatistical procedure was using Inverse Distance Weighted (IDW) estimate method in 500m × 500 m grid squares. The study population was random sampling junior high school students to complete both baseline and follow-up lung function tests and standard questionnaires in Taipei area. The lung function tests followed the 1977 criteria of the American Thoracic Society (ATS) Snowbird Conference, and the ATS update statement of 1987. To assess household and school environments related to its effects on lung function, the regression model was used. The controlling variables included gender, age, body weight, body height, parental education level, exercise, cigarette smoking, environmental tobacco smoking, alcohol beverage consumption, child’s bedroom, and allergens. The result showed that the increasing concentration of air pollution (ozone and nitrogen oxides) was related to decreased lung function (FVC and FEV1). It is important to assess children exposed to air pollution in both school and household environments to protect children from the risk of respiratory health.

M5.64 Site-Specific Assessment Based on Inhalable Fraction of Nickel Species in Ambient Air. Tsunemi K*; Research Center for Chemical Risk Management, National Institute of Advanced Industrial Science and Technolog k-tsunemi@aist.go.jp

Abstract: Measuring the total amount of a metal without regard to speciation may introduce uncertainties into inhalation exposure estimates. Metal speciation affects a range of processes that change how the metal is deposited in the respiratory tract and subsequently, distributed throughout the body and, consequently, its potential toxicity (U.S. EPA, Framework for Inorganic Metals Risk Assessment, 2004). Thus, in this study, site-specific risk assessment was conducted based on inhalable fraction of nickel species in ambient air. First, distribution of nickel species in ambient air was analyzed in five areas with high concentration of nickel in Japan. Second, the result of the distribution of nickel species was compared with existing nickel speciation data for specific sites of nickel sources. Thus, the main sources of nickel emission were identified in each area by the qualitative relationship between nickel sources and distribution of nickel species. Third, exposure levels for each nickel species were expressed based on measurements of the total nickel exposures, taking into account the analyzed speciation data. Fourth, cancer unit risk and NOAEL of non-cancer toxicity through inhalation of each nickel species were derived from the toxicological data of animal tests and epidemiological studies. Finally, site-specific risk of human health was calculated for risk characterization. As a result, there is no significant non-cancer risk to humans exposed by inhalation in all areas, however, cancer risk was calculated above 10E-5 in some areas. Therefore, we identified the industries in these areas which should examine to reduce nickel emission.

M5.65 Formal Uncertainty Analysis in the Interpretation of Organophosphate Pesticide Metabolite Concentrations. Vigoren EM*, Griffith WC, Krogstad FTO, Coronado GD, Thompson B, Faustman EM; University of Washington gorst@u.washington.edu

Abstract: As part of a community intervention study to reduce organophosphate pesticide (OP) exposure in farmworker populations, we measured urinary OP metabolite concentrations in urine from farmworkers and their children and OP residue concentrations in dust collected from the farmworkers’ homes and vehicles. Analyses of these pesticide biomarker data were aided by a formal approach to uncertainty in which we specified three types of uncertainty in the data. Structural uncertainty was observed in the need to classify subpopulations. Farmworkers that worked in apple and/or pear orchards (pome fruit) had higher exposures than those who did not work with these crops. Parameter uncertainty was apparent in shifts in the distribution of the metabolite and residue concentrations over time. These shifts may have been due to changes in OP usage (partly as a result of regulatory changes), timing of collection (vis-à-vis OP application), an actual reduction in OP use, and/or improved personal protective practices. Metabolite and residue concentrations alone are not sufficient to answer these questions, but they can direct the path of further analysis. Stochastic uncertainty was omnipresent in the variability that we observed within and between individuals. Increased sample size could improve our understanding of, but not necessarily reduce, this uncertainty. Classifying and modeling the types of uncertainty present in the analysis of pesticide biomarkers can help to formulate how we incorporate and interpret uncertainty in the analysis of our dataset. With a better understanding of uncertainty, we can clearly identify specific factors that contribute to exposure changes, especially for sensitive populations such as children. This will help us to more clearly communicate the results and minimize ambiguity in conveying information to decision-makers, stakeholders and the community. [This publication was made possible by grants PO1 ES009601 from NIEHS, and EPA RD-83170901.]

M5.66 HALF-LIVES OF DIOXINS, FURANS, AND PCBS AS A FUNCTION OF AGE, BODY FAT, BREASTFEEDING, AND SMOKING STATUS. Milbrath MO, Chang C-W, Emond C, Franzblau A, Garabrant D, Gillespie B, Wenger Y, Adriaens P, Jolliet O*; University of Michigan, School of Public Health ojolliet@umich.edu

Abstract: It is necessary to know the half-life of each congener of polychlorinated dibenzodioxins (PCDDs), polychlorinated dibenzofurnans (PCDFs), and dioxin-like polychlorinated biphenyls (PCBs) in the body to estimate the contribution of past exposures on current serum concentrations. A literature search was conducted for measured and modeled kinetic values for 29 congeners. Sources that provide consistent data for numerous congeners and that are within the range of reported values are used to determine reference half-life values for an infant and for a middle age male.The half-lives of polychlorinated dibenzodioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and polychlorinated biphenyls (PCBs) exhibit inter-individual variation based on different individual characteristics. Measured and modeled data from the literature were then compared to examine the effects of age, percent body fat, and total body fat on the half-lives of 29 different congeners. Two strategies are proposed for adjusting half-lives for an individual based on a predefined reference value. An equation based on a linear relationship with age that is corrected for percent body fat, smoking status, and time spent breastfeeding a child provides results that are consistent with previously reported observations. Alternatively, an equation based on a linear relationship between half-life and total body fat adjusted for smoking status and time spent breastfeeding a child is proposed. This method requires further testing and validation against individual measurements.

M5.67 Minimizing risk of exposure to pharmaceuticals in the environment. Ruhoy IS*, Hassenzahl DM; UNIVERSITY OF NEVADA, LAS VEGAS ilene.ruhoy@unlv.edu

Abstract: The disposal of unused pharmaceuticals has caused increasing concern among health care professionals, environmental scientists, policymakers, and the public. Studies have demonstrated levels of drug residues in water systems, and have also found adverse effects on aquatic organisms, such as altered endocrine and neurologic physiology. It is unknown whether this poses a health hazard for humans through exposure from drinking water and food substances. Considering the multitudes of pharmaceuticals used daily and the varying levels of exposure possible, a risk assessment is complicated. This paper discusses the extent to which we can hope to lessen and modify the risk to humans. The discussion focuses on four general strategies for risk abatement (Morgan, 1981): modify the environment, modify exposure processes, modify effects processes, and mitigate or compensate for effects. Several environmental modifications are possible. Wastewater treatment plants may serve to rid the waters of most, but not all, active pharmaceutical compounds. Not all WWTPs maintain technology to be maximally effective. Alternatively, medications can be disposed of in manners other than sewerage. Prescription practices can be better monitored to prevent accumulation of medications, thereby discounting the need for disposal. Exposure is prevalent from many different facets of daily life. Beyond WWTPs, there are limited means by which to modify exposure and effects processes. It should not be recommended for consumers to decrease the water or food they consume. Finally, the effects can be mitigated for by providing incentives for drug manufacturers to be good product stewards. Further research is needed to develop disposal and water treatment technology. Consumer and health care professional education can serve to assess a patient’s potential exposures. The nature of this risk suggests a front-loaded risk abatement strategy. Modification of the environment holds the greatest potential for minimizing risk to humans from pharmaceutical compounds in the environment.

M5.68 Integrated risk analysis model to estimate health effects due to primary fine particle (PM2.5) emissions from different sectors. Tainio M*, Karvosenoja N, Sofiev M, Karppinen A, Kupiainen K, Kangas L, Tuomisto JT, Pekkanen J, Kukkonen J; National Public Health Institute (KTL), Finnish Environmental Institute, Finnish Meteorological Institute, University of Kuopio marko.tainio@ktl.fi

Abstract: Air pollution, especially fine particle (PM2.5), are causing significant adverse health effects worldwide. The development of reasonable mitigation strategies requires complex integrated models that combine action, emissions, exposure, and health effects. The scope of the study is to build an integrated risk analysis model to estimate Finnish fine particles emission, combine it with European inventories and evaluate the associated health effects in European countries. The integrated model combined information from emissions, dispersion, exposure, and health effects of Finnish primary fine particles. The emissions were estimated based on activity and emission factor information from 20 pollution source categories. The dispersion of particles was modeled using the Air Quality and Emergency Modeling System SILAM with 10 and 30 km spatial resolution over the Europe. The resulted concentration maps were combined with population data to obtain exposure estimates for six main pollution source categories. The exposure model was based on intake fraction (iF) method. The health effects were predicted using concentration-response estimates from U.S. epidemiological cohort studies. All the information was combined to one integrated risk model. Two pollution source categories, domestic combustion and traffic, were estimated to cause most of the primary fine particles health effects in Finland. The intake fraction for fine particles from traffic was notably larger than that for other main pollution source categories. Over half of the associated health effects were predicted to occur in Finnish population. The integrated risk model was able to estimate the most significant pollution source categories, traffic and domestic combustion, and guide both decision making and further research.

M5.69 Evaluation of Biomonitoring Data for Dioxin-Like Compounds in Workers at a Primary Magnesium Production Facility. Haws LC*, Scott LLF, Staskal DF, Harris MA, Finley BL; ChemRisk, USA lhaws@chemrisk.com

Abstract: In 2004, the National Institute of Occupational Safety and Health (NIOSH) conducted a Health Hazard Evaluation (HHE) at a primary magnesium production facility in Rowley, Utah. The HHE included blood testing for dioxin-like compounds (DLCs) in 30 workers performing jobs believed to have the greatest potential for exposure and who had the longest job tenure. Blood concentration data for these individuals was obtained from NIOSH and evaluated using two techniques: 1) comparison of mean and maximum concentrations in the HHE report to published reference ranges, and 2) comparison of individual results to the 95th percentile and maximum value for 2001-2002 male NHANES participants aged 40-59 years. The goal of these analyses was to determine if dioxin concentrations in the workers was similar to what is observed in the general US population and to determine if the two techniques yielded similar results given that dioxin concentrations have been shown to vary based these demographics. When summary statistics from the HHE were compared to published reference values, the mean lipid-adjusted TEQ for the 30 workers exceeded the mean reference values, while the maximum value for the workers was within the range of published reference values. For example, the TEQ in the USM cohort ranged from 12 to 147 ppt (mean=37.3 ppt), whereas the TEQ of the general U.S. population based on NHANES data ranges from 8 to 208.1 ppt (mean=30.4 ppt). However, comparison of the individual workers to male NHANES participants aged 40-59 years revealed that while the majority of workers in the USM cohort were below the 95th percentile of the NHANES, only 7 individuals had TEQs between the 95th percentile and maximum value. In conclusion, rigorous evaluation of the HHE data indicates that site-related exposures to DLCs do not appear to result in increased dioxin body burdens in these workers.

M5.70 Benzene in soft drinks and other beverages: Do measured levels pose a human health risk? Tachovsky JA*, Scott LF, Haws LC, Williams ES, Harris MH; Chemrisk atachovsky@chemrisk.com

Abstract: A recent study by the Center for Food Safety and Nutrition (CFSAN) reported that several beverages contained levels of benzene at levels that exceeded the EPA drinking water standard of 5 ppb. One of the beverages, Crystal Light Sunrise Classic Orange (CLSCO), was reported to have concentrations as high as 87.9 ppb, though a newer formulation registered levels <1 ppb in CFSAN testing. A stratified random sampling approach was developed and samples of CLSCO were obtained from retail outlets in the Houston, Texas area. Three formulations were obtained during sampling: 16 oz. old and new formulations, and 64 oz. formulation. Analysis of the samples revealed mean benzene concentrations of 90 ppb in 16 oz. old formulation bottles, 0.24 ppb in the 16 oz. new formulation product, and 3.38 ppb in the 64 oz. product. Point estimates of risk and hazard were developed using two techniques. The first technique was an upper bound calculation using average body weights and a 95% upper confidence limit on the mean benzene concentration. The second technique allowed exposure to different formulations of CLSCO at different times and varied body weight with age. A probabilistic risk assessment was also conducted, using distributions for benzene concentration and exposure parameters. The results indicate that point estimates developed based on allowing variation in formulation and body weight match upper bound point calculations for the maximally exposed scenario and allow examination of different exposure scenarios. Teen risk and hazard index was slightly greater than adult risk and hazard, but probably not outside the range of uncertainty. Monte-carlo simulations show probabilistic risk estimates at 95% of the risk distribution tend to overestimate upper bound point estimates performed using the 95% upper confidence limit of the mean. In all cases, cancer risk and non-cancer hazard associated with ingestion of benzene in CLSCO are below levels of concern.

M5.71 Use of Multi-pathways Intake Fraction and blood half-live to determine adequate Toxic Equivalency Factors (TEF) to weight emissions and blood concentrations. Jolliet O*, Soucy G, Wenger Y, Deschênes L, Dettling J, Humbert S, Margni M; University of Michigan, School of Public Health ojolliet@umich.edu

Abstract: It is common to use Toxic Equivalency Factors (TEF) to compare emission of various congeners of Dioxin-Like compounds or PAH. Based, on dose-response experiment, these TEF are adequate to compare intake, but are also often used innapropriately to produce TEQ equivalent emissions or to compare serum concentrations of various congeners. The present paper develop a method to consistently compare both emissions and serum concentrations and applies it to the 29 WHO dioxin-like compounds and the 16 PAH's. To relate emissions to intake, a multi-scale multimedia fate and exposure model is being developed with high resolution in the great Lake area, to determine Intake Fractions - the fractions of an emission that is taken in by the obverall population. For PAH's it shows that 3% of emissions correspond to 53% intake which account for 98-99% of toxic Equivalents. It demonstrates that it is essential to account for both t for TEFs and intake fraction to weight PAH emissions. For serum concentration, it is important to relate intake to serum concentration. This is achieved by accounting for half-lives and pharmacokinetic in the body. The method is illustarted using the 29 Dioxin-congeners, leading to variation of up to one order of magnitude compared to use of TEF only.

M5.72 Case studies for the Development of a Pathophysiological Progression Model. Dodge DG*, Zhao QJ, Haber LT, Goodman JE, Pagan I, Foureman GL, Rhomberg LR; Gradient Corporation; Toxicology Excellence for Risk Assessment (TERA); US Environmental Protection Agency ddodge@gradientcorp.com

Abstract: U.S. EPA's National Center for Environmental Assessment is developing approaches to identify/categorize toxicological endpoints along a pathophysiological continuum, using the framework presented by Schulte and colleagues (exposure, internal dose, biologically effective dose, early biological effect, altered structure/function, and clinical disease) as a starting point. Two critical effect endpoints – fatty liver (steatosis) and pulmonary inflammation – have been developed as case studies for population of the model. Steatosis is an accumulation of triglycerides in hepatocytes and occurs when the rate of import or synthesis of fatty acids by hepatocytes exceeds the rate of export or catabolism. Carbon tetrachloride, for example, is believed to affect export by reducing the level of very low-density lipoproteins (VLDL), the major transport vehicle for triglycerides out of the liver. Steatosis can sensitize the liver to further injury such as inflammation, fibrosis, and cirrhosis. The endpoint selected for the second case study, pulmonary inflammation, is an important airways response that can be induced and/or exacerbated by certain inhaled toxicants. An example chemical is used to illustrate a progression of effects from different degrees of inflammation to more severe pulmonary interstitial disease such as fibrosis. For both the steatosis and pulmonary inflammation case studies, biomarkers of exposure and effect are also discussed. These case studies include descriptive criteria and graphical approaches for placing the selected endpoints along a pathophysiological continuum from no effect to adversity.

M5.73 Considerations Regarding the Structure and Application of a Pathophysiological Progression Model . Haber LT*, Rhomberg LR, Goodman JE, Dodge DG, Zhao QJ, Pagan I, Foureman GL; Toxicology Excellence for Risk Assessment (TERA); Gradient Corporation; US Environmental Protection Agency ddodge@gradientcorp.com

Abstract: A pathophysiological progression model has been described by Schulte and colleagues (exposure, internal dose, biologically effective dose, early biological effect, altered structure/function, and clinical disease). Although this model has been very useful in organizing thinking about biomarkers and observed endpoints, we are considering additional enhancements to improve its application in risk assessment. Identification of the appropriate “box” to place specific endpoints along a continuum has traditionally been based on professional judgment and weight of evidence considerations. The existing framework would benefit from a more rigorous approach and specified rationales. It may be useful to establish subcategories within the continuum boxes that allows for considerations such as level of organization (gene, molecule, organelle, etc.), exposure duration required to observe an effect, reversibility, and whether the endpoint represents adaptation. As part of this work, we are also evaluating alternative structures for categorizing endpoints. For example, systems biology focuses on feedback loops and considers toxic effects as deviations from normal feedback and homeostatic mechanisms. An enhancement being considered is the incorporation of the dimension of time (since the progression reflects a causal series of relationships, rather than a specified time progression). Another alternative approach is to consider whether the biomarker is upstream or downstream of the target, or whether it reflects an effect with a common cause. Case studies supporting these analyses are presented in the accompanying poster by Dodge et al.

M5.74 What Dose of Chloral Hydrate Can Be Really Hazardous to Human Health? Byczkowski JZ*; Consultant, Fairborn, Ohio januszb@aol.com

Abstract: Chloral hydrate is a ubiquitous by-product of water chlorination and it is produced in the organism as ultimate metabolite of trichloroethylene. It is also one of the oldest of synthetic drugs, still in use principally as a hypnotic in the treatment of insomnia and as a routine sedative. Even though, it is considered a safe drug, an overdose - as with any systemic drug - may be lethal. Based on in vitro genotoxicity assays and rodent studies, it was suggested that chloral hydrate may be classified as a "possible human carcinogen". The aim of the present report was to reconstruct a dosimetry of chloral hydrate from some well defined, notorious cases of human poisoning (based on available autopsy information) and from the recent study that examined the relationship between chloral hydrate exposure and cancer risk in humans. The data were analyzed using a computer-aided physiologically based pharmacokinetic (PBPK) model, originally developed with data from mice and dogs, and re-calibrated to reflect human physiology. The results of modeling are discussed in relation to current reference dose (RfD) for acute and chronic exposures to chloral hydrate.

M5.75 Deriving Skin Notations Based on the Improved Strategy Developed by the National Institute for Occupational Safety and Health. Dotson GS, Gadagbui BK*, Kroner O, Nance P, Lentz TJ, Maier A; Toxicology Excellence for Risk Assessment (TERA) maier@tera.org

Abstract: Industrial hygienists have traditionally relied on inhalation-based exposure guidance (occupational exposure limits or OELs) for assessing risks to worker health. These OELs do not directly account for potential skin exposures. To warn workers of potential impacts of dermal contact, skin notations are assigned for chemicals that can be absorbed through the skin. NIOSH has developed an improved strategy for assigning these skin notations. This new strategy was developed to address limitations associated with the current skin notations and to provide a systematic approach for assigning such notations. The new system relies on a weight-of evidence approach based on the findings of an extensive review of reports of human exposure and health effects, empirical data from animal toxicity studies and considerations provided by mathematical models. The modeling component includes both skin absorption estimation as well as structure-activity predictions for skin irritation and sensitization. Benefits of the new system include improved consistency in skin notation assignments through the use of more fully-defined criteria, and the inclusion of information that distinguishes between systemic and point-of-contact health effects associated with dermal exposure to chemicals. Additionally, the strategy has been designed to improve the use of skin notations beyond their historical interpretation by providing supplement information on the potential for point-of-contact effects, which enhances efforts to protect workers’ health. Scientific and data interpretation issues associated with the application of this revised methodology have been identified in its application for over 20 case-study chemicals. Key issues have centered on assessing adequacy of data sets and the definition of cut points and criteria for assessing “significant” dermal absorption and systemic toxicity.

M5.76 Evaluation of low-dose effects of acrylamide in the mouse micronucleus assay. . Friedman MA, Zeiger E*, Recio L, Swenberg J, Fennell TR, Haseman JK; Errol Zeiger Consulting killrat@bellsouth.net

Abstract: Acrylamide is an industrial chemical that is also formed in substantial levels in cooked food, with daily dietary intakes estimated at 0.075 mg. In order to determine whether acrylamide risk should be assessed using a linear extrapolation through zero, a non-linear (i.e., quadratic) model, or by a threshold model, an extensive mouse micronucleus (MN) assay was conducted. Groups of 10 mice were treated with 0, 0.125, 0.25, 0.5, 1.0, 2.0, 4.0, 6.0, 8.0, 12.0, 16.0, and 24.0 mg/kg daily for 28 days, and MN were measured in peripheral blood erythrocytes. Above 2 mg/kg the responses increased linearly with dose. MN-PCE in peripheral blood increased from a vehicle control level of 4.0/1000 to 5.5/1000 at 24 mg/kg; the corresponding increases in MN- NCE were from 1.7 to 2.2/1000. The same animals were also used for measurements of hemoglobin (Hb) adducts formed from acrylamide and its metabolite, glycidamide, and liver DNA adducts. There was no statistically significant increase in the incidence of micronuclei in peripheral blood PCE or NCE at doses up to 2 mg/kg. Regression analysis performed on the log-dose scale revealed significant non-linearity and a response that was consistent with a threshold dose of approximately 2 mg/kg. When this analysis was conducted on the dose scale, the MN response was consistent with a linear model, but was also consistent with a threshold model, with a threshold dose of 1-2 mg/kg. Statistical analyses are being performed to relate MN frequency as a function of Hb adducts (a measure of internal dose). The relationships among MN, Hb adducts, and liver DNA adducts will also be examined. The implications of these types of data analyses are critical to assessment of the public health risk from acrylamide exposure. This study was supported, in part, by the Polyelectrolyte Producers Group.

M5.77 Use of exposure markers located downstream of the target organ may result in a misleading quantitative risk assessment. Gamo M*; Advanced Industrial Science and Technolgy (AIST) masashi-gamo@asit.go.jp

Abstract: In the risk assessment of chemical substances, although the exposure levels are usually measured as the amount of intake or concentration in the target organ (or blood), surrogate exposure markers such as concentration in hair or urine are also occasionally used. In this study, we demonstrate that the use of such surrogate exposure markers that are located downstream of the target organ may result in a misleading quantitative risk assessment. This error arises from the interindividual variability in the relationship between the concentrations in the target organ and the surrogate exposure markers. It has been suggested that in an epidemiological study, the total prevalence rate in the subject population affects the dose-response relationship obtained on the basis of surrogate exposure markers. That is, a higher prevalence rate in the population would result in a dose-response relationship that is steeper and shifted to the left. This implies that the dose-response relationship obtained in a population is not applicable to the assessment of another population that has different exposure levels and therefore a different total prevalence rate. Further, the assessment would be overprotective if an uncertainty factor is applied in order to take into account the interindividual variability between the concentration in the target organ and the surrogate exposure marker. Case studies on the risk assessment of cadmium and methylmercury will be discussed.

M5.78 Simulating the effects of polymorphisms and abundance of CYP-450 enzymes in metabolism of chlorpyrifos. Griffith WC*, Ramaprasad J, Faustman EM; University of Washington griffith@u.washington.edu

Abstract: The initial step in the metabolism by CYP450 enzymes of the organophosphate pesticide chlorpyrifos (CP) typically involves conversion to either a more toxic CP-oxon or to a relatively non-toxic compound, TCP. Four different CYP450 enzymes account for the majority on the metabolism,1A2, 2B6, 2C9, and 2C19, and each has different rates of production of CP-oxon and TCP. The rate at which these reactions occur has been the focus of a great deal of recent research, since these rates control the amount of the CP-oxon produced and how quickly it is detoxified. Genetic variability in the enzymes involved in the metabolism can greatly alter these rates. We look at the genetic and phenotype variability in the CYP-450s using recently published data mined from the scientific literature to simulate the potential role of genetic polymorphisms in reducing reaction rates of the CYP450s and variability in the abundance of individual CYP-450s. These simulations indicate that variability in the abundance of the individual CYP-450s is likely to have as large an effect on variability in the amount of CP-oxon produced as are genetic polymorphisms of CYP-450s. This occurs because all four enzymes exert an effect on the rate of CP-oxon production and when a polymorphism occurs in one CYP-450 the other enzymes are still effective in producing CP-oxon. . Support for this research was from Center for Child Health Risk Research, R 826886-01-0 and P01 ES09601.

M5.79 Age-related intra-human variability in nasal dosimetry. Kimbell JS*, Garcia GJM, Stanek J, Foureman GL; The Hamner Institutes for Health Sciences, Research Triangle Park, NC and National Center for Environmental Assessment, U.S. EPA, Research Triangle Park, NC kimbell@thehamner.org

Abstract: Nasal extraction of inhaled, soluble, reactive gases is expected to be moderate to high. Default dosimetric methods adopted by the U.S. EPA assume that such gases are completely extracted by the nose and uniformly distributed over the entire nasal surface area. A previous study examined these assumptions by simulating inspiratory nasal uptake using three-dimensional (3D) computational fluid dynamics (CFD) models based on MRI scans of the nasal passages of four adults. This study has been extended to children. 3D nasal passage reconstructions were created for a 7-yr-old boy and an 8-yr-old girl from CT scans using Mimics (Materialise, Ann Arbor, MI). Resting minute volumes (cyclic) for the 7- and 8-yr-old were calculated from estimated body masses (24.0 and 27.2 kg, respectively) to be 5.5 and 5.8 L/min, respectively. Maps of localized wall mass-flux of the gas were generated for theoretical conditions of high (maximal) and moderate (80%) nasal extraction (NE). Predicted high NE was estimated at 92% in the 2 children and 87% in the 4 adults. In the high NE case, peak values of simulated wall mass-flux were up to 14 times higher than average values in adults and 8 times higher than average flux values in the children. A similar trend was predicted in the moderate NE case with peak flux values predicted to be up to 12 times higher than average values in adults and 4 times higher in the children. In both NE cases, average nasal flux with the parameters used was predicted to be two-fold higher in the children than in the adults, while the children’s peak local flux was of the same magnitude as the adults’ value. These results in both adults and children challenge the assumptions of complete extraction and uniform distribution in the nose for such gases. This research was funded by the U.S. EPA (EP06C000405).

M5.80 A Stratified Meta-Analysis of Source-Specific Particulate Matter (PM10) and Its Health Effects. Li Y*; University of North Carolina at Chapel Hill liying@email.unc.edu

Abstract: Earlier epidemiological studies have consistently found that airborne particulate matter less than or equal to 10 µm in diameter (PM10) can affect people’s health in many ways, such as cardiopulmonary diseases and related premature deaths. In recent years, studies are increasingly addressing the sources of particles as a factor that plays an important role in the health effects. For instance, some argued that particles from different sources are not equally toxic to the exposed population, but some of them such as particles from combustion-related sources may pose a greater health risk. Given this evidence, it is not appropriate to assume that every pollution abatement strategy has the same health impact per unit change in the mass concentrations. However, currently the source-specific PM10 health risks are not yet sufficiently understood due to the difficulty in source apportionment. To explore the source-specific PM10 health risk quantitatively based on earlier epidemiological evidence, this study conducted a stratified meta-analysis by grouping similar time-series studies and pooling their estimates of the association between short-term exposure to PM10 and mortality. The sources of particles are classified into three categories: mobile, industrial and crustal. Separate meta-analyses were conducted within each source group. The pooled estimates were: a 10 µg/m3 increase in PM10 from mobile sources were associated with a 1.22% increase in daily mortality, whereas an equivalent increase in PM10 from industrial and crustal sources were associated with 0.58% and 0.55% increases in daily mortality, respectively. The results of this study suggest that particles from mobile sources may pose the greatest health risk, but particles from industrial and crustal sources are not ruled out either. Given this, PM10 regulation strategies should not only focus on a national ambient standard, but also particle sources, particularly traffic-related sources.

M5.81 Development of Methodologies for Cancer Risk Estimation from the in vitro genotoxicity Test. Nakayama A*, Sasamoto Y, Fukami T, Morisawa S; Kyoto University nakayama@risk.env.kyoto-u.ae.jp

Abstract: It is an important and urgent issue to evaluate the risks of chemicals in environment. Basically animal based researches are performed to evaluate human risks. However, it costs time and money to obtain sufficient toxicological data for most chemical agents. Risk assessment would never be completed before some agents actually affected human health and/or the ecosystem. Another problem is that animals do not react in the same way to chemical agents in their absorption, distribution, metabolism, elimination (ADME) and response as humans. Extrapolation of lab animals to humans causes uncertainty in risk estimation because of the species differences. Here in this study we propose a new framework of quantitative cancer risk estimation by applying an in vitro transformation assay. The framework consists of four parts: 1. In vitro transformation assay 2. Dose-response analysis 3. Predicition of target tissue dose with a physiologically based pharmacokinetic model. 4. Evaluation of tumorigenic transformation probability at the target tissue. By using this methodology, we estimated the risk of liver tumor induced by DDT (dichloro-diphenyl-trichloro-ethane). The estimated slope factor was about 0.1(mg/kg/day)^-1, which is about third part of U.S. EPA’s estimation.

M5.82 Development of a Pathophysiological Progression Continuum Model for Selected Toxicological Endpoints. Pagan I*, Haber LT, Rhomberg LR, Goodman JE, Dodge DG, Zhao QJ, Foureman GL; US Environmental Protection Agency Pagan.Ines@epa.gov

Abstract: The Environmental Protection Agency (EPA), Office of Research and Development (ORD), National Center for Environmental Assessment (NCEA) is currently developing methodologies and information sources to use as guidelines for exposure to various compounds, especially toxic air pollutants. This project is being proposed as a means to advance our understanding on grading endpoint severity that may be used in making estimates of safe exposures, with a focus on inhalation exposure scenarios. The goal of this project is to develop a model to identify/categorize toxicological endpoints along a pathophysiological progression continuum (from no effect to adverse effect). The characterization of toxicity endpoints with regard to severity and placement along a pathophysiological continuum will better inform the use of these endpoints in human health risk assessments. An endpoint is defined in this presentation as an objective or subjective measure that serves as an indicator of a response (normal, adaptive and/or pathological) to chemical/toxicant exposure. A compendium containing the categorization of endpoints along a pathophysiological continuum would serve to do the following: (1) weigh the relative importance of effects induced by a chemical or across chemicals, (2) provide information valuable in the determination of species sensitivity and relevance of animal effects to humans, (3) provide interpretative, descriptive and graphical information that can aid in the standardization and uniformity of chemical health assessments, and (4) serve as a reference source for chemical-mode of action (MOA) characterization. The diagram proposed by Schulte and colleagues including exposure, internal dose, biologically effective dose, early biological effect, altered structure/function, and clinical disease is being considered as one of the graphical models for representation of selected endpoints along the pathophysiological continuum.

M5.83 Use of Animal Acute Toxicity Data to Derive Immediately Dangerous to Life or Health Concentrations: Extrapolating to Human Effect Thresholds . Parker A*, Maier A, Gadagbui BK, Dotson GS; Toxicology Excellence for Risk Assessment (TERA) maier@tera.org

Abstract: Immediately Dangerous to Life or Health (IDLH) concentrations have a long history of use in industrial settings in defining work practice and respiratory protection requirements for entry into potential high exposure environments. We explored approaches for extrapolating from acute toxicity data in animals to estimate human effect thresholds that might serve as the basis for deriving an IDLH value. Ratios of animal to human effect levels were compared. Based on our analysis we propose refining the current use of a default factor of 10 for extrapolating to an IDLH value from either an LC50 (the concentration estimated to cause a 50% mortality rate in an acute toxicity study) or an animal lethality threshold (commonly estimated as an animal LCLO, LC10, etc.). We also examined the effect(s) of various modes of action on ratios of animal-to-human effect levels. We hypothesized that potent irritants would have a greater LC50/human serious effect threshold ratio than other chemicals. Our results were mixed with a significant mode of action effect observed for a subset of 20 chemicals, but not in a broader analysis of current IDLH values. Approaches for using other indicators of irritant potential were also investigated. Lessons learned from this research effort were applied to the derivation of new IDLH values for several chemicals of interest to homeland security applications.

M5.84 Modeling illness endpoints for pathogen dose response. Swartout J*; U.S. EPA swartout.jeff@epa.gov

Abstract: There is a significant body of literature addressing microbial dose-response modeling, but primarily for the infection endpoint. There are relatively fewer papers presenting modeling for illness endpoints. While infection is a necessary precursor to illness and is of importance for secondary disease transmission in the population, illness endpoints are key to benefit-cost assessment and determination of risk management options. Illness response modeling is more complicated than infection modeling, as it must be conditional on infection, but may not be dose dependent in the same manner as for infection. The risk of illness usually is presented as a morbidity ratio, which is the probability of illness in the infected population. Morbidity ratios either can be fixed (constant over dose) or dose-dependent (increasing with dose). In the former case, although the probability of illness will vary for individuals, there is an average morbidity ratio for the population that is independent of dose. In the latter case, the probability of illness increases with dose as it does for infection, but still may not reach 100%. Unconstrained modeling of illness directly as a function of dose often can result in erroneously high risks. In some cases, however, the observation of illness may be more certain than the measure of infection, such that direct modeling of illness may be more reliable. Methods are presented for constrained modeling of illness for both fixed and dose-dependent morbidity ratios, and for unconstrained illness modeling, with consideration for a resistant fraction of the infected population.

M5.85 A Quantitative Risk Assessment Model of Cigarette Mainstream Smoke Constituents. WILSON CL*, POTTS RJ, BODNAR JA, GARNER CD, BORGERDING MF; Research and Development, RJ Reynolds Tobacco Co, Winston-Salem NC wilsonc5@rjrt.com

Abstract: A quantitative cancer risk assessment of mainstream smoke (MSS) constituents was performed for thirty-five (35) U.S. commercial cigarette brands. Products represented a broad range of “tar” yields (0.9 to 19 mg) as determined by machine smoking using the Federal Trade Commission puffing regimen. Cigarettes were machine-smoked using an intense puffing regimen (60 ml puff volume every 30 seconds with 2 second duration; 50% vent blocking) and yields of twenty-three (23) individual MSS constituents measured. The constituents studied in the risk assessment were selected based on their quantitation in MSS, evidence of being a known or suspected human carcinogen, and availability of cancer potency data. Compounds meeting these criteria fell into a range of chemical classes, including polynuclear aromatic hydrocarbons, semi-volatile organics, heavy metals, tobacco-specific nitrosamines, aromatic amines, and carbonyl compounds. Assumptions included in the point estimate model include 95th percentile daily cigarette consumption data from the 1999-2002 National Health and Examination Survey (NHANES) and exposure duration, averaging time, and lifespan consistent with current cancer risk assessment practices. Cancer risk estimates for the individual MSS constituents ranged from 2.4 x 10-7 (dibenz-[a,h]-anthracene) to 2.5 x 10-2 (1,3-butadiene). Summation of cancer risk values calculated from all individual constituents suggested that several compound classes weighed more heavily in the overall calculated cancer risk for the cigarettes in the study. Compared to previously reported studies, this modeling approach more closely aligns with the observed smoking-related cancer incidence in the U.S. population, with the outcome dependent on selection of risk parameter assumptions such as smoking regimen, exposure scenarios, and availability and selection of cancer potency data. The model does not account for variability in cigarette yield, internal dosimetry, or interaction between chemical constituents.

M5.86 The Risk Information Exchange (RiskIE): A database to communicate in-progress risk assessment work. Wullenweber A*, Kroner O, Maier A, Wexler P, Rak A, Tomljanovic C; Toxicology Excellence for Risk Assessment (TERA), National Library of Medicine (NLM), Noblis, and Concurrent Technologies Corporation (CTC) wullenweber@tera.org

Abstract: The rate of chemical synthesis and use has outpaced both the development of risk values and the resolution of risk assessment methodology questions - spurring diverse agencies to develop risk assessments to meet their immediate needs. Organizations have their own risk assessment schedule, chemical priority list, methodology, and means of communication. Thus, we investigated the hypothesis that although there is a large investment in resources for risk value development in general, there is limited coverage of, and accessibility to, full risk assessments for new chemicals. Furthermore, we assessed whether inadequate communication across organizations hinders data sharing and collaboration, resulting in overlapping efforts. We compiled a list of the in-progress work of multiple organizations and identified chemicals for which assessments are currently underway or recently completed for two or more groups. We also conducted an informal survey of risk assessors to evaluate the mechanisms used to identify ongoing risk value development activities. Identifying such activities posed considerable challenges, suggesting an opportunity to develop systems for enhanced information sharing about in-progress work. To help address increased demand, insufficient resources, and inadequate communication among organizations, a risk assessment notification system was created - the Risk Information Exchange (RiskIE). RiskIE is available at http://allianceforrisk.org and will ultimately be part of the National Library of Medicine’s TOXNET system (http://toxnet.nlm.nih.gov/). RiskIE will contain notifications about in-progress chemical risk value and methodology documents. The system will also include links to information that is not chemical specific such as training modules, white papers and risk methods documents. RiskIE will provide a risk information portal to enhance data sharing and collaboration among groups developing and using human health risk assessment information.

M5.87 Quantification of the Effects of Host Age on the Dose Response of Variola major in Suckling Mice. Weir M.H.*, Haas C.N.; Drexel University mw88@drexel.edu

Abstract: The disease smallpox (Variola major being the causative agent) prior to eradication ravaged the world for centuries irregardless of class or region killing and disfiguring millions. The potential use as a bioterrorist agent spurned the researchers at the Center for Advanced Microbial Risk Assessment (CAMRA) to quantify the dose response relationship. Intraperitoneal exposure dose response data from open literature for V. major the was summarized and fit to dose response models. The dose response data showing good fit to the models were seen to have age delineation in the source document for the Yamada strain of V. major injected intraperitoneally into suckling mice. This age delineation was then used to modify the dose response models to include the dependency of host age into the optimized parameters of the dose response models used. The exponential and beta Poison models showed an improvement in the goodness of fit on the order of 7.8% and 72% respectively, compared to no age dependency included in the models.

M5.88 Comparison of forward and reverse dosimetry estimates using urinary biomarkers of exposure from the CTEPP study. Okino MS*, Morgan MK, Gerlach R, Tsang A, Tornero-Velez R, Power FW, Chang DT, Dary CC; US EPA okino.miles@epa.gov

Abstract: Biomonitoring for human health risk assessment is attractive because biomarkers are an indicator of the chemical dose that entered the body and reflect aggregate exposures received by all routes. However, for non-persistent chemicals (i.e., organophosphates and pyrethroids) and with intermittent exposures, quantitative relationships between biomarkers of exposure and internal dose are often unclear. The Children’s Total Exposure to Persistent Pesticides and Other Persistent Organic Pollutants (CTEPP) study offers a unique opportunity to compare forward and reverse dose estimates from urinary biomarkers of exposure. In this study, environmental concentrations were collected of parent compounds and metabolites in multiple media, including soil, air, dust and food. In conjunction with the environmental measurements, urinary biomarkers were collected, with a subset of participants providing multiple samples over a day. ‘Forward’ dose estimates based on the environmental concentrations and steady-state exposure calculations revealed important routes of exposure, but were not able to account for all the measured urinary biomarker concentrations. ‘Reverse’ dose estimates were calculated using pharmacokinetic models to investigate hypotheses about the relative timing of exposure events and incorporate the longitudinal measurements. Uncertainties in the estimates illustrate important data and information required for interpretation of urinary biomarkers of exposure, and the results provide another way to evaluate the possibilities for the apparent discrepancy between the exposure and biomarker levels. Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy.

M5.89 Computational modeling: reaching for understanding. Hattis D*; Clark University dhattis@aol.com

Abstract: Biological sciences are in the midst of a series of transformations in problem definitions, research tools, and in the economic importance of its applications. Once biologists sought primarily to connect structures with functions, and qualitatively describe causal pathways. In rare cases where computations were required it was often sufficient to whip out a slide rule. While the causal pathway questions are still of fundamental importance, practical applications require us to increasingly address more engineering-like issues of “How much?” “How fast?” and “According to what quantitative functional relationships?” Computational modeling tools offer qualitatively oriented biologists a way to elucidate possible quantitative implications of our improved qualitative understanding of biological systems. In the future we face tremendous challenges and opportunities to elucidate the practical operation of interrelated homeostatic feedbacks at different levels of biological organization from basic enzymatic biochemical pathways through subcellular and intercellular signaling, maintenance of tissue and organ function and inter-system neural and hormonal regulation. Issues for the next decades include how the set points for the numerous monitored parameters are set at all these organizational levels during development, and re-set according to changing needs as the organism encounters different environmental circumstances. How do we understand and analyze both the temporal variability of these systems and implications of interindividual differences in people in different life stages and at various levels of deterioration of homeostatic controls to different perturbations? Informed by future insights in these areas the implications of different modes of actions of toxicants can be explored, and we can hope that by the 100th anniversary of the 1954 Lehman and Fitzhugh paper, uncertainty factors will seem as quaint as the slide rule now seems as a computing device.

M5.90 Dose-time-response modelling for cumulative multinomial data. Chen D*; Department of Mathematics and Statistics, South Dakota State University din.chen@sdstate.edu

Abstract: In this presentation, a dose-time-response modelling approach will be outlined for cumulative multinomial count data. This data type can be resulted from some quantal bioassay in toxicological, pharmaceutical and biological control experiments where a response, such as mortality, in a group of animals is recorded over time points under different dose levels in the course of the experiment. The application of the typical logit and probit analyses is then not valid any more in this situation since it neglects the dependency on time and also the possible interaction of time and dose concentration on the response in the experiment. In this presentation, a dose-time-response model is proposed for this type of experiment and a cumulative multinomial generalized linear model that incorporates time and the other experimental conditions as covariates is developed by the theory of maximum likelihood estimation. Both the point estimator and confidence bands for $ED_{50}(t)$, the concentration of a toxicant that will kill 50% of the animals by a specific time, $t$; as well as $LT_{50}(d)$, the time to 50% mortalities for a specific concentration, $d$, is then formulated in closed form from the newly proposed dose-time-response model. Finally, the newly proposed model is considered for a real data set to demonstrate the application.

M5.91 EFFECT OF AGE AND HISTORICAL INTAKE ON BLOOD DIOXIN CONCENTRATIONS: PHARMAKOKINETIC MODELING TO SUPPORT STATISTICAL ANALYSES. Jolliet O*, Wenger Y, Milbrath MO, Chang C-W, Chen Q, Garabrant D, Towey T, Adriaens P, Gillespie B; University of Michigan, School of Public Health ojolliet@umich.edu

Abstract: We develop a method to convert historical intake of dioxin-like compounds of a given individual into a 2005 actualized intake value. Three successive correction factors need to be applied. 1) Present food intakes are corrected to account for the change of intake with age, as a child eats less than an adult. 2) The second factor accounts for historical changes in dioxin concentration in the food chain, with a peak around 1968. 3) The half-life of each congener in the body is used to decay past intake over time. This approach is used as a pre-treatment of food intake variables - expressed in a 2005 actualized number of meals over lifetime - before performing the statistical regression analysis. It also provides insights on an adequate statistical model for predicting blood as a function of food intake and age. The method is applied to the University of Michigan Dioxin Exposure Study in Midland and Saginaw, involving more than 950 participants whose serum Dioxin concentration was sampled. Results of various linear and non linear model for the statistical analysis are systematically compared.

M5.92 The University of Michigan dioxin exposure study project overview. Garabrant D*, Franzblau A, Adriaens P, Gillespie B, Demond A, Lepkowski J, Olsen K, Hedgeman E, Knutson K, Zwica L; University of Michigan dhg@umich.edu

Abstract: The University of Michigan Dioxin Exposure Study (UMDES) is designed to assess exposures to dioxins, furans and coplanar polychlorinated biphenyls (PCBs), describe the pattern of serum dioxin, furan and PCB (collectively ‘dioxins’) levels among adults and to understand factors that explain variation in serum dioxin levels. The study was undertaken in response to concerns among the population of Midland and Saginaw Counties that dioxin-like compounds from the Dow Chemical Company facilities in Midland have contaminated areas of the City of Midland and sediments in the Tittabawassee River flood plain. There is concern that residents’ body burdens of dioxins may be elevated because of environmental contamination. To address these concerns the study was designed to measure the serum levels in a multistage random sample of the population in the region and to estimate each individual’s past exposure to various factors that are believed to contribute to the body burden of dioxin-like compounds. By measuring factors that reflect potential exposure to dioxins, furans and co-planar PCBs through air, water, soil, food intake, occupations, and various recreational activities, the study identified factors that correlate with (and explain variation in) serum congener levels. 946 people, sampled from five geographically-defined populations by using a two-stage area probability household sample design, participated in an interview, gave blood samples, soil samples and household dust samples for analysis of the WHO 29 dioxin-like compounds. The study subjects live in Midland County, Saginaw County, and part of Bay County, both in and out of the Tittabawassee River flood plain; and in a region of Michigan (Jackson and Calhoun Counties) that has no known industrial point-sources of dioxins like Dow. An additional central goal of the study is to communicate the results and the implications of the results in an effective manner to the populations impacted by the study.

M5.93 Relationship between serum dioxin levels and residential proximity to contaminated soils: Serum dioxin concentrations from the University of Michigan Dioxin Exposure Study. Hedgeman E*, Hong B, Chen Q, Olson K, LaDronka K, Ward B, Gillespie B, Lepkowski J, Franzblau A, Garabrant D; University of Michigan hedgeman@umich.edu

Abstract: As part of the University of Michigan Dioxin Exposure Study, serum samples were analyzed for 29 dioxin-like compounds from 946 adults living in Michigan. Populations for sampling were stratified based on residential proximity to the Dow Chemical facility in Midland, Michigan and the downstream Tittabawassee River flood plain – both known areas of contamination with dioxins; a reference population was selected from two counties 100 miles southwest of the Dow Chemical facility – these counties had only background exposures to dioxin-like compounds. Adults 18 years of age and older who had lived in their residence for five or more years were selected for participation using a two-stage, area probability household sample design. The final population weighted sample was 56% female and 94% white with a mean age of 52 years and a mean BMI of 29 kg/m2. Lifetime exposure to the 29 dioxin-like compounds was estimated with a serum total toxic equivalency measure (TEQ); the TEQ is a single number representing the weighted sum of the individual dioxin-like compounds. Residents living immediately adjacent to the Tittabawassee River flood plain had a mean population weighted serum TEQ of 34.5 pg/g serum lipids (parts per trillion, ppt); residents living downwind of the Dow Chemical plant had a mean serum TEQ of 28.3 ppt; residents living in the reference community had a mean serum TEQ of 29.7 ppt. After adjustment for age, age squared and BMI, the mean serum TEQs for residents living in these three regions were not statistically different (p > 0.05). These results would indicate that living on contaminated soils is not directly associated with serum TEQ. Additional serum dioxin concentration data will be provided.

M5.94 SENSITIVITY ANALYSIS OF SOIL DIOXIN CLUSTER PROFILES TO REGIONAL SPECIFICATION FOR A COMMUNITY IN MICHIGAN, USA. Adriaens P, Towey T*, Demond A, Gillespie B, Garabrant DH, Franzblau A; University of Michigan adriaens@umich.edu

Abstract: Sources of exposure to dioxins are reflected in blood serum and often exhibit a regional character depending on the pathways deemed relevant to its contribution. With soil as a potential direct or indirect pathway of exposure, the sensitivity of hierarchical cluster analysis (HCA) of PCDD/PCDF concentrations in soil samples was investigated through comparisons with an initial understanding of PCDD/PCDF sources. Specifically, the sensitivity of HCA results to linkage method and cluster quantity was evaluated. For the soil samples collected as part of the University of Michigan Dioxin Exposure Study (UMDES), complete linkage best reflected the initial understanding of the data. Increasing the number of clusters offered additional resolution of potentially contributing PCDD/PCDF congener patterns. This analysis helped constrain the areas where variable sources of exposure are relevant, and may play a role in better defining exposure models for the UMDES population.

M5.95 Linear regression modeling to predict household dust TEQ and TCDD concentration. Knutson K*, Zwica L, Lee S-Y, Hong B, Towey T, Chen Q, Lepkowski J, Gillespie BW, Franzblau A, Garabrant D; University of Michigan kknutson@umich.edu

Abstract: The University of Michigan Dioxin Exposure Study (UMDES) was undertaken in response to concerns of environmental contamination in the Tittabawassee River floodplain in Midland and Saginaw Counties in Michigan and in areas within the City of Midland, Michigan. Household dust was determined to be one of many potential explanatory factors for the variation in serum levels of polychlorinated dibenzodioxins (PCDDs), polychlorinated dibenzofurans (PCDFs) and polychlorinated biphenyls (PCBs) and household dust samples were obtained from eligible participants in the UMDES. Linear regression models were used to determine which factors explained the variations in household dust toxic equivalency (TEQ, calculated based on the 2005 World Health Organization TEFs) and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) concentrations measured from the participants in the UMDES and to quantify how much of the variation each factor explained. Potential dust predictor variables were derived from the UMDES questionnaire, dust sampling field data notes and results from the household dust and soil sampling laboratory analyses. Household dust samples were obtained from 764 study participants over the course of three sample seasons. For the TEQ regression model, a higher soil dioxin TEQ from around the house provided the greatest explanation of the variation in the model and was associated with a higher household dust TEQ (Est., 0.336; Δ R2, 10.754). For the TCDD concentration regression model, a higher soil dioxin TCDD concentration from around the house provided the greatest explanation of the variation in the model and was associated with a higher TCDD concentration in household dust (Est., 0.261; Δ R2, 10.078).

M5.96 Predictors of serum 2378-TCDD and 23478-PentaCDF concentration in a background population in Michigan, USA and in a representative USA sample. Gillespie BW, Chang C-W, Hedgeman E, Hong B, Chen Q*, Jolliet O, Knutson K, Demond A, Franzblau A, Garabrant D; UNIVERSITY OF MICHIGAN bgillesp@umich.edu

Abstract: U.S. background levels of 2,3,7,8-TCDD and 2,3,4,7,8-PentaCDF were estimated for 251 subjects in a 2-county area of Michigan (the University of Michigan Dioxin Exposure Study, or UMDES), and for 1145 subjects in a U.S. population-based sample (National Health and Nutrition Examination Survey, or NHANES). Mean levels by age and gender for each sample are graphically presented. The additional effects of weight loss, smoking, breastfeeding, and other variables were also investigated using regression models. In general, the vast majority of background serum concentration values were less than 10 ppt for TCDD, and less than 20 ppt for PentaCDF. Higher serum concentrations of both TCDD and PentaCDF tended to occur with older age, female gender, recent weight loss, and consumption of some foods. Smoking and breastfeeding tended to reduce serum TCDD concentrations. These two factors, as well as weight gain, tended to reduce serum PentaCDF. Hispanics had significantly lower serum concentrations of PentaCDF than other ethnic groups. Serum concentrations in UMDES and NHANES were similar. The UMDES had much lower limits of detection, allowing for more precision in modeling, but the NHANES data had more racial/ethnic diversity, and allowed for generalization of results to the U.S. population.

M5.97 A follow-up investigation of homes with ‘high’ concentrations of PCDDs, PCDFs and dioxin-like PCBs in house dust. Franzblau A*, Zwica L, Knutson K, Chen Q, Lee S-Y, Hong B, Adriaens P, Demond A, Garabrant DH, Gillespie BW; UNIVERSITY OF MICHIGAN afranz@umich.edu

Abstract: As part of the University of Michigan Dioxin Exposure Study (UMDES), the 29 congeners of polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs) and dioxin-like polychlorinated biphenyls (PCBs) that have consensus toxic equivalency factors (TEFs) were measured in house dust of 764 subjects who were a representative sample of the general population in five Michigan counties. Thirteen subjects with the highest toxic equivalency (TEQ – based on 2005 TEFs) in house dust were re-interviewed to determine why these results were elevated. The TEQ’s in house dust ranged from 150 parts per trillion (ppt) to 1,750 ppt. The congeners contributing most to the TEQ among outliers varied, with some cases dominated by PCDDs (n=6), PCDFs (n=1) and PCBs (n=6), respectively. Contamination of soil adjacent to the home appeared to be a source for contamination of house dust in only 2 out of 13 cases. Some homes with outlier dust results were recently constructed (i.e., less than 10 years old). In one case the likely source of contamination of house dust was an old carpet and/or carpet pad. Cigarette ash has a low TEQ and appears to not be an important contributor to the dioxin contamination of house dust. In two cases residents had occupations that may have afforded opportunity for exposures to dioxin-like chemicals and secondary contamination of house dust, however, the congener profiles in house dust in these 2 cases did not appear to differ conspicuously from other cases in which there was no suspicious occupational history. A remote structural fire did not appear to be contributing factor to dioxins in house dust. In most cases the dominant source(s) of house dust contamination in homes with ‘outlier’ values were not related to adjacent soil; in most cases the source(s) of contamination are unknown, but may be related to uncharacterized sources within homes.

M5.98 Risk Minimization in Cultural Developments. Borodyanska IN*, Shamo TI; Musikhochschule Nürnberg, Rodnik Foundation borodyanska@freenet.de

Abstract: Economic globalization and fragmentation of policy as a result of ethnico-religious conflicts all over the world has led not to " cultural association ", and more likely to technical unification, uniformity without the present unity. The culture of today is no more culture of any district, but the culture of time. The main task of the culture consists in putting in pawn a basis for originality of the person, to reduce communicative risk, to influence social gradation, kinds and ways of manufacture and consumption at a wide social and economic level. On the other hand, annually millions of people voluntary or because of the political reasons settle in various parts of globe that leads to cultural enrichment, to cultural internationalism, to occurrence of a new cultural substance. We regard as of paramount importance not assimilation, and process of integration which most finally is expressing in preservation of the native language and together with it in a wide spectrum of spiritual ideas, cultural development, especially, painting and the music, without having language and conceptual barriers. In modern Germany, with the population more, than 82 million people, the Russian-speaking population is 5 millions. These people are separated by national., religious, to attributes and historical experience, and only the intercultural space unites them, considerably reducing risk of loss of the originality and promoting adaptation to new conditions of life. This process is traced on examples of occurrence in Germany, and across all Europe Russian-speaking cultural societies, to expansion of dialogue between cultures, that finally, promotes social and political stability in the world.

M5.99 Risk manageability on decabromodiphenyl ether and alternatives. Tokai A*, Yamaguchi H, Tsunemi K, Iwata M; National Institute for Advanced Industrial Science and Technology a-toukai@aist.go.jp

Abstract: We have tackled three-year project on exploring substitution principle with the example of flame retardant in Japan. We have done full- scale risk assessment for decabromodiphenyl ether and their substitutes. Part of it was already presented at this meeting. In this presentation, we characterized the risk profile of decabromodiphenyl ether and discuss their risk manageability. We mainly looked at human health risk and then discussed the risk manageability of decabromodiphenyl ether in connection with reducing the consumption quantity, substitution and waste management. Through this work, further implications of this analysis was also discussed.

M5.100 Approaches: Mechanism-based categories of processes that cause adverse health effects—implications for quantitative risk assessments. Hattis D*; Clark University dhattis@aol.com

Abstract: This discussion will explore the risk assessment implications of a basic categorization of the processes that lead to adverse effects. A fundamental division relates to the reversibility of the underlying causal events that occur at low chemical doses and/or at preclinical stages of a pathological process. Given a period of no further exposure to a toxicant, is it expected that the system returns completely to a basal state, or has something changed that does not ordinarily revert to baseline. If the answer to this first question is “yes” then the process belongs in a broad category of traditional toxicological effects. If the answer is “no” then it matters how many irreversible steps are needed to produce a noticeably bad outcome. If only a few irreversible steps are needed, the process is in the “molecular biological” category, if many thousands of irreversible steps are needed (e.g. deaths of neurons in a particular area of the brain, destruction of alveolar septa in the lung leading to emphysema) then it needs to be analyzed as a “chronic cumulative” process. The talk will explore various implications of these categorizations for approaches to quantitative risk assessment.

M5.101 Risk Management Process for Developing F-35 Global Sustainment Solution. Wicker, Ph.D. D B*; Lockheed Martin Aeronautics Company dorothy.b.wicker@lmco.com

Abstract: The F-35 Lightning II is redefining the way we do business. F-35 is a transformational weapon system using innovative, integrated management concepts based on affordability paradigms with international partnerships. The four program pillars are affordability, lethality, survivability and supportability. The F-35 Autonomic Logistics (AL) concept incorporates smart, reliable aircraft; agile support; integrated electronic training of pilots and maintainers; ALIS (AL Information System); integrated infrastructure and partnering with government and industry for best value. The affordability challenge is key to the AL concept. The Risk Management Process being implemented on F-35 is key to achieving affordability in the long term through PBL (performance base logistics). The risk process is designed to be reactive and proactive while integrating many system engineering tools. The IMF (Integrated Management Framework) provides a common database for viewing and managing risk. The F-35 program is subdivided into organizational layers. Risk is managed at all program tiers. AL and GS (Global Sustainment) maintain risk lists at all relevant tiers. Tiered risk mitigation plans ensure efforts are focused at the right level. Risk levels (scores) are a combination of probability of occurrence and consequence of occurrence. Quantification of risks is based on lessons learned using risk scoring templates. Proactive risk management is achieved through the use of KSDIs (Key System Developments & Integrations). AL has successfully completed one KSDI in Support Systems and is currently tracking KSDIs in ALIS, Training Systems and GS. The risk management process is very dynamic with risks being added to and removed from the risk lists as potential risks are identified and successfully avoided by mitigation. This paper describes how the risk management process is implemented for AL (Training System, Support System, ALIS) and GS on the F-35.

M5.102 Upstream and Integrated Oversight Assessment for Nanotechnology: A bidirectional approach to anticipatory risk analysis. Kuzma J*, Paradise J; University of MN kuzma007@umn.edu

Abstract: We have developed two novel approaches to help society and policy makers anticipate and prepare for challenges posed by convergent applications of nanotechnology. The first, Upstream Oversight Assessment (UOA), has potential as a key element of anticipatory governance and is a preparatory tool. UOA anticipates the features of emerging applications; their endpoints of use; the extent to which users, workers, ecosystems, or consumers may be exposed; the nature of the material and its safety; whether and where the technologies might fit into current regulatory systems; the strengths and weaknesses of these systems in light of these novel applications; and the possible social concerns related to oversight for them. The goal is to identify data, risk assessment, regulatory policy, and engagement needs for overseeing these products so they can be addressed prior to market entry. Combined with this approach, we have developed methods for integrated oversight assessment (IOA). IOA blends risk analysis, social science, public administration, legal, public policy, and ethical perspectives for evaluating oversight systems. Multiple methods, including historical analysis, expert elicitation, and behavioral consensus, were employed to develop a broad set of criteria for evaluation. Criteria were placed in four categories reflecting the development, attributes, evolution, and outcomes of oversight systems. These criteria illuminate how oversight systems develop, operate, change and affect society and are being applied to historical case studies of oversight for technologies such as gene therapy, genetically engineered organisms, human drugs, medical devices, and industrial chemicals. From the criteria and their comparisons within and between historical oversight systems, evidence-supported lessons for nanotechnology oversight will be proposed. More broadly, we suggest both UOA and IOA as tools for anticipating oversight needs and analyzing outcomes and tradeoffs of oversight systems for emerging technologies.

M5.103 Dermal risk assessment on cosmetics sterilization preservatives of the restricted ingredients. Hwang JH*, Koo TH, Ryeom TK, Woo JM, Kim YJ, Son SJ, Cho SJ, Jang DD, Lee HM; National Institute of Toxicological Research rain@kfda.go.kr

Abstract: Strategy of risk assessment can be used to decide the harmful effect by the exposure to human bodies and to establish the standard for the safe use of chemicals. In this research, we tried to show the standard and management methods about cosmetics sterilization preservatives which have been specified regulation limits according to the regulation about the assignment and standard of cosmetics resources and the test methods through scientific assessment such as confirming the continuous usage or possible enhancement of the present restricted ingredients by performing risk assessment considering the actual usage for cosmetics and cosmetics sterilization preservatives of manufacturers. 7 ingredients (benzyl alcohol, benzoic acid, sorbic acid, captan, formic acid, phenol, formaldehyde) which have been indicated toxicities for risk assessment were chosen as objects for the skin exposure risk assessment out of 61 restricted ingredients. We made use of the estimated skin permeability coefficient (Kp-Pre ADME) and the dermal RfD toxic value etc. based on the daily human skin exposure dosage needed for the skin exposure risk assessment. And we established the exposure scenario based on the actual usage survey for major customers of cosmetics (20-65 years old women) and made use of face surface area measuring 100 women (20-65 years old) actually. To estimate human exposure dosage of cosmetics sterilization preservatives based on customers\\\\\\\' exposure scenario we executed the investigation of the use materials dosage intended for 20 cosmetics industries and decided whether the present exposure is safe according the investigation result. Indicating the management methods for evaluating sterilization preservatives was based on toxicity information like the skin stimulating grade, exposure frequency, body accumulating level and we established the guideline for the safety margin to keep and indicated the plan for improvement and enhanced management methods for restricted ingredients.

M5.104 When “bads” become tradable goods: The benefit-risk analysis of e-waste trading from U.S. to developing countries. Liao C, Chu H*; California State University at Chico chliao@csuchico.edu

Abstract: Electronic waste, or e-waste, has been a fast growing solid waste in the past decade worldwide. Studies found that e-waste is growing in almost three times faster than the municipal waste stream. E-waste contains various substances such as lead, mercury, chromium, and PVC. When these substances are landfilled or incinerated, contaminants are generated and may cause various health effects, such as damages in blood, reproduction, and nerve systems when people are exposed to them. In U.S., one of the popular disposing strategies is to export the e-waste to developing countries in Asia and Africa, such as China and Nigeria, despite the absence of local environmental regulations and the fact that most of these countries do not have adequate disposal technology to minimize the adverse health effects caused by these waste. The goal of this study is to consider both the economic benefit generated by e-waste trading as well as the risks caused by the e-waste. An international trade model is built to describe current international trade market of e-waste from U.S. to developing countries. We use the abatement expenses in U.S. and developing countries to illustrate the incentive to trade. The model allows us to quantify the monetary gains (i.e. the benefits) from exporting e-waste on both U.S. and exporting destinations. The risks (derived from the increase of adverse health effects) are externalities to local waste disposal companies because the local regulations are not enforced. After building the model of current scenario, we then are able to predict what the trade market will become if local environmental regulations are enforced. Unlike traditional benefit-risk-cost analysis, the risks of adverse health effects are transformed to cost; then are compared against the benefits generated from e-waste trading.

M5.105 Evaluation of a Mental Models risk communication instrument for carbon monoxide intoxication. Gurian PL*, Galada HC, Velázquez –Angulo G, Corella-Barud V, Pérez FG, Flores S, Montoya T; Drexel University pgurian@drexel.edu

Abstract: Low income residents of northern México rely on unvented heaters during the winter, a practice which puts them at elevated risk for carbon monoxide intoxication. The Mental Models framework is an empirically-based approach to risk communication which has been found to be effective at identifying gaps in the public’s understanding of technological risk. However, there are still only a limited number of evaluations of risk communication instruments developed using this methodology, particularly for lower socioeconomic status populations. In this study we present results of a pretest/posttest evaluation of the risk communication instrument developed using the Mental Models procedure. Significant gains are reported in addressing the targeted knowledge gaps. Expressed intentions to adopt risk mitigating behaviors also increase after reading the risk communication instrument. A coupon-based incentive program to encourage the adoption of carbon monoxide alarms has been implemented and will provide feedback on the performance the risk communication instrument in producing the desired behavioral change in the target population. This study supports the use of the Mental Models approach for use with lower socioeconomic status groups as it demonstrates that these groups have causal models of technological risks, and deficiencies in these models can be effectively addressed by targeted risk communication efforts.

M5.106 Predicting Severe Winter Coastal Storm Damage. Hondula DM*, Dolan R; University of Virginia dmh4j@virginia.edu

Abstract: Atlantic coast storms called “northeasters,” are responsible for major landscape change along the North Carolina barrier islands. Over the past 40 years, residents of this region have experienced the destructive forces of several “named” storms that caused large-scale redistribution of barrier island sand and loss of coastal structures and infrastructure. While most of the population living on the islands is familiar with the wintertime storms, the damage and scars of the super northeasters, such as the Ash Wednesday storm of 1962, have slipped away from the public’s memories. In our research, we compare the damage zones of the 1962 storm, as depicted on aerial photos taken soon after the storm, with a set of photos taken in the same area in 2002. With these two sets of high-resolution aerial photos we are able to estimate the extent of new development that has taken place since the Ash Wednesday storm. We have defined three zones across the islands on the 1962 aerial photos: (1) the zone of destruction on the seaward edge of the islands where the storm waves occurred, (2) the zone of structural damage, and (3) the zone of flood damage at the landward edge of the storm surge. Factors we are considering include: the rate of coastal erosion, the rate of development, changing value of the dollar since 1962, improvement in construction, increase in property values, and the reduced natural protection due to development. In comparing the values of six factors with the 1962 damage data yields the predicted dollar value for storm damage if (or when) another storm of the magnitude of the Ash Wednesday storm should occur. This model also provides an opportunity to estimate the rate of increase in the potential losses through time as shoreline erosion continues to progressively reduce the buffer between the present day development and the edge of the sea.

M5.107 Risk-Based Prioritization for Nuclear Material Disposition – An Approach Combining Decision and Statistical Science Techniques. Hoffman J, Kelly E*, Koehler A; Los Alamos National Laboratory ekelly@lanl.gov

Abstract: Stabilization and disposition of thousands of containers of plutonium-bearing materials are underway at several Department of Energy facilities. Following recommendations from the Defense Nuclear Facilities Safety Board the decision was made to prioritize these activities based on risk reduction. The goal was to ensure that the containers with the highest risks were scheduled for processing first. The challenge was to develop a method for assigning risk to individual containers.At one of the facilities, Los Alamos National Laboratory (LANL), the initial approach for prioritization used a theoretical risk value, which was an estimated worker dose multiplied by a container failure probability index. This approach was adjusted as warranted using expert judgment. However, facility management desired a more systematic method for incorporating expert judgment, data from surveillance activities and information from actual container failures into their prioritization decisions. This paper describes how Multi-Attribute Utility Theory (MAUT) and CART® (Classification And Regression Trees) were combined with expert judgment and surveillance data to develop a risk-based prioritization approach. Using MAUT, nuclear material experts developed an algorithm for converting complex container surveillance data into a risk ranking. The ranked containers, along with data available for all containers, provided a “learn” data set for CART®. The CART® model was then used to identify high priority items in the entire population of containers. The CART® based prioritization resulted in substantially increased odds of finding high priority items (38% versus less than 8%) and identified items that would have been missed by the theoretical risk method.

M5.108 Exposure factors: How to characterize the data? Öberg T*, Filipsson M, Bergbäck B; University of Kalmar, 39182 Kalmar, Sweden tomas.oberg@hik.se

Abstract: The exposure to chemicals is, apart from level of contamination and release, controlled by human behavior, physiological characteristics and environmental factors. The quantifiable statistics used to describe this information are called exposure factors, and are instrumental for any risk assessment. Evaluated data on exposure factors are often accessed from the literature or compilations such as the Exposure Factors Handbook from the U.S. EPA. However, these factors may differ between countries and regions, and we have recently collected and evaluated a similar range of country-specific data from Sweden. Our compilation includes the population age distribution, moving patterns, body weights, skin surface areas, activity patterns, dietary intakes and home produced items, drinking water consumption and private wells, residential building characteristics, and soil parameters. One problem that we faced in this study was the choice on how to best describe and present the available data. Some examples from previous compilations are arithmetic means, standard deviations, percentiles, and specific parametric distributions. We have in earlier work highlighted the uncertainty in choosing distributions, and the overconfidence that is conveyed in assigning single numbers or specific distributions. In the present study, we have instead chosen to estimate and present both parameters (mean, standard deviation, kurtosis, and skewness) and multiple percentiles (1-99%) with uncertainty ranges (95% confidence intervals). These uncertainty estimations were performed by resampling (bootstrapping) from the primary data. This tabulated compilation can serve as a basis for estimating exposure with several different approaches; deterministic, intervals, Monte Carlo-simulations, or probability bounds analysis. In this poster we will exemplify and discuss how the compilation and presentation of data on exposure factors affect the methodological choices and outcome of an exposure assessment.

M5.109 Perceptions and beliefs about weed management: Mental models of Ohio grain and produce farmers. Wilson R*, Tucker M, Hooker N, LeJeune J, Doohan D; The Ohio State University and Purdue University wilson.1376@osu.edu

Abstract: Experts have long sought to understand the weak adoption by farmers of integrated weed management approaches and more recently, herbicide resistance mitigation strategies. Attempts to increase adoption of these strategies have relied largely on a methodology that is less than adequate for these contexts. In order to increase the effectiveness of engagement efforts there is a need for a more circular process of communication about the risks, which incorporates the perceptions, values, and existing beliefs of the target audience. A better understanding of the underlying drivers for decision making in these contexts will allow future efforts to more effectively communicate the need for integrated and preventive tactics when managing the risks associated with weeds (i.e., invasives, herbicide use and resistance, etc.). The objective of the research reported here is to establish a baseline understanding of the key factors that drive farmer decision making. Thirty Ohio farmers participated in an in-depth interview where they were asked to reflect on how weeds are introduced and spread, what risks and benefits they pose, and what management strategies they are familiar with and which they prefer. Their responses were mapped, coded and analyzed for dominant beliefs and major decision making influences. The results indicate that farmers largely attribute the introduction and movement of problematic weeds to factors outside their control (e.g., the environment, specific plant characteristics). They do believe in diverse and integrated management, but their focus is on control as opposed to prevention. In general, they tend to receive messages about integrated and preventive approaches, but do not always put them into practice due to underlying beliefs about the inevitability of new weeds being introduced and spread on their farm.

M5.110 Ingestion of home-grown produce: Managing uncertainties in risk assessment. Quiring SJ*, Kabli HA, Scheffler LD, Ruoff WL; URS Corporation sharon_quiring@urscorp.com

Abstract: Baseline human health risk assessments conducted at four radiological waste sites at the Hanford Nuclear Site in Washington State identified ingestion of home-grown produce (fruits and vegetables) at three of the four sites as the most significant exposure pathway for future residents by an order of magnitude. Two main factors driving excess cancer risks estimates are: 1) concentrations in produce tissue (modeled from soil concentrations in these assessments) and 2) produce ingestion rates. Soil concentrations in a hypothetical home garden were estimated by assuming a specific volume of buried waste would be brought to the surface, spread throughout a garden area, and mixed with existing soil. Soil concentrations can, therefore, vary depending on assumptions of volume, waste thickness, and garden size. Similarly, ingestion rates of home-grown produce can also vary depending on assumptions regarding over-all produce intake, percentage of intake that is home-grown, and whether average intake rates or upper-bound intake rates are used. At the four sites evaluated in this study, baseline assessment cancer risks from ingesting produce ranged from 10-5 to 10-1, assuming a 100 m2 garden size and average home-grown produce ingestion rates recommended by U.S. EPA. Using EPA’s 2007 food intake study and other sources of ingestion rates, and decreasing conservative assumptions for ingestion rates and garden size, demonstrated that some sites identified in the baseline assessment as having risks above EPA’s target cancer goal risk range of 10-6 to 10-4 could have risks within or below the acceptable range. The range of credible garden sizes varied less, and thus affected risk results less, than varying ingestion rates. This analysis indicates that a range of risk estimates, including a thorough presentation and discussion of uncertainties for key exposure assumptions, is necessary to ensure that risk managers have sufficient information to make sound risk management decisions.

M5.111 Interdicting smuggled nuclear material. Morton DP, Popova E*, Schneider E; The University of Texas at Austin elmira@mail.utexas.edu

Abstract: This research concerns optimally designing a global network of radiation detectors to thwart a smuggler who attempts to smuggle illicit nuclear material through that network. There is a significant ongoing research and development effort in the US to build higher fidelity radiation detectors. We investigate how to best deploy those detectors. The goal is to design a network of detectors to maximize the probability a smuggler is detected. An interdiction model assumes an intelligent adversary, who, in turn, solves an optimization model to select the smuggling route that best avoids detection. In addition to this model of the smuggler’s behavior, a key input to the decision-support system is the probability a radiation detector will correctly sense smuggled nuclear material. Monte Carlo radiation transport calculations are used to parameterize these detection probabilities as functions of system variables such as the type of detector, its detection algorithm, the type and quantity of smuggled material, the manner in which that material is shielded, etc. This multidisciplinary research effort combines optimization and probability modeling from operations research with radiation transport calculations from nuclear and radiation engineering.

M5.112 Improving EPA chemical risk assessment information availability: Using DSSTox database technology and IRIS data in computational toxicology. Backus GS*, Johnson M, Richard A; US EPA backus.gillian@epa.gov

Abstract: Computational toxicology has provided new tools for mining databases in a relational manner. EPA’s Integrated Risk Information System (IRIS) (www.epa.gov/iris) provides public access to searchable documents that describe health effects of substances via hazard identification and dose-response analysis. IRIS also presents qualitative and quantitative reference values and cancer weight-of-evidence characterization. Revisions to the IRIS Web site, currently underway, will increase usability and transparency for a wider audience via improved search capabilities and greater access to critical and time-sensitive information. Linking IRIS to computational toxicology information can advance information available to risk assessors. The mission of the EPA’s Distributed Structure-Searchable Toxicity (DSSTox) Database project is to improve structure-activity and predictive toxicology capabilities. The DSSTox Web site (www.epa.gov/ncct/dsstox) provides a public forum for querying chemical structure files associated with downloadable toxicity data, including IRIS. The DSSTox/IRIS data file (IRISTR), available on the DSSTox Web site, includes 34 IRIS-specific information fields and documentation, as well as quick links to all of the IRIS chemical substances (544) on the IRIS QuickView page. This file has been integrated into the DSSTox Web site structure-browser that permits users to structure search across one or more DSSTox files, and retrieve content from IRISTR as well as other DSSTox data files. The DSSTox structure files additionally serve to interface IRIS and other DSSTox data files with large public resources, such as PubChem. This EPA collaboration has enhanced the usability and functionality of both databases to help risk assessors and decision makers quickly find the information they need. This abstract does not represent EPA policy.

M5.113 Evaluation of the Benchmark Dose for point of departure determination: An alternative to “Dr. NO(AEL)”. Izadi H, Grundy JE*, Bose R; Health Canada, Government of Canada jean_grundy@hc-sc.gc.ca

Abstract: The New Substances Assessment and Control Bureau (NSACB), Health Canada, is responsible for the human health risk assessment of ‘new’ chemicals and polymers during their pre- and early marketing phases as required under the mandate of the Canadian Environmental Protection Act (CEPA). The point of departure (POD) used to estimate the dose below which toxicity is not expected in humans, are traditional thresholds (NOAEL, LOAEL); known to be sensitive to the study design (e.g., dose intervals and group sizes). The benchmark dose (BMD) method has been proposed as an alternative to this approach, yielding a more accurate POD derived from all the study data. The U.S. EPA Benchmark program (BMDS version 1.4.1) was used to retrospectively compare traditional toxicity thresholds against Benchmark-derived PODs, for 49 chemicals and one polymer notified under the New Substance Notification Regulations. For comparability and consistency, repeated-dose studies were limited to oral gavage studies with NOAEL or LOAEL lower than 500 mg/kg/day. Specific cases where Benchmark provided an obvious advantage in health risk assessment were identified, for instance studies where dose ranges and/or intervals were widely spaced. The BMD generally was higher than the NOAEL and lower than the LOAEL. While NOAEL and LOAEL derivation was affected by the distribution of dose groups, the BMD approach performed consistently, independent of dose distribution. Model fits were not affected by dose distribution or the toxicity end-point selection for the chemical under question; the BMDL was generally constant at 57% of the BMD. Although Benchmark modelling is not always ideal, or practicable, the Benchmark approach was nonetheless shown to provide a mathematical, objective alternative to traditional methods of POD derivation, yielding more precise estimates of the true no-adverse-effect level, an important consideration in situations where the margin of exposure is relatively narrow.

M5.114 Parent’s knowledge and control of asthmatic children’s home environment: A pilot study of mental models. Biksey TM*, Wu F; Biksey - WSP Environmental Strategies LLC, Pittsburgh, Pennsylvania; Wu - Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania thomas.biksey@wspgroup.com

Abstract: Parents’ knowledge and control of asthma triggers in home environments can help to improve asthmatic children’s respiratory health. In this pilot study, we conducted in-depth interviews with twelve parents of asthmatic children to determine how much they knew about asthma triggers in home environments, if and how they attempted to control those triggers, and the information they received and trusted regarding asthma triggers. Of the twelve parents interviewed, all knew about the potential link between environmental tobacco smoke exposure and asthma symptoms. Ten of the parents knew about the link between pet exposure and asthma symptoms, eight knew about the link between house dust mite exposure and asthma symptoms, and four knew about the link between mold exposure and asthma symptoms. While the eight Caucasian parents in the study reported receiving information from their physicians about environmental triggers of asthma, the four black parents reported having received no such information from their physicians. Yet all parents reported having greater trust in information received from physicians than from other sources, such as books or the Internet. While black and Caucasian parents did not have statistically significant differences in knowledge of asthma triggers, Caucasian parents reported doing significantly more actions to control or remove the triggers from their homes. This pilot study represents the first step in designing communications to better fill gaps in parents’ knowledge about mitigating asthma triggers in asthmatic children’s home environments. The study suggests that one possible reason that African-American children have higher asthma rates is that their parents are getting less information from doctors about how to control home environmental asthma triggers.

M5.115 Mapping Emergence: Network Analysis and Construction of Climate Change Risk in the Media. Klinsky S*; University of British Columbia klinskys@interchange.ubc.ca

Abstract: Attempting to understand the role of the media in public construction of complex risks has been of interest to both the mass communication and risk analysis communities. This work builds on insights from both these communities and attempts to use network analysis to explore the development of media discourses about climate change. Widely applied in a variety of disciplines, attention to both the roles of actors and network structure has contributed to our understanding of mechanisms that are neither internal nor external to actors, but depend on the interactions among them. Media portrayals of climate change have been investigated by a number of researchers but most of this work has either used frame analysis to delineate trends in coverage, or qualitative content analysis to explore particular aspects of representation. Complex issues such as climate change present a challenge to both strategies because they consist of multiple story lines and emerge from the interactions of many actors. This poster presents the results from an experimental network analysis of Canadian media coverage of climate change between 1997 and 2006. The technique successfully provided a systematic view of the roles of key actors in several distinct storylines, and created compelling visualizations of the emergent nature of climate discourse. The primary limitations of network analysis in this case resulted from its ambiguous stance on agency and from practical coding decisions, both of which suggest the utility of further exploration.

M5.117 Literature Survey on Risks of Escape from Underground Spaces. Tanaka A*; National Institute of Advanced Industrial Science and Technology a.tanaka@aist.go.jp

Abstract: Based on a survey of the literatures on 22 cases of evacuation from underground spaces, we analyze and discuss the effect of hazardous factors on escape. These cases involve road tunnels, railway tunnels, mines, and construction sites. The types of disaster include fire, automobile collision, electrical power failure due to earthquakes, and terror-based occurrence. Our analysis shows that evacuation success or failure dependson (1) urgency of the event and (2) the distance requiredto reach shelters or exits, regardless of how underground facilities are used. Based on the analysis of these two factors, we concluded that the maximum distance of safe evacuation is within 400 m. We compared urgency and rescuer access time in disasters involving power failures in underground spaces. In the case of power failures, the risk to human life, and the need for immediate escape are small. In some fires, the contamination of escape routes is far faster than rescuerscan arrive, and survey results show the importance of measures and designs for self-escape, especially in underground spaces.

M5.118 SERRA: Use of a Knowledge Base Approach to Facilitate Rapid Risk Assessment of Environmental Agents. Singer L*, Ritondo M, Rogers J, Chappie D, Hines S, Taylor M, Nichols T, Baumel I, Sonich-Mullin C, McKean D; Battelle, Battelle, Battelle, Battelle, Battelle, Battelle, US EPA NHSRC, US EPA NHSRC, US EPA NHSRC, US EPA NHSRC hiness@battelle.org

Abstract: US EPA’s National Homeland Security Research Center is charged with developing tools to aid in the response to terrorism incidents. The assessment of risk is a core component of this effort, as it is critical to address exposures of the public, protect infrastructure, and support overall decontamination efforts. Risk assessment of biological agents is hindered by the limited availability of centralized, comprehensive information. The SERRA (Support for Environmental Rapid Risk Assessment) software tool provides a dynamic infrastructure to organize, maintain, and access available relevant information. SERRA is composed of two sections, a knowledge base and a literature repository. The knowledge base, organized by agent, is broken down into a taxonomic structure of information mined from the open source literature and analyzed by subject matter experts. While the emphasis has been on disease transmission, dose-response, agent fate and transport, risk assessment techniques, and decontamination, other fields have been populated with information on disease characteristics, prophylaxis and treatment, and personal protection. Content is linked back to its full-text source document in the literature repository, where the user may access a document criteria report designed to assist in the assessment of its scientific relevance. SERRA contains data for 4 biological agents with further additions planned, and provides an extensive view of biological agent behavior that can enhance risk-based response plans. Its content can expedite sampling plan development, provide data to derive infectious dose values, describe sensitive subpopulations, and assist in the selection of decontamination strategies. This resource is intended to have widespread application in the development of the microbial risk assessment process.

M5.119 Towards risk-based microbiological criteria. Barrucci F.*, Ricci A.; Istituto Zooprofilattico Sperimentale delle Venezie fbarrucci@izsvenezie.it

Abstract: As the scientific basis for the food safety was strengthened, new food control approaches were developed. Food safety management has evolved from an approach strongly relying on end-product testing to a preventive approach based on process control, which is primarily hazard-based. In the Hazard Analysis Critical Control Point (HACCP) system decisions, standards and actions are based on information on relevant hazards and are aimed at eliminating or reducing exposure to such hazards, with the expectation that there will be a reduction in risk. This evolution resulted in a radical revision of the European Union food safety hygiene rules, from the proposals set out in the White Paper on Food Safety (2000) to the Regulation (CE) 178/2002. Minimum hygiene requirements were established and HACCP principles were incorporated in the new hygiene package legislation. More recently, there is growing need for risk-based food safety management systems: the hazard-based control systems do not directly link the effectiveness of control measures to the expected level of health protection, e.g., a reduction in the number of foodborne illnesses occurring in a country. With the advances in microbiological risk assessment techniques, it is becoming evident that a system where decisions, standards and actions are based on specific knowledge of risk, are aimed at achieving an established level of health protection and are explained and validated in these terms, is possible. In this contest we analyze the evolution of microbiological criteria as the food safety system has evolved: microbiological tests increased their significance and changed their purpose from determining the acceptability of food batches, to validating processes and procedures based on HACCP and GHP principles, to verifying the level of hazard in specific steps of the food chain in comparison to specific targets established in order to guarantee an appropriate level of health protection.

M5.120 Assessing the needs of the working poor during a pandemic. Bier VB*, Zach LS, King SB, O\\\'Sullivan T, Burgos I; University of Wisconsin-Madison bier@engr.wisc.edu

Abstract: An influenza pandemic could cause enormous suffering. One non-pharmaceutical intervention for controlling the spread of disease is social distancing—including closing of schools and non-critical businesses such as entertainment venues. The hospitality industry could also experience dramatic downturns due to loss of demand. The Wisconsin Department of Health and Family Services asked us to look into policy options for addressing the needs of the working poor in a pandemic. Trust for America’s Health has estimated that a severe pandemic would result in a 2% loss of annual gross domestic product (GDP) for the state of Wisconsin—more accurately, an 8% loss concentrated during the peak three months of the pandemic. Our analysis showed that these losses would not affect the various counties of Wisconsin equally. The largest counties, with the greatest employment, would be responsible for the largest fraction of state GDP loss, but would not be particularly hard hit in percentage terms. By contrast, the counties that are hardest hit would be mainly those that are heavily dependent on tourism. Moreover, the average wages of the jobs lost due to a pandemic would be extremely low—less than $26,000 per year in most counties, and less than $14,000 per year in some (perhaps in part due to part-time or seasonal employment). Low-income workers typically do not have sufficient resources to get by for more than a few days or weeks without working. In addition to the risk of job loss due to loss in demand, many low-income jobs also do not offer paid sick leave, and some do not offer sick leave at all. We identified a number of policy options to address these concerns, including facilitating liquidity, compensating individuals who are isolated (due to illness) or quarantined (due to possible exposure), enhancing job security, strengthening social capital in low-income communities, and extending moratoriums on utility cutoffs.

M5.122 Effects of fear of insufficient benefits on risk perception of nuclear power: An experimental research of risk communication. Tsuchida S*, Tsujikawa N, Koike F, Nagaoka Y, Tanigaki T; Kansai University tsuchida@kansai-u.ac.jp

Abstract: Effects of fear of future insufficient benefits on risk perception, risk judgments, and risk information exploring behaviors were examined with a laboratory experiment. Three hundred and twenty-three individuals between 20 and 59 years old who lived in Osaka area, Japan were randomly sampled and participated in the experiment. At the experiment, participants watched a video putting fear of future insufficient energy supply into them (experiment condition) or a video explaining energy supply with no fear (control condition), and then they explored a database of nuclear power which consisted of scientific, technological, economic, social, and institutional information of nuclear power. Participants’ information exploring behaviors including the length of time spent, risk perception of nuclear power, trust in people who made the video and the database, and disposition to trust in other persons were measured and analyzed.

M5.123 Evidence-Based Dose Response Assessment for Thyroid Tumorigenesis from Acrylamide . Dourson ML*, Hertzberg RC, Allen BC, Haber LT, Parker AL, Kroner OL, Maier A, Kohrman MJ; TERA; BRUCE ALLEN CONSULTING haber@tera.org

Abstract: Acrylamide is commonly found in various foods. Cancer studies in rats have reported increases in tumors of the thyroid, mammary tissues, tunica vaginalis of the testis, and sometimes tumors at other sites. We review relevant studies on acrylamide’s DNA toxicity, tumor formation and the manner of its tumor formation. We find, as do others, that glycidamide, a metabolite of acrylamide, causes mutations, but that acrylamide does not. We also find that thyroid tumors are most consistently sensitive in rats, being evoked in each of 4 long-term experiments. We evaluate the common manners of this tumor formation in the thyroid, including both mutagenicity and thyroid growth-stimulation. Consistent with the overall weight of the evidence, we conclude that both of these manners or modes of action may be occurring, with the mutagenic mode of action determining the low-dose response and growth stimulation dominating the response at higher doses. Using U.S. EPA methods, we “decouple” these modes of action and determine that the probit model, unlike the multistage model, best reflects the likely underlying “decoupled” biology. We use the probit model to identify a health-protective, linear cancer slope factor (SF) of 0.030 (mg/kg-day)-1 for the low area of the dose response curve associated with possible mutagenicity. We also identify a Reference Dose (RfD) in the range of 0.05 to 0.02 mg/kg-day for the high area of the dose response curve associated with the growth stimulation. This latter value can be used to determine the upper range of risk.

M5.124 Maximize Utility of Field Screening Results using Regression Analysis. Swanson WL*, Michael DI; Neptune and Co., Inc. wswanson@neptuneinc.org

Abstract: Field screening techniques are often used to evaluate sediment contamination because they produce results quicker and at a lower cost than fixed lab analysis. Because of these qualities, field screening techniques can maximize coverage of contaminant assessment, making these techniques desirable for directing sampling plans, defining boundaries of contaminated areas, and locating hot spots or reference areas. When there is a need for a more quantitative determination of contamination, a subset of the screening samples, covering the range of screening results, are submitted to a fixed lab for analysis. Regression analysis can be performed to evaluate the quantitative relationship, or calibration curve, between the field-screening data and laboratory-based confirmatory analysis data. The purpose of the evaluation is to determine whether an adjustment to the screening data is warranted, and if so, to recommend a defensible adjustment. Adjusted screening data can then be combined with fixed lab data to increase the data support to spatial assessments and site decisions. The regression analysis approach that will be discussed is a stepwise process to determine the best overall model based on maximizing R2, minimizing Lack of Fit error, and identifying and removing outliers. Outlier identification and testing was based on three criterion: 1) absolute standard residuals to identify outlying Y (Lab) values, 2) hat matrix elements to identify X (screening) values with strong leverage in determining the fitted value of Y, and 3) Cook’s distance measures of the overall influence of an individual observation on the regression equation. The process will be illustrated using data from several sediment studies utilizing PCB immunoassays and XRF data coupled with fixed laboratory measurements.

M5.125 Evidence-based assessment of human relevance and mode of action for tunica vaginalis mesotheliomas resulting from oral exposure to acrylamide. Haber LT*, Maier A, Kroner OL; TOXICOLOGY EXCELLENCE FOR RISK ASSESSMENT haber@tera.org

Abstract: The human relevance and mode of action of acrylamide-induced tunica vaginalis mesotheliomas (TVMs), a tumor of the scrotum, was evaluated based on the available data on acrylamide and general biology considerations. TVMs are most common in F344 rats, suggesting an association with the hormonal milieu unique to F344s, and suggesting an association with Leydig cell tumors, which occur in F344 rats at a very high incidence. These hypotheses were biologically plausible, but direct data on acrylamide were lacking for several key events; some of the gaps could be filled in based on other biology information. Overall, the MOA data indicate that some relatively minor fraction of the acrylamide-induced TVM response in rats may be relevant to humans. Multiple MOAs may apply, and some contribution from a mutagenic MOA is plausible, along with a likely influence from LCTs or from the same hormonal changes that result in higher LCT incidence in F344 rats. The data reviewed are not sufficient to distinguish between there being a causal relationship between LCTs and TVMs, and the hypothesis that these tumor types are responding to some other influence (e.g., hormonal mileu of the F344 rat). However, in light of the very low incidence of TVMs in humans, the most appropriate upper bound estimate of the risk of acrylamide-induced TVMs in humans is below de minimis levels.

M5.126 A Comprehensive Evaluation of the Economics of Food Safety in Ohio. Scharff RL*, McDowell J, Medeiros L; Ohio State University scharff.8@osu.edu

Abstract: Past efforts to evaluate the economic burden of risks from foodborne illness have generally taken the form of studies focused on a single or a small number of "important" pathogens. As a result, the economic impact of many less important pathogens has not been sufficiently explored. The result is that currently available studies only provide cost estimates for fewer than 4 million of the 76 million cases of foodborne illness estimated by Mead et. al (1999) are incomplete and, as a result, underestimate the efficacy of broad-based intervention programs. We present a model that enhances the oft-cited USDA Economic Research Service (ERS) model. Our model uses a more comprehensive set of pathogens, includes pathogen-specific hospitalization costs, and includes a quality-adjusted-life-day (QALD) measure to account for lost quality of life. We also use Monte Carlo simulations to evaluate the effects of uncertainty in our analysis. The resulting annual expected economic cost of foodborne illness for each Ohio resident is as high as $400. This is five to ten times greater than would be predicted by the ERS model. To illustrate the usefulness of this comprehensive approach, we combine our results with data from the Ohio Family Nutrition Program (FNP) to evaluate the effectiveness of a food safety education program.

Society For Risk Analysis Annual Meeting 2007

Risk 007: Agents of Analysis

Session Schedule & Abstracts

M5
For Your Eyes Only: Poster Reception

Room: Marriott Riverwalk Ballroom 6:00 - 8:00 PM

Society For Risk Analysis Annual Meeting 2007

Risk 007: Agents of Analysis

Session Schedule & Abstracts

M5For Your Eyes Only: Poster Reception

Room: Marriott Riverwalk Ballroom 6:00 - 8:00 PM

M5
For Your Eyes Only: Poster Reception