SRA Logo (print)


Society For Risk Analysis Annual Meeting 2009

Risk Analysis: The Evolution of a Science

Session Schedule & Abstracts


* Disclaimer: All presentations represent the views of the authors, and not the organizations that support their research. Please apply the standard disclaimer that any opinions, findings, and conclusions or recommendations in abstracts, posters, and presentations at the meeting are those of the authors and do not necessarily reflect the views of any other organization or agency. Meeting attendees and authors should be aware that this disclaimer is intended to apply to all abstracts contained in this document. Authors who wish to emphasize this disclaimer should do so in their presentation or poster. In an effort to make the abstracts as concise as possible and easy for meeting participants to read, the abstracts have been formatted such that they exclude references to papers, affiliations, and/or funding sources. Authors who wish to provide attendees with this information should do so in their presentation or poster.

Common abbreviations


Room:    



.0    A Pareto-based multi-criteria approach to building integrated pest risk maps. Yemshanov D*; Canadian Forest Service   dyemshan@nrcan.gc.ca

Abstract: Invasive species risk maps often provide broad policy guidance on where to allocate resources for pest monitoring and regulation but they also often portray individual risk components as separate products. Building integrated risk maps usually relies on various multi-criteria analysis techniques and requires a prior knowledge of decision makers` preferences and the importance of particular risk components. This information is often poorly understood, or is simply missing, for many new invasive threats. This study proposes an alternative approach - building integrated risk maps using the principle of Pareto dominance. The technique analyzes the elements of the risk map in dimensions of risk components and does not require prior knowledge of the relative importance of individual risk components. We define integrated risk rankings as the subsequent Pareto frontiers in dimensions of individual risk components and estimate them using a Goldberg algorithm. Each frontier outlines the best tradeoff between individual risk components and therefore represents a nominal risk rank. The approach is demonstrated with an example of a forest invasive pest recently detected in North America, Sirex noctilio Fabricius. The results provide a spatial representation of integrated risks and uncertainties and show major geographic hotspots where the consideration of trade-offs between multiple risk components changes integrated risk rankings. Overall, a Pareto-based aggregation may be a better alternative to multi-criteria techniques based on compensatory and outranking principles in cases where knowledge about the importance of individual risk components is scarce or missing. The nominal nature of risk rankings also makes integrated risk maps a more useful tool in prioritizing pest surveillance and regulation efforts.

.0    Pest risk maps, information gaps and the design of early warning programs for invasive species. Yemshanov D*, Koch FH, Ben-Haim Y, Smith WD; Canadian Forest Service   dyemshan@nrcan.gc.ca

Abstract: Integrated risk maps and pest risk assessments provide broad guidance for establishing pest surveillance programs for invasive species but they rarely account for knowledge gaps about the new threat or how these gaps can be reduced. In this study we demonstrate how the notion of information gaps and potential knowledge gains could be used in prioritizing large-scale surveillance activities. We illustrate this approach with the example of an invasive pest recently detected in North America, Sirex noctilio Fabricius. First, we formulate the existing knowledge about the pest into a stochastic model and use the model to estimate the expected utility of surveillance efforts across the landscape. The expected utility accounts for the abundance, economic value and susceptibility of the host resource and the benefits of timely S. noctilio detections. Next, we make use of the info-gap framework to explore two alternative pest surveillance strategies. The first is to aim at timely detection and maximize the robustness to uncertainty about S. noctilio behavior; the second is to maximize the potential knowledge gain about the pest`s via unanticipated finds. In our case, we are uncertain both about the likelihood that this species is present at any given location and the odds of its detection, but we seek a geographical allocation of survey priorities that maximizes the range of uncertainty over which the expected benefits from successful detections will nevertheless be gained. The results include two sets of spatial outputs that can be used to prioritize surveillance efforts. The first set prioritizes certain S. noctilio detections and the second maximizes the potential of a survey to enhance the existing knowledge about the pest via unanticipated finds. These outputs are then aggregated via the Pareto ranking technique in a single priority map that outlines the survey regions with the best trade-offs between both surveillance strategies.

.    Probabilistic Risk Analysis and Terrorism Risk. Ezell BE*; Virginia Modeling, Analysis, and Simulation Center, Old Dominion University   bezell@odu.edu

Abstract: "There is only one logical and workable interpretation of probability and it is that of degrees of belief." – G. Apostolakis (1990) For more than thirty years, probabilistic risk analysis (PRA) has been a major tool for assessing risks and informing risk management decisions by government and businesses, in areas as diverse as industrial safety, environmental protection, and medical decision making. The more recent application of PRA to terrorism risks is new however, and not uncontroversial. Here, we take a broad view of PRA, including any probabilistic approach involving tools like event trees, fault trees, and decision trees. We also introduce other tools such as, social network analysis, and game theoretic approaches, which may prove to be useful in dealing with the intelligent adversary. A major challenge in risk analysis of terrorism is the fact that terrorists, unlike nature or engineered systems, are “intelligent adversaries” and may adapt to our defensive measures. The National Research Council’s Committee on Methodological Improvements to the Department of Homeland Security’s (DHS) Biological Agent Risk Analysis (referred hereafter as the NRC Committee) has argued that because of this adaptive nature, alternative tools like decision trees are needed to assess the risks of terrorist events. While we do not take issue here with the possible value of these alternative approaches, we aim to make a case 1) that PRA is an important and useful approach for quantifying terrorism risks and has value in guiding risk management decisions; 2) event trees can be used as part of a terrorism PRA to decompose the universe of terrorism scenarios; and 3) decision trees, like all approaches, have limitations. In the case of applications for terrorism risk analysis, decision tree limitations may be difficult to surmount in that the adversaries’ objective functions and level of ability to predict tree outcomes are unknown.

.    The emerging role of risk analysis in U.S. Army Corps of Engineers’ decisionmaking . Haimes YY*; University of Virginia   haimes@virginia.edu

Abstract: Following Hurricanes Katrina and Rita, the U.S. Army Corps of Engineers (USACE) revised its Strategic “Campaign Plan” with 12 Actions for Change, to “Effectively Implement a Comprehensive Systems Approach: comprehensively designing, constructing, maintaining, and updating our engineered systems to be more robust, with full participation of all stakeholders.” For example, the first two Actions speak to the very essence of the Society for Risk Analysis: “Point 1 Employ integrated, comprehensive and systems-based approach; Point 2 Employ Risk-Based Concepts in Planning, Design, Construction, Operations and Major Maintenance.” Leaders from the four major research centers of the USACE (Hydrologic Engineering Center; Institute for Water Resources; Engineering Research and Development Center; and the New Orleans District) will present in this Symposium their perspectives on the challenges and opportunities that the Corps faces in the deployment of the 12 Actions for Change, focusing on the need for better risk assessment, management, and communication. (Note: During the last two years Yacov Y. Haimes has been the Arthur Maass and Gilbert White Fellow at the USACE Institute for Water Resources (IWR), focusing on the 12 Actions for Change.)

.    Addressing vulnerabilities of electric power system including interdependencies. Kroeger W*, Eusgeld I, Probst P; ETH, Zurich   kroeger@mavt.ethz.ch

Abstract: The electric power system is regarded as one of the most critical infrastructures (CI) in industrialized countries. In continental Europe it comprises a large-area, sycchronized grid of decentralized control applying common rules (UCTE Handbook) and principles (internal liberalized market). It has witnessed dramatic changes, both technological and organizational, and is subject to an increasing set of threats including economic pressure and malicious attacks.The Swiss Federal office of Civil Protection is committed to develop a strategy to adequately protest CI based on sound scientific input. For this purpose, we have developed and applied an analytical framework comprising three steps, (1) a preparatory phrase to frame the task (problem), to provide key information about the system, to develop the database, etc., (2) a screening analysis focuses on the evaluation of statistical data, i.e. recently experienced major blackouts, and the identification of "obvious" vulnerabilities and interdependencies, e.g. by network-theory based topology-driven analysis. The third step, called in depth-analysis, calls for detailed modeling and simulation (M&S) techniques following a systems' approach. After careful evaluation a two layered agent-based /object-oriented modeling approach turned out to be most promising and was applied to the Swiss power system. A list of vulnerabilities and recommendations to reduce and manage them has been developed based on current detailed understanding of the system, evaluation of accident scenarios and analytical results. The susceptibility of modern SCADA-systems to malicious attacks often using non-dedicated ICT has found out be of tense of system operations (TSOs) and cooperation between them are the areas which deserve attention.

.    Probabilistic Risk Analysis and Bioterrorism Risk. Ezell B*; Virginia Modeling, Analysis, and Simulation Center   bezell@odu.edu

Abstract: There is only one logical and workable interpretation of probability and it is that of degrees of belief – G. Apostolakis (1990) Since the terrorist attacks of September 11th, 2001, and the subsequent establishment of the Department of Homeland Security (DHS), considerable effort has been applied to the challenge of risk analysis in the security domain. DHS, industry, and the academic risk analysis communities have all invested heavily in the development of tools and approaches that can assist decision makers in effectively allocating limited resources across the vast array of potential investments that could mitigate risks from terrorism and other threats to the homeland. While considerable progress has been made in approaches for terrorism risk analysis, there remain a number of challenges and limitations to each method currently in use. In this paper we explore a number of current and potential approaches for terrorism risk analysis, focusing particularly on recent discussions regarding the applicability of probabilistic, and decision analytic approaches in terrorism risk analysis. The National Research Council’s Committee on Methodological Improvements to the Department of Homeland Security’s (DHS) Biological Agent Risk Analysis (referred hereafter as the NRC Committee) has argued that because of this adaptive nature, alternative tools like decision trees are needed to assess the risks of terrorist events. While we do not take issue here with the possible value of these alternative approaches, we aim to make a case 1) that PRA is an important and useful approach for quantifying terrorism risks and has value in guiding risk management decisions; 2) event trees can be used as part of a terrorism PRA to decompose the universe of terrorism scenarios; and 3) decision trees, like all approaches, have limitations. In the case of applications for terrorism risk analysis, decision tree limitations may be difficult to surmount in that the adversaries’ objective functions and level of ability to predict tree outcomes are unknown.

.    Challenges in cross-domain risk assessment in homeland security. Bennett SP*, Ezell BC, Levine ES; U.S. Department of Homeland Security   steve.bennett@dhs.gov

Abstract: Since its inception following September 11th 2001, the concept of “homeland security” has become a multi-domain enterprise, covering preparation, prevention, response, and recovery for (at minimum) terrorism, natural hazards, and border/immigration issues. Since its birth in 2003 through this year, the Department of Homeland Security (DHS) has enjoyed an average annual budget increase of approximately 8% per year. In such a budget environment, prioritization is not as challenging as it is in the current environment in which budgets are expected to be more constrained. Consequently, as a risk-management organization charged with managing risk across many disparate domains, DHS requires sound processes and techniques for risk analysis to support decision making in a fiscally-constrained environment. While DHS has made much progress in its risk analytic capabilities since 2003, the most advanced approaches are for the most part, still applied to single or narrow groups of domains. This fact presents a significant challenge to decision makers responsible for “optimizing” the allocation of limited resources such that risk is best managed or reduced. While intra-domain risk analysis is important for resource allocation within a domain, cross-domain (or “all hazards”) risk analysis is perhaps the more important area of decision support required by decision makers, since it is intended to support the decisions related to broad strategic prioritization. This presentation will discuss the DHS decision making landscape, as well as many of the technical and organizational challenges associated with cross-domain risk analysis. Specific attention will be devoted to: variable risk perception and tolerance across domains within DHS responsibility, the challenges in probability/frequency estimation for terrorism/intentional events for comparison against historically-derived natural hazard frequencies, and promising recent work in CBRN terrorism risk assessment as an example.

.    Novel exposure pathways for poultry to H5N1 HPAI virus from wild birds and poultry products. Gale P*, Kosmider R, Irvine R, Cook A, Munyinyi D, Breed A; Veterinary Laboratories Agency -Weybridge, Surrey, KT15 3NB, UK   p.gale@vla.defra.gsi.gov.uk

Abstract: The pathways presented here form part of a qualitative risk assessment for exposure of poultry to H5N1 highly pathogenic avian influenza (HPAI) virus in different sections of the poultry industry in Great Britain (GB), Italy (IT) and the Netherlands (NL), given its release through wild birds or imported poultry meat and products. Exposure pathways were constructed based on epidemiological evidence and consideration of the possible fates of wild birds, and routes of dissemination of meat and products. Pathways will be presented for the transmission of H5N1 HPAI from wild birds and other wild animals, which may gain access to backyard and commercial flocks. Environmental pathways, through persistence or dissemination of virus in faeces or pharyngeal secretions are considered as well as routes involving companion animals (cats), domestic livestock (pigs) and insects. Quantitative data are available to populate the pathways, including viral loads in poultry tissues, virus shedding data and decay rates of the virus in water, fomite surfaces and poultry manures. In addition, questionnaires are used to elicit expert opinion on the qualitative risks of exposure through the component parts of the pathways. Results will be analysed to rank the risks to different poultry sectors through each route, exploring differences, if any, between the three countries. The exposure pathways investigated represent those already known through epidemiological investigation of H5N1 HPAI outbreaks in poultry, together with those identified through consideration of the possible fates of infected material including wild bird carcasses, infected viscera/tissues, respiratory secretions, faeces and poultry products. The role of rodents, for example, is potentially important since they have been shown to move the carcases of dead birds. This approach may identify new routes, such that the risks can be assessed and managed.

.    Norovirus Human Challenge Studies: “Many a slip twixt the cup and the lip”. Moe CL*, Teunis PFM; Emory University   clmoe@sph.emory.edu

Abstract: Quantitative microbial risk assessment for food and water safety relies on dose-response data for the target pathogen. In the absence of reliable animal models, human challenge studies have been necessary to determine the dose-response relationships for noroviruses. The results of a dose-ranging study and low dose study with Norwalk virus and a dose-ranging study with Snow Mountain virus will be presented. The factors that must be considered in the optimal design and implementation of these types of studies as well as the interpretation of the results will be reviewed. This includes ethical considerations to protect the rights and safety of the subjects and technical challenges with measuring “dose” and various aspects of host response and susceptibility.

.    Funding Food Science and Nutrition Research: Financial Conflicts and Scientific Integrity . Hentges E*; ILSI North America   ehentges@ilsi.org

Abstract: There has been significant public debate about the susceptibility of research to bias as a result of industry funding. The nutrition and food science community has faced similar issues; specifically that industry-funded research is biased towards results that favor the sponsors. Given the critical role that industry has played and will continue to play in the research process, the International Life Sciences Institute (ILSI) North America published conflict-of-interest guidelines for industry funding of nutrition, food, and food safety science research in order to ensure the integrity and credibility of the scientific record. Eight principles have been developed specifying ground rules for industry-sponsored research. These principles are intended to be dynamic, prompting ongoing discussion by the nutrition research community and further refinement. This presentation will provide participants with information on the nature of academic-industry relationships in research including identifying key issues, present views from different stakeholders, and offer a discussion of the eight principles.

.    The Management of Ethical Risk and the Ethics of Risk Management. Saner MA*; CARLETON UNIVERSITY, OTTAWA, CANADA   marc@saner.ca

Abstract: When we hear the word “ethics” at work, we think first of workplace situations: treating colleagues well or badly, handling the temptation to personal gain at public expense, and so on. Audit, HR, training and other oversight mechanisms are in charge of managing these so-called “ethical risks”. Within organizations, ethics is also very much present in other places, however. For example, in the policymaking setting where societal and stakeholder interests must be balanced. Or, we can find them at the boundary of workplace and public policy: how to advise decision-makers in an impartial way, for example. It is rare that the word “ethics” is used explicitly in these situations. Yet another range of ethics-related issues is even more “embedded”: What value systems do we use when we select endpoints in risk assessment and regulation? How do select standards in science and or statistical analysis? How do we establish standards of accountability where many people are involved in complex regulatory decisions and there is an incident of public harm? These are examples of what we could call ethics in risk management. The management of ethical risk and the ethics embedded in risk management and policy making are entangled. In this paper I will clarify this entanglement, provide a holistic assessment of the intersection of risk and ethics, and show how a naïve commitment to “accountability” not only raises the ethical risk but is also highly problematic from a pure ethics perspective.

.    Placeholder: for open discussion to complete the second (of two)sessions. Calabrese, Maynard, Jones, Lewis and others *; UNIVERSITY OF MASSACHUSETTS, SANOFI-AVENTIS, INTEGRATIVE POLICY & SCIENCE   stevencraiglewis@comcast.net

Abstract: As a generalized biological phenomenon, hormesis is now acknowledged in textbooks of basic and applied toxicology. Citation have grown from 16/year (ca 1980s) to nearly 2300 in 2008. Such growth indicates substantial acceptance among scientists. Beyond essential micronutrients (i.e., small amounts of certain metals are required for good health and longevity), hormetic responses are well-documented in radiation biology, as well as environmental, chemical and pharmaceutical toxicology. Species that display hormetic responses reach from single-celled organisms to complex plants and animals. Nevertheless, considerations of hormesis have not significantly influenced risk assessment. This two-part symposium will [1] review the state of the science, [2] expand awareness of recent scientific advances, [3] present a case for applying the hormetic principle in risk assessment, [4] show widening acceptance among scientists, and [5] challenge creative thinking about incorporating hormesis into risk assessment. Sessions will end with a full period of moderated discussion on scientific readiness*: to incorporate hormetic responses into risk assessment; to apply hormetic dose-responses in safety assessment and regulatory decision-making. *All participants are welcome; see short-list (from >150 inter-disciplinary thought-leaders, past-presidents, etc); any who register will be personally invited to participate; Linda Birnbaum, Mike Bolger, Tony Cox, Kenny Crump, Vicki Dellarco, Don Elliott, Bill Farland, Penny Fenner-Crisp, Bernie Goldstein, John Graham, George Gray, Dale Hattis, Roger Kasperson, Bob Kavlock, Dan Krewski, Granger Morgan, Warner North, Resha Putzrath, Christopher Portier, Peter Preuss, Joe Rodricks, Jon Samet, Jennifer Sass, Rita Schoeny, Bern Schwetz, Bill Sette, Bob Sielken, Mitch Small, Bob Tardiff, Kim Thompson, John Vandenberg, Vanessa Vu, Jonathan Wiener, Jim Wilson, Lauren Zeise, Hal Zenick, Rae Zimmerman.



[back to schedule]