Revisiting the rationale for holistic, integrated risk assessments

At the 2009 SETAC NA session “Learning How to Make Decisions that Integrate Social–Ecological Landscape Interactions,” we detailed the need for formalized approaches to environmental management decision making, especially as society faces the challenges of sustainability. A formalized approach is necessary because humans have an inherent tendency to focus on narrow slices of information (Gladwell 2005) and therefore are likely to miss important information, even diagnostic characteristics that intellectually they know to be important. Humans also tend to discount the occurrence of rare and unanticipated outcomes (Taleb 2007). Therefore, we suggest that some or all environmental management decision- making processes be subsumed within a holistic integrated risk assessment framework. The central themes underlying the need for such a transformative shift in the way we perform risk assessments emerge from 2 observations. First, the complexity of most of today's issues that a society is compelled to face involve nested ecological processes acting across wide spatial and temporal scale ranges. Second, the well-being of individual humans and communities extends far beyond mere exposure to chemicals. Either observation should be seen as a compelling reason to improve current practices; together they become not only compelling, but humbling. For example, consider the multitude of effects to fisheries, wildlife, and the overall social fabric along the Gulf Coast from the British Petroleum Horizon platform explosion (20 April 2010) and subsequent leakage of oil, all effects that are not directly linked to toxicity of the oil or the dispersants. Reviewers of our initial submission of this commentary expressed that integrated human health and ecological risk assessment was nothing new, that guidance already existed (e.g., UNEP-IETC 1996; USEPA 1997; Suter et al. 2001), and that many projects already were using integrated approaches. However, these documents speak mostly about planning for, and then sharing, data in both the ecological risk and human health risk components (e.g., measurements of chemicals in fish tissues could be collected in a way that provides whole-body concentrations for wildlife and filet portions for human dietary considerations). Although the occasional journal article includes “integrated” in the title, it is rare for holistic perspectives to have played a significant role in these assessments. Clearly, we did not communicate our concerns effectively about the way risk assessments are practiced, nor our suggestions for improvements. We appear to have aptly demonstrated Churchill's observation about America and Great Britain, that the 2 nations were divided by a common language. Central to the miscommunication are different understandings of what it means to conduct an integrated assessment. Initially, it would seem that “holistic” and “integrated” are redundant terms. Indeed, in anticipation of this we had deliberately coupled holistic with integrated, reflecting our experiences that practitioners tend to address single issues (stovepiping) with a belief that integration can be achieved by adding up the parts at the end of the exercise, amounting to a last-minute, post hoc combination of 2 or more stand-alone efforts. These post hoc efforts tend to miss opportunities to ask questions about systems linkages and interactions and fail to properly define the level of analysis needed to answer such questions. Our primary concern about the practice of risk assessment is this tendency for technical approaches that focus on single issues in ways that do not acknowledge the interconnectedness of complex ecological systems. We include socioeconomic and cultural dynamics of humans as embedded components within ecological systems. The disparate parts of fragmented assessments (e.g., one examining exposures to stressors in fish, another exploring wildlife, a third considering groundwater, and a fourth estimating exposures of humans) cannot be connected late in the process. The whole is different from the sum of the parts, a concept attributed to Aristotle approximately 10 AD and perhaps even to Euclid some 300 y earlier. Different questions are required at different scales; different data are needed to explore dynamic systems; different analyses are needed to develop robust conclusions. As environmental management becomes more challenging and more sophisticated, so does the need for truly integrated risk assessments, that is, ones emerging from a holistic perspective and structured using a systems-framework from the outset. Holistic approaches consider relevant, dynamic crosslinkages that exist within complex systems as well as the breadth and depth of affected stakeholders' values throughout problem formulation. The key is to get different data, not just more data. Typically, when facing a difficult decision, calls are made for more data, in the belief that the mere addition of information will reduce recognized uncertainties and thus make decisions easier. Although additional data may strengthen confidence regarding central tendencies and other statistical descriptors, their existence rarely eases the difficulties of decision making (see Taleb 2007). Whether dealing with economic considerations, cultural distinctions, chemical management, forestry, agriculture, or wildlife resources, it is imperative we acknowledge that any and all of these issues reside wholly within an ecological system. This presents us with one of the first great enigmas. Ecological systems are unbounded, abstract constructs with multifaceted, highly dimensional, and dynamic processes, composed of organisms and abiotic structures that interact through modulated feedback loops to yield a multitude of potential responses. One can rightly conclude that the complexity of ecological systems greatly constrains our ability to forecast outcomes with much certitude. Historically, holistic efforts tended to be restricted to qualitative descriptions and “spaghetti diagrams” illustrating the connectedness of all components and processes of a system, disregarding scaling issues, and largely devoid of mechanism and quantification of process flows. We suggest the need for a hybrid process, one that involves the construction of appropriately scaled and nested conceptual diagrams showing the qualitative relationships among valued entities with spatial context and temporal position (Froese and Orenstein 2010; McCormick 2010). A holistic and integrated risk assessment process also needs to include constructive iterative dialogue about those valued entities that leads to consensus decision criteria. A final honing of the nested conceptual models results in explicitly linking valued entities with decision criteria. This part of the process fosters development of transparent analyses accessible to all stakeholders, promotes equity of tradeoffs among stakeholders, and achieves consistency of management decisions for similar issues. Gladwell (2005) posits that humans are wired to make critical decisions quickly, and to do so we tend to ignore most information at our disposal. He characterizes this operational decision process as acting on “thin slices” of information. Whereas this approach to decision making had great survival value for our ancestors needing to elude a lion, or for us today as we attempt to cross a street with sporadic traffic, it does not serve us well as we engage in deliberative processes on wicked problems (Rittel and Webber 1973), that is, situations for which there are no right answers, but for which there are many wrong ones. Indeed, this tendency of humans may explain the disconnect that many scientists sense between their good work and the ultimate environmental management decisions; we simply see different thin slices. To counteract our predisposition to hone in on certain bits of information, and not necessarily the most critical ones, it appears that we require formal deliberative processes that curtail our instincts as we strive to make wiser decisions regarding environmental management options. Sufficient awareness of the challenges of environmental management existed immediately following World War II, prompting the World Health Organization (1946) to define “health” in the preamble of its founding documents as “a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity.” Despite this insightful appraisal of health, practitioners of human health risk assessment have scaled back to a much simpler, much easier, and arguably much less useful approach that focuses solely on risks from chemicals. We can make the case that current assessment practices actually do not even consider “risk” from chemicals because most assessments stop once some estimate of potential exposure to a limited selection of chemicals is produced. In other words, a tacit assumption has evolved that a potential exposure is equivalent to an effect. Going back to the examples of integrated risk assessment guidance, for example, the Framework for the Integration of Health and Ecological Risk Assessment from the WHO Workshop on Integrated Approaches to Human Health and Environmental Risk Assessment (Suter et al. 2001), it is this narrower definition of health that contributes to a belief that harmonizing the data collection and molecular–mechanistic approaches to chemical assessments for humans and wildlife fulfills the integration requirement. This approach, however, fails to consider the broad perspective of well-being, which arises from a balance of physical, social, and spiritual determinants of health (Health Canada 2001; WHO 2010). Foster consideration of all relevant perspectives Encourage recognition and/or acceptance that other perspectives exist and have merit Establish a dialogue about decision criteria Resist our predisposition to act on “thin slices” of information, but rather to consider simultaneously all agreed relevant inputs and all relevant perspectives Thus, taking a holistic and integrated approach to risk assessment becomes a more relevant consideration of human and ecological risks from exposure to all relevant stressors (chemical, biological, and physical). It also includes explicit consideration of social, economic, mental, and spiritual well-being. Moreover, the approach can produce information compatible with multicriteria decision analysis tools, that is, formalized, structured procedures that allow us to confound our propensity to act on thin slices of information. What decisions are to be made? Who will make the decisions? What input will other affected stakeholders have regarding the decisions to be made? What factors (e.g., cost, time) will influence the work to be done? Consensus on these questions, which cross the value spectrum of affected stakeholders, is highly desired. Several passes through these questions may be required, and variations of these questions are likely to be added to the list. Moving on to the technical components of Problem Formulation should be considered only after consensus on management goals and values has been achieved. A deep consideration of conceptual modeling and problem formulation processes will certainly provide a more robust elicitation of problems and a more fully integrated assessment of risk. However, to access the values espoused by the WHO when addressing the current and future state of individual and societal well-being, we need the holistic perspective offered by sustainability assessments. Much like the problem formulation process, deciding what is meant when saying “this is sustainable” requires careful consideration and iterative evaluation. Sustainability is about values. And values change with every person, every community, every organization, every country, every time. Sustainability is about the efficient and equitable translation of the ultimate means of production to the ultimate ends of human society (i.e., a process with the ultimate goal of conserving human society, rather than the planetary resources society depends on). To discuss sustainability for any group, regardless of size, we must first acknowledge what we value in order to answer these 4 questions: 1) what is it we want to sustain, 2) for whom do we want to sustain it, 3) for how long do we want to sustain it, and 4) at what cost are we willing to sustain it? Thus, from a systems perspective a unifying definition of sustainability would be “maintaining, or fostering the development of, the systemic contexts that produce the goods, services, and amenities that people need or value, at an acceptable cost, for as long as they are needed or valued.” Sustainability is a process, a way of thinking, that can get an individual, but more typically a group, organization, or society, to move toward agreed objectives. Sustainability assessments are intended to anchor management objectives within the realities of how the ecological system works, while still allowing a little wiggle room in meeting a broad range of societal objectives. Subsuming traditional risk assessment into a sustainability assessment framework allows us to continue with what is known and understood (e.g., EnRA, LCA), and puts those efforts in context with the greater societal need for effective and efficient use of ecological goods and services. Combining the holism of sustainability, the integration of all types of environmental stressors, and the expanded valuation of human well-being will serve to address the values of all stakeholders in any holistic, integrated risk assessment process. Our proposed hybrid offers a formalized risk assessment process to foster stakeholder respect, reach consensus on management decision criteria, and develop a project conceptual model that can guide problem formulation, data acquisition, and analysis. Our intent in proposing the broad scale adoption of a holistic and integrated risk assessment framework for all social–ecological landscape decision making processes reflects our belief that such a process will reduce overall cost for the project proponents, reduce regulatory oversight burdens, reduce stakeholder uncertainty, and increase the overall effectiveness of our use of ecological goods and services consistent with sustainability. Disclaimer—The peer-review process for this article was managed by the Editors and Editorial Board, without the involvement L. Kapustka.