Environmental Health Sciences Decision Making: Risk Management, Evidence, and Ethics: Workshop Summary (2009)

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

4 Stakeholder Perspectives on Environmental Health Sciences Decision Making In all scientific endeavors, whether of an environmental nature or not, there is the need to balance conflict of interest, address issues of bias, and understand the ethical implications involved in research. Opinions differ as to how to strike such a balance and conduct science that is viewed as credible and sound. In this chapter, which covered the panel discussion of four stakeholders, the individuals considered issues of disclosure of conflicts, criteria to ensure equal weighting of research, and peer review of data. The overall discussion from this panel and the general workshop discussion are captured in the next chapter. Full Disclosure of Conflicts of Interest Several preliminary studies of Rofecoxib (Vioxx) had conflicting interpreta- tion of the same sets of data. Some researchers argued that the data supported an increased rate of cardiovascular disease, while company scientists suggested that the results could be attributed to possible confounding factors, including the impact of another chemical product. Eventually, this product was removed from the market. David Michaels of The George Washington University suggested that neither of these interpretations was a result of intentional misleading; however, the public may have been served better if, following these conflicting interpreta- tions, an independent review had been conducted. He further noted that eliminating all conflict of interest also would mean barring individuals who are employed by product defense or litigation support firms from serving on federal scientific advisory panels. These individuals should be viewed more as advocates than as impartial scientists, since scientists who are hired to defend products in a regulatory or legal arena are not paid to provide unbiased opinions but rather to promote the interests of the party that hired them, said Michaels. A central tenet of the process is full disclosure and publication of conflicts 45 

46 ENVIRONMENTAL HEALTH SCIENCES DECISION MAKING in publicly and privately supported science, asserted Michaels. Full disclosure implies having this information available to the public and not just leaving the decision about managing conflicts of interest to the judgment of an agency or an editor. Disclosure needs to encompass the entire research enterprise and needs to be applied equally to the publishing of research and the regulatory setting. The Credibility of Science Protecting scientific integrity and credibility given human fallibility is an ongoing challenge. Ultimately, all scientific findings must be judged on their merits, whatever the source of funding, argued Myron Harrison of ExxonMobil Corporation. In an ideal world, the people who sit in judgment should not have conflicts and should have the best expertise to render a decision. However, they need to be guided by an established set of rigorous criteria that must be equally applicable to research from all sources. It is a reality that the science used in public health is particularly unstable and uncertain, and therefore scientific disagreement and controversy should be expected. In the face of this uncertainty, other human factors, such as personal beliefs and values, often play a large role. Scientists are not usually trained in methods, such as argumentation, that try to establish particular and contingent “truths” in the realm of human affairs. Thus, there is often a misuse of empirical evidence to support decisions that are primarily value based, noted Harrison. The challenge remains: How can an agency optimize the credibility of sci- ence used in rule making? Some characteristics of good research can strengthen its credibility: • Using good lab practices and good epidemiological practices, which include such tools as research protocols, auditable data management practices, and publication of all results. • Protecting human subjects in all settings, including private institutions (oversight by institutional review boards also addresses the value and quality of research). • Applying rigorous peer review not only for the purposes of journals, but also separately in the rule-making process. • Disclosing all potential conflicts of interest. • Implementing strong management systems, including external reviews, to oversee the priorities and conduct of the research program.   There is a basic challenge to publishing all data, as studies reporting negative findings (lack of an effect) are more difficult to publish in peer-reviewed journals than those that show an effect. 

STAKEHOLDER PERSPECTIVES 47 Asymmetry in Decision Making Due to the challenges of a very high burden of proof faced by many regula- tory agencies, too few environmental health decisions are actually made, noted John Balbus of the Environmental Defense Fund. For example, under the Toxic Substances Control Act (TSCA), the current law governing industrial chemicals, the Environmental Protection Agency (EPA) must demonstrate that a chemical “presents or will present an unreasonable risk” before it can take any regulatory action. Yet there is no routine requirement for the maker of that chemical to generate data indicating safety, and the EPA must present evidence of potential harm even to require testing on a case-by-case basis. In practice, this presumption of innocence for industrial chemicals creates such a large evidentiary burden on the agency that it has essentially abandoned efforts to regulate them under the TSCA. An additional hindrance to environmental health decision making is the growing mistrust of risk assessment. When risk assessments were first put in place, the goal in general was to determine a level of exposure that was presumed, in the face of uncertainty, to be well below the level expected to cause harm. Over time, government risk assessors have been challenged by the regulated industry to increase the precision of risk estimates. But because of such factors as the substantial variability in susceptibility in the population and the reality that indi- viduals are exposed to many different environmental agents, determining actual risk—whether to the population as a whole or to any given individual—is an elusive and unrealistic goal. By pursuing precise estimates of actual risk, asses- sors are now more likely to end up with an inadequately protective outcome for an unknown percentage of the population. Furthermore, asymmetry in the regulatory process impedes decision making based on sound science, asserted Balbus. On one hand, the work of agency scien- tists in the early phases of the rule-making process undergoes intense scrutiny and review by expert scientific advisory committees. On the other hand, in the latter phases, such as during the finalization of air pollution standards, the agency must respond to and may even incorporate comments or data from studies that have not had to undergo such rigorous scientific review. Balbus called for equally rigorous review of all data that may be incorporated into the rule-making process. Data Development for Risk Assessment In order to move the environmental health decision-making process forward, there is a need to think strategically about how data can inform risk, noted Wil- liam Farland of Colorado State University. Focusing on basic instead of applied research and on disease-based instead of topic-based research creates an inability to generate the type of data necessary for rigorous assessment of chemicals. As noted often during the workshop, the absence of data does not equal the absence of risk. The United States needs to have a commitment to sound science while 

48 ENVIRONMENTAL HEALTH SCIENCES DECISION MAKING participating as a global partner in trade and guidance to developing countries, asserted Farland. Science is a moving target, and it is essential to think about what information is needed to inform decision making. One idea is the development of a systematic approach to working with data and weighing the evidence. In discussing scientific evidence, it is common to “take studies off the table” until the process reaches a point at which there isn’t enough information to make a decision. If a decision has been made, there is general reluctance in the United States to revisit the science and the decision, either because of antibacksliding regulations or the inability to change the regulation. To move forward, a new decision paradigm is needed in which there is the flexibility to take into account new insights and scientific information, asserted Farland. This approach would not create an environment in which the discussion of risk based on the information is avoided. Currently, although most state and federal regulations are not designed to protect individuals, they protect the public without defining what the public is or how many individuals constitute the pub- lic. As part of a new paradigm, researchers and policy makers would carefully consider whether current federal regulations are in fact designed to adequately protect individuals, especially those in vulnerable subpopulations. Any proce- dural change, noted Farland, is an opportunity to engage stakeholders on how these regulations are structured to address these populations and under what context. Finally, the paradigm should incorporate evaluation into the decision- making process, as assessing the impact of a decision is vital to the success of future decision making.