"You’ve got to think about what kind of decision will be made with these risk assessments. You want to match the assessment to the question you are trying to answer," said Zeise, who is chief of reproductive and cancer hazard assessment at the California's Office of Environmental Health Hazard Assessment but was speaking only as a member of the national committee. "Making this an explicit exercise frames the degree of the analysis and the complexity and what questions should be asked."
The scientists proposed bringing regulators, scientists and stakeholders together before the assessment begins to set a deadline and determine what key questions need to be addressed. They warned, however, that they are “mindful of concerns about political interference in the process” and that it is imperative that risk assessments “not be inappropriately influenced” by regulators and others.
“This needs to be a formal process, open and transparent, with clear deadlines," Zeise said.
Another key recommendation is that EPA set a clear standard that requires strong evidence whenever the agency decides to disregard common scientific assumptions. Under the committee's proposal, the EPA could cast aside these so-called "defaults," or assumptions, only when the alternative is "clearly superior." This, for instance, would mean that the EPA would have to consider animal cancer data relevant to humans unless there was new scientific evidence for a chemical that clearly outweighed that long-held assumption.
“We thought the EPA's approach was unclear and we raised some concerns about it. What we’re suggesting is that they develop an evidentiary standard. Assessments can get bogged down quite a bit because of this whole issue of defaults," Zeise said.
In the report, one unidentified scientist on the committee criticized a new EPA policy that allows the agency to ignore sound scientific inferences whenever new data questions them. The scientist called it a "definitive and troubling shift" that can lead to "the paralysis of having to re-examine generic information with every new risk assessment."
The agency also needs a consistent approach to handle scientific uncertainties and to ensure that the differences between individuals, such as infants or pregnant women, are considered, the panel reported. Many U.S. policymakers, in the absence of certainty, decide against regulating some substances. The panel said while uncertainty cannot be eliminated, it can be better communicated to regulators.
Cranor said he is most enthusiastic about the recommendations to focus on how people's individual risks vary. He said it is "very important" for the EPA to gauge the dangers that chemicals, such as perchlorate, pose to children, who have a bigger body burden of contaminants, and people with diseases. “Things like this have to be taken into account. I think this NAS report opens up that possibility," said Cranor, author of the book Toxic Torts: Science, Law, and the Possibility of Justice.
The scientists told the EPA to put more emphasis on non-cancer effects of contaminants, to develop more sophisticated ways to analyze cumulative exposures to multiple chemicals, and to overhaul how it determines what dose of a chemical is considered safe. The agency should examine each chemical's effects on human health, such as how it contributes to a disease, and not assume there is a minimum exposure, or threshold, for causing effects, the committee said.
Don't wait for new technology
Because the changes are substantial, the panel recommended that the EPA phase in the new strategy with a series of demonstration projects.