I’m very grateful to Dr. Pasky Pascual, a lawyer and scientist, and recently retired career U.S. EPA expert. He was kind enough to guest write this blog with me, sharing his technical expertise on this topic, as well as his vast experience on chemical risk assessment policy. Thanks, Pasky!
A rather arcane and largely discredited chemical industry-sponsored publication is still circulating, zombie-like, at the Trump Administration’s EPA science meetings, despite being basically dead among experts. (See, for example, the December 2019 EPA TSCA Science Advisors report on 1-Bromopropane, p. 54, 62).
— Jennifer Sass (@JBSass) March 17, 2020
The paper is funded by the chemical industry, and co-authored by four people: Ted Simon, Yiliang Zhu, Michael Dourson, and Nancy Beck (the paper is cited as Simon et al. 2016). Nancy Beck currently serves as the Deputy Assistant Administrator for EPA Toxics office and has been the political appointee making Agency decisions on toxic chemical policies, many of which have drawn fire from scientists and public health advocates, as well as having drawn legal challenges. She was recently nominated to become the next Chair of the Consumer Product Safety Commission (CPSC), which plays an important role in protecting the public from dangerous products, including those containing toxic chemicals.
The Simon et al. 2016 paper — hereafter called the Beck-Dourson method — is published in Regulatory Toxicology and Pharmacology, the journal of choice for industry advocacy pieces posing as published science (see blog by EDF’s Richard Denison on Reg Tox Pharm’s corporate ties). The journal’s Editorial Board includes members with ties to tobacco, big pharma, and agrochemical companies (see Thacker, 2018). One of the members of the journal’s editorial board and an author on this paper, Michael Dourson, was a Trump nominee to run EPA’s Toxics office, but ultimately he was rejected by Congress for his career-long defense of highly toxic chemicals including TCE and PFAS “forever chemicals.”
The American Chemistry Council (ACC) — the trade association for chemical manufacturers, which reportedly spent almost $8 million lobbying in 2019 — was the sole funder of the Simon et al. (2016) paper. The paper is important, because it helps illustrate why Beck, who worked for ACC when she co-authored this paper — is the worst possible choice to Chair the CPSC.
The Beck-Dourson paper fails to follow best practices
The Beck-Dourson method claims to be a way to implement recommendations by the National Academies of Science to use a Bayesian benchmark dose method to enhance risk assessment, but instead it strips away health protections, weakens the acceptable exposure limit, and would result in increased hazardous chemical exposures to consumers, workers, and communities.
Best practices for risk assessment are described in numerous reports of the National Academies of Science over the last decade that have recommended that risk assessors apply a benchmark dose method. Properly applied, it can enhance study information. It can also facilitate presenting the results as a probabilistic distribution which can provide more information than a single risk estimate. Overall, these advancements can increase the reliability of dose-response modeling. The seminal National Academies report, Science and Decisions: Advancing Risk Assessment, recommended these approaches (NAS 2009), as did a more recent guidance of the World Health Organization (IPCS 2014, 2017).
While the Beck-Dourson method tries to hang its hat on its opening gambit, that it is the answer to the call of the National Academies of Sciences, the truth is that whereas the National Academies, the World Health Organization, and other experts have successfully developed Bayesian methods for dose-response modeling, Nancy Beck and her co-authors advocate applying a bent and broken version of Bayesian methods to the wrong end of the risk assessment.
The problem with the Beck-Dourson method is that it is applied to each individual Uncertainty Factor (UF), making assumptions without evidence for each distribution. This conflates variability in the data, with uncertainty. Inherent uncertainties in risk assessment are unavoidable and include things like differences between laboratory test animals and human biology, or differences in biology between individuals, or data gaps and lack of information. Is there any reason to believe this uncertainty, or degree of unknowing, translated into the UF, follows a predictable distribution (like a “bell curve”)? If there is, the authors have not stated it, they merely assume it without any basis. Since there is in fact no basis to justify this central assumption, their entire, proposed methodology falls apart.
While the WHO method also employs a probability distribution to each uncertainty factor, it properly separates the issues of uncertainty and variability, and bases its distributions on historical data, making it more accurate. Consider the variability to the response of different individuals to the corona virus, which is now being represented as a bell curve in many media outlets. Because all humans share an evolutionary history, it is reasonable to assume that most of us respond in generally the same way to the virus (the peak of the curve), with some being either more susceptible or less susceptible because of particular, individual biological mechanisms (the two extremes of the curve). Does the same assumption apply to how much we don’t know about the risks of a chemical? That our ignorance conforms to a distribution that can be mathematically simulated? Beck-Dourson fails to show this, or even attempts to justify what such a distribution might be. In short, applying probabilistic, Bayesian methods to uncertainties from organism variability is sound; using them to capture degrees of unknowing is without empirical basis.
The Beck-Dourson method puts people and communities at risk
I’ve extracted the values from the Beck-Dourson paper that presents 19 individual chemical examples (for some of the chemicals, both an adjusted RfD and RfC is presented) to show how their method stacks up against EPA’s existing chemical assessments from the Integrated Risk Information System, or ‘IRIS’ program. These are the evaluations that are used by regulators and decision-makers all over the U.S. and in many parts of the world when calculating clean up levels and exposure limits for factories, refineries, and other waste-producing operations.
The Beck-Dourson method increases the acceptable amount of exposure (called the Reference Dose, RfD, or Reference Concentration, RfC, if we ingest or inhale it) to a staggeringly long list of toxic chemicals, including some of the top 10 chemicals currently being evaluated under the Toxic Substances Control Act (TSCA). Moreover, over half of the chemicals (11 of 19) that the Beck-Dourson paper claims are safer are linked to cancer.
For some toxic chemicals, the Beck-Dourson method would increase the acceptable exposure limit (RfD or RfC) by up to 150-fold! Even dioxin, one of the most toxic chemicals known, would have a 3-fold increase in the acceptable exposure limit.
The acceptable exposure limit for Methylene chloride, a chemical in paint strippers that can kill an exposed worker within minutes, and also cause cancer, would be 8.3-fold “safer,” according to Beck and Dourson. And 1,4 dioxane, a cancer-causing chemical that is polluting drinking water throughout the U.S. would end up with an exposure limit 33x less protective of public health, under the Beck-Dourson method.
See Beck-Dourson paper, Table 3, extracted here. Note that a bigger RfD/RfC value means more chemical is still an ‘acceptable exposure’, thus pushing exposure levels that can harm people into the ‘acceptable’ range.
National Academies rejects the Beck-Dourson method
A 2019 National Academies report explicitly recently rejected the Beck-Dourson approach (NAS 2019). From the 2019 NAS report: “The committee also concurs with DOD’s use of the BMDL [benchmark dose limit] rather than the method recommended by Simon (2016) [the Beck-Dourson paper] of using the BMD with a log-normal distribution based UF. Although [the Beck-Dourson paper] recommended applying the UFs in a stage-wise fashion, the committee supports DOD’s decision to apply all UFs after derivation of the [Point of Departure].” In other words, any method – including the Beck-Dourson method- that thinks it can apply a Bayesian analysis to each individual Uncertainty Factor is misguided, and too flawed to be used.
Instead of the flawed and discredited Beck-Dourson method, the National Academies committee recommends using the probabilistic framework developed by the World Health Organization with experts that include current and former U.S. government career scientists (Chiu and Slob 2015; WHO/IPCS 2018; Chiu et al. 2018), “which provides a more comprehensive and rigorous framework for evaluating uncertainties than that of Simon et al. (2016).” (NAS 2019) In other words, Beck-Dourson BAD, and U.S. government career scientists GOOD. It’s actually that simple (see details at W.A. Chiu on Bayesian EPA-HQ-OPPT-2019-0236-0039).
Meet the real experts, the U.S. government career scientists
The EPA IRIS program experts conduct rigorous hazard-based assessment on many of the most dangerous industrial chemicals that pollute our environment and put workers, families, and communities at risk. Unlike other federal government evaluations, the IRIS program is unique in that it provides the acceptable exposure limit value—the RfD or RfC—that is used by decision-makers to set health-protective exposure limits, reduce toxic emissions, and determine clean up levels at hazardous waste sites. The IRIS evaluations may also inform product manufacturers and formulators wanting to make safer products. In short, IRIS sets the numbers that supports the regulations to protect people but may be costly to polluters. And that’s why the IRIS program has been a target of the chemical industry, including Nancy Beck, for years (see NRDC Delay Game report for details).
The National Academies has repeatedly endorsed the EPA’s IRIS chemical assessments (see NAS 2018, and NAS 2014). Its most recent report specifically defended the IRIS program’s methods, giving its near-unanimous approval (one lone dissenter was an industry consultant) of the IRIS inorganic arsenic evaluation, “EPA’s hierarchical, Bayesian meta-analysis approach is an accepted and contemporary technique … In the opinion of the committee, this Bayesian technique provides an appropriate means of determining the shape of the dose response curve using actual data from a body of evidence that includes low-dose exposures within the observed range of human exposures.” (NAS 2019) The National Academies is saying there that what the US. Government career scientist are doing is best practices, and it should continue its good work. Bravo, EPA IRIS program!
Beck and Dourson, discredited by science, elevated by Trump
This isn’t the first time that Nancy Beck has put forward proposals for how to assess the risk of chemicals that were then roundly rejected by the National Academies. In 2006, when working for the Office of Management and Budget (OMB), Beck issued a draft Risk Assessment Bulletin that purported to instruct all federal agencies how to conduct risk assessments. The draft prompted widespread criticism at its sweeping overreach of power to review Agency actions. The NAS called the draft bulletin “fundamentally flawed” and unanimously recommended that it be withdrawn, a move without precedent. It was withdrawn, at the time.
Fast forward a few years and Beck is installed by the Trump Administration to run EPA’s Toxics Office (pending the nomination of Michael Dourson which never succeeded). Beck notoriously re-wrote EPA rules for administering the recently revised Toxic Substances Control Act (TSCA) to weaken protections from asbestos, “forever chemical” PFAS, and toxic flame retardants. Her re-write of the rules was eventually thrown out by a Federal Court of Appeals (see blog by NRDC chemical policy expert Daniel Rosenberg for details). Beck has also been involved in re-writing EPA’s Science rules to make it harder to consider evidence of harm to humans from toxic chemicals, pesticides, air pollution and lead when setting health protections. The list goes on.
Everything Nancy Beck does has the same result—to weaken health protections from toxic chemicals. Her zombie-like pro-industry approach relentlessly seeks to undermine the protection of people’s health from chemicals that can cause cancer, learning disabilities and other harms. She should no more oversee protecting consumers than Dracula should oversee protecting the nation’s blood supply.
— Jennifer Sass (@JBSass) March 10, 2020
Jennifer Sass brings to her job a highly specialized expertise in U.S. chemicals policy. Much of her work is focused on understanding and explaining the science behind toxic chemical regulation and on advocating for regulations that are consistent with science, health policy, and environmental law. She frequently provides testimony and scientific briefings for members of Congress and federal advisory committees and is a lecturer at George Washington University’s department of environmental and occupational health. She has a master’s degree and a PhD in anatomy and cell biology from the University of Saskatchewan and has done postdoctoral work in toxicology at the University of Maryland. She is based in Washington, D.C.