EPA TSCA Systematic Review for Chemicals Is Fatally Flawed
By Jennifer Sass, Natural Resources Defense Council
The EPA Toxics Office is pushing a new approach to evaluating scientific studies that is setting off alarm bells with health scientists everywhere. It was developed in near-secrecy by the EPA office tasked with assessing chemicals under the new Toxic Substances Control Act (TSCA), the office now headed by former chemical industry lobbyist Nancy Beck. That may explain why it is so favorable to industry, so harmful to public health, and so inconsistent with global best practices.
The document is called, “Application of Systematic Review in TSCA Risk Evaluations” (see EPA’s website for details).
The TSCA Systematic Review conflicts with accepted methods in several important ways, including but not limited to:
- never been peer reviewed;
- flawed use of scoring;
- similar to the discredited Censor Science Rule;
- favors industry-sponsored studies;
- excluded non-industry studies.
These represent fatal flaws in the approach, rendering it hopelessly compromised by bias towards regulated industry sponsored studies.
Regulators, lawmakers, researchers, product manufacturers, and others want chemical assessments based on an evaluation and synthesis of all the evidence. This can include whole animal studies, cellular and in vitro studies (test tube and petri dish studies), and wide-ranging human data. All these study types have strengths and limitations; a systematic way to collect all the relevant information, assess the quality and reliability of each study, and then integrate all the studies together will lead to the most accurate assessment. A Systematic Review is supposed to be a systematic and transparent method to evaluate the quality of data and to support evidence-based decision making.
What is a systematic review?
Unfortunately, the TSCA Systematic Review document is less about evaluating the quality of evidence, and more about eliminating it altogether. It describes a head-in-sand approach to any study reporting that a toxic chemical is, well, toxic.
How will it hurt us?
Prior to the chemical industry’s takeover of the Toxics office, EPA career experts conducted a systematic review of non-industry studies that linked exposures to the toxic solvent trichloroethylene (TCE) during fetal development to congenital heart defects (see EPA 2016 Proposed Rule, p. 10-12). Based on this evidence, EPA managers under the Obama Administration proposed bans on the uses and products that posed the most dangerous risk.
In response, the chemical industry conducted its own review with funding to corporate consulting firm ToxStrategies from Entek International, whose Oregon-based battery parts operations has been repeatedly fined for violations related to its TCE pollution including allegedly poisoning its workers. Its review used an approach very similar to the one described in the TSCA Systematic Review and concluded that the studies showing harm were all flawed. This is a real-world example of how a faulty systematic review process can lead to the inevitable exclusion of studies that show the potential harm posed by a chemical from a supposedly neutral assessment.
In summary, many excellent studies demonstrating links between environmental exposures and adverse health effects will be excluded under the TSCA Systematic Review, resulting in an inaccurate and unprotective chemical assessment.
This proposed TSCA Systematic Review is only the latest gift from Nancy Beck to the regulated industry, which makes it increasingly difficult for the EPA to consider epidemiology and other human health data that links industrial chemicals and pollutants to adverse health effects.
Discrediting human health studies, and blocking their use by regulatory agencies, is a Big Tobacco strategy, as recently described in The Hill:
“The tobacco companies lost that battle, but it is now clear that the war did not end but simply shifted arenas…The same shady tactics are being used today by many of the same characters to try to create false uncertainty about the science demonstrating the public health risks of pollution…Using the false mantle of transparency to cast doubt on public health science and create barriers to regulations is not a new idea. It is a decades-old industry trope pioneered by tobacco companies concerned about the increasing number of studies linking tobacco smoke with health problems. Big Tobacco realized the best chance of avoiding regulation lay not in attempting to combat each study as it came out, but rather in changing the standards to which those studies were held. To do this, they sought to use transparency and reproducibility as rhetorical cover to raise questions about the reliability of the science on secondhand smoke.”
With this TSCA Systematic Review in hand, the TSCA program is now set up to follow the tobacco industry’s footsteps to benefit the chemical industry and other polluters, rolling back important health protections from toxic chemicals.
Here are the low-lights of the TSCA Systematic Review
Never been peer reviewed: This is the first time that the TSCA Systematic Review has been made public. It has not undergone a transparent, public, and rigorous peer review process. This violates existing EPA peer review requirements that require documents like the TSCA Systematic Review that are “highly influential,” “novel, controversial, or precedent-setting,” or have “significant interagency interest” to undergo peer review before being implemented (see the EPA Peer Review Handbook, and the OMB Peer Review Bulletin). Yet, it is already being applied to the TSCA chemical assessments currently underway (for example, see the 2018 TCE Problem Formulation, Section 1.4, p. 17). Moreover, it will be used retroactively to undo chemical assessments that were completed under the Obama Administration and found harm from toxic chemicals like methylene chloride, trichloroethylene, and asbestos.
Flawed use of scoring: The TSCA Systematic Review uses an arbitrary scoring system, which may classify studies as “poor” or “unacceptable” quality based on inconsequential missing information. Other systematic review methodologies do not use numerical scoring systems for assessing study quality, which the National Academies recommend strongly against using (NAS 2014).
Similar to the discredited Censor Science Rule: Although the TSCA Systematic Review fails to align with established chemical assessment methods, it does manage to dovetail disturbingly well with the discredited Science Transparency Rule, characterized by Reuters as a “concession to big business that has long requested such restrictions”. Both the Censor Science rule and the TSCA Systematic Review would make it hard or impossible for EPA to include important human health and toxicology studies it its chemical hazard assessments if there is any information that is missing or not made public (details in my previous blog). Ultimately, both proposals will hamstring EPA’s use of scientific information, which will severely harm EPA’s work quality and public credibility.
Favors industry-sponsored studies: The TSCA Systematic Review favors regulatory studies that are sponsored by industry to gain regulatory approval of its products, over hypothesis-driven research which is mostly publicly-funded and conducted at Universities. It does this by favoring studies that meet the information reporting requirements described in the Good Laboratory Practices (GLP) standards required for regulatory studies. The GLP standards were established for industry laboratories in response to serious widespread fraudulent practices documented by government inspectors in the 1970s. That is why it is a requirement only for industry-sponsored studies. In many cases GLP studies are not published, not subjected to public scientific scrutiny, and not independently peer reviewed.
There is widespread acknowledgement across the scientific and medical community that regulated industry sponsorship can lead to biased study design, biased study conduct, and biased reporting of study results—all leading to a favorable outcome for the regulated industry sponsor (Lundh et al, 2017). However, EPA’s TSCA Systematic Review fails to address financial or other conflicts of interest at all.
Excludes non-industry studies: While moving industry-sponsored studies to the front of the line, the TSCA Systematic Review will take no chances that a study showing harm might sneak into its chemical assessments. So, it is also setting an impossibly high bar for occupational and environmental epidemiology studies such as ones showing that leaded gasoline is harmful for children, or that air pollution (particularly from fossil fuel combustion) is linked to debilitating diseases and premature death. The TSCA Systematic Review has a lengthy list of “serious flaws that would make epidemiological studies unacceptable” (Table H-8, p. 231), and the failure to report on almost any aspect of the study is grounds for exclusion. For example, grounds to exclude an epidemiology study include if it fails to report the number of study subjects at each stage of the study, or the reasons why some study subjects may no longer be included—neither of which necessarily compromise the study quality.
This approach will exclude many older studies due to changes in reporting conventions over time. It would also exclude studies with rare health outcomes that may be significant even in a study with design or reporting limitations.
What should EPA do?
The EPA TSCA program should use a systematic review protocol that is established through a transparent and scientific process, and meets globally established best practices, such as the NTP-OHAT method or the Navigational Guide (NavGuide) and SYRINA for endocrine disrupting chemicals.
If finalized as proposed, the TSCA Systematic Review will not align with globally consistent criteria for classifying chemicals according to their health, physical and environmental hazards. We can then expect Nancy Beck and the chemical industry to press for adoption of its fundamentally flawed and industry-friendly approach by other agencies and other countries (including via trade negotiations). This will lead to weaker hazard labels, material safety data sheets for workers, and other hazard communication information, as well as chemical assessments—thereby weakening health protection worldwide.
What can you do?
Anyone from the public, including you, can submit comments to the docket at the www.regulations.gov website at this link: Docket ID: EPA-HQ-OPPT-2018-0210. Comments are due Thurs Aug 16th.
Let EPA political appointees know that no chemical evaluation ever got better by eliminating informative studies. Instead, EPA should be focusing on conducting accurate chemical assessments using all the available information in a way that is consistent with current best practices for systematic review including discarding arbitrary study scoring, assessing study quality not reporting quality, and addressing the risk of bias from industry sponsored studies.