Using shoddy methods, EPA says chemical is not risky

As many times as my parents used it, “because I said so” is never a good rationale– but that’s basically what EPA wants us to believe for why the chemical Pigment Violet 29 is not risky. We blogged previously that the meager available data does not support this conclusion, and EPA’s release of additional information- not once, but twice– only shows the data and the Agency’s methods are even worse than we thought.

We already know there’s a serious lack of data on this chemical. Does it cause cancer? There are no tests. Is it a hormone disruptor? Again, no data. Is it toxic to the brain? You guessed it– nothing. Even for existing Pigment Violet studies, a number had serious methodological problems that called the results into question, which we pointed out in our previous comments. Yet, in Nov 2018, EPA relied on these studies to say Pigment Violet is not risky.

Then, in April 2019, EPA did a 180, changing its ratings of two acute inhalation toxicity studies from “medium quality” to “unacceptable.” While this appropriately acknowledges the deficiencies in the studies, this left the Agency with no reliable studies to determine if Pigment Violet poses health hazards when breathed in (inhaled). EPA states that inhalation is the primary way it expects workers would contact Pigment Violet. Despite now having no data on inhalation hazards, EPA did not revise its conclusion that PV 29 poses no risks by inhalation. What?

Jump to June 2019, and EPA releases a new inhalation risk characterization analysis purporting to demonstrate that Pigment Violet does not pose inhalation risks. Problem is, the calculations start with numerous assumptions that lack empirical data, including that Pigment Violet is:

  • Poorly respirable
  • Poorly soluble/ low solubility
  • Poorly absorbed
  • Not metabolized
  • Not inherently toxic

Case in point regarding solubility, EU government scientists recently stated that the Pigment Violet water solubility data are “questionable,” that Pigment Violet could present persistence, bioaccumulation and toxicity concerns, and more data are needed.

All of these flawed analyses and conclusions get to a major underlying problem with EPA’s approach to the science. That is the TSCA “systematic review” method, and especially how it evaluates study quality.

In a recent peer-reviewed commentary, we analyzed the TSCA method and found it is inconsistent with empirically based, validated systematic review methods in use by other government agencies. We also looked at the available animal toxicity data on Pigment Violet and compared EPA’s study quality ratings to a pilot risk of bias evaluation using the Navigation Guide, our systematic review method that the National Academies recommended for use in chemical evaluations (for more on systematic review and risk of bias, check out this blog).

Below is what we found—in this type of table, more red means higher the risk of bias. And there’s a lot of red.

Fig-1-PV-29-blog.jpg
Figure: Initial evaluation of Pigment Violet 29 animal toxicity studies with the Navigation Guide risk of bias assessment, finding the evidence base is overall probably high risk of bias.

Our risk of bias analysis reveals that the overall evidence base of Pigment Violet animal toxicity studies is probably high risk of bias, in contrast with EPA’s conclusions that many of these same studies are “high” and “medium” quality. Our analyses indicate that the available data on Pigment Violet is inadequate for EPA’s risk evaluation both because of quality deficiencies and critical data gaps.

Is Pigment Violet risky? At this point, we don’t know, and neither does EPA. It is critical for the Agency to use the new authorities Congress gave it in reformed TSCA to obtain adequate data on Pigment Violet, complete a new risk evaluation, and ensure the protection of workers, communities, and other populations facing health threats.