EPA Advancing Transparency Rule as Science Board Pushes Back

The Environmental Protection Agency is planning to finalize a rule later this year requiring that all data and models from certain types of studies be made “publicly available in a manner sufficient for independent validation” before the agency can use them to defend major regulations. EPA has justified the rule as a way to increase the transparency and quality of the regulatory process in view of widespread concerns over the replicability of scientific findings.

Introduced in April 2018 by former EPA Administrator Scott Pruitt, the proposal has attracted widespread criticism, including from many science organizations and public health groups who argue it would hamstring regulatory action by unduly restricting the scientific research available to the agency. EPA’s Science Advisory Board (SAB) also established a work group to examine the rule shortly after Pruitt announced it, complaining they had not been consulted or even informed about it despite the relevance of their expertise.

At an SAB meeting last month, current EPA Administrator Andrew Wheeler resolved to consult with the board more frequently and asked it to weigh in on how the agency should handle proprietary and personally identifiable information under the transparency rule. However, the board voted at the same meeting to undertake a more expansive review, though Board Chair Michael Honeycutt acknowledged little time remains to complete a thorough examination of the rule before EPA is expected to finalize it.

SAB members skeptical of rule’s scope

Although the transparency rule was first proposed by his predecessor, Wheeler has made clear that he supports it. He told SAB at its meeting last month,

Maria Doa, a senior science advisor at EPA, offered further details on the agency’s progress in implementing the rule, saying the goal is to finalize it by year’s end. Summarizing the nearly 600,000 comments received, she said many addressed the variety of ways that concepts such as “validation” and “replication” can be defined, an issue the board’s work group also raised in a memorandum last year. Doa did not say how EPA plans to define the terms, though she indicated her familiarity with a recent National Academies study on replicability and reproducibility in science.

The work group noted the rule’s “highly controversial” nature, citing its similiarities to legislation that advanced in Congress on largely party-line votes. Among its other concerns, the work group also observed that the EPA did not assess the proposed rule’s impact on its regulatory processes, and noted the rule could “have the effect of removing legal, ethical, and peer-reviewed studies of health effects as sources to support the agency’s regulatory efforts.” While recognizing the rule “might stimulate researchers to make stronger efforts toward transparency so that their work may be considered in regulatory deliberations,” the work group suggested, “It might be easier to accomplish the rule’s objectives if the focus were on future studies rather than on studies that are already designed and published with terms that make complete transparency difficult or impossible to accomplish.”

Doa affirmed at the meeting that the rule, as proposed, would apply both “retrospectively and prospectively for dose-response data and models.” She also said it requires that “all underlying raw data and computer codes, regardless of who generated or funded it,” be publicly available for a study to be usable.

SAB member Richard Smith, a biostatistician at the University of North Carolina, cautioned EPA against requiring full access to all raw data, remarking, “Raw data can go back, in some cases, 50 or 100 years. You can be talking about individual experimenters, what they do in their labs. I do not see that as being remotely practical.” Smith urged the agency to consider limiting the requirement to the final “analysis dataset” on which a given study is based. Doa, though, stated that, “based on the proposed rule as written, the data would not be aggregated.”

Relations rocky between EPA and SAB

The relationship between EPA and SAB has been a subject of concern throughout the Trump administration. Months before EPA announced its transparency rule, it also instituted new requirements for membership on all its advisory panels, including that members may not simultaneously hold EPA grants. While EPA framed the requirements as a means of preventing conflicts of interest, critics condemned it as a way of curbing the influence of academic experts and increasing the prevalence of industry perspectives. Pruitt appointed Honeycutt, director of the Toxicology Division of the Texas Commission on Environmental Quality, as SAB chair the same day he announced the policy change.

Notwithstanding the shakeup, EPA’s handling of the formulation and rollout of the transparency rule, among other matters, has left a wedge between it and SAB. In April this year, Wheeler wrote a lengthy letter to Honeycutt and the board, referencing various concerns SAB has raised and resolving to work more closely with them. He stated he has asked the EPA’s Office of Policy to “strengthen” certain principles within the agency’s regulatory development process, including its commitment to “timeliness and early notification,” to “transparency and consistency,” and to coordination among its various advisory panels. The board, he noted, will receive “more rapid and frequent” briefings on major regulations “soon after they are proposed” as well as “specific charge questions.” Nevertheless, Wheeler also stressed that final policy decisions rest with EPA alone and will reflect “a wide range of considerations” beyond scientific ones.

At the June SAB meeting, some public commenters asserted that the relationship between EPA and the board remains inadequate. Bernard Goldstein, who led EPA’s Office of Research and Development during the Reagan administration, argued Wheeler’s narrow request for SAB to review how its transparency rule would handle personal and propriety information was unacceptable, telling board members ,

The SAB held its vote on whether to conduct a broader review than requested at the end of the meeting, approving it nearly unanimously.