A note on the Joint statement on EPA proposed rule and public availability of data

Today a group of editorial leaders from several  publishers, including PLOS, have co-published an Editorial highlighting concerns about the “Strengthening Transparency in Regulatory Science” rule proposed by the U.S. Environmental Protection Agency (EPA). This comes eighteen months after the same group published an Editorial when the rule was first proposed, in April 2018. 

Our concern is that the EPA proposal will eliminate from consideration in policy-making evidence from many peer-reviewed studies of prime importance to public health. The new proposed requirements do not reflect the norms or capabilities of the field and overreach even PLOS’ progressive data sharing policy.

PLOS has always supported open sharing of research data.The 2014 update to the PLOS Data policy introduced a requirement for data sharing at publication and was the first of its kind to be applied by a publisher broadly across many disciplines. Since then, PLOS journals have published more than 124,000 articles with a data availability statement indicating where and how relevant data can be accessed, and many other journals have followed suit by adopting encompassing data policies. We continue to champion data sharing as an essential practice of Open Science that supports reproducibility and re-analysis thereby increasing trust and fostering faster scientific developments. 

But our policy also recognizes that open public sharing of research data is not always possible, especially in studies where participant privacy can be at risk. Where that may be the case, we consider exceptions and we work with authors, editors and reviewers to find ways to satisfy the need for scientific scrutiny while protecting research participants’ privacy. This sometimes means that the data cannot be made publicly available and our experience shows that finding appropriate solutions can be challenging and costly. Too often, researchers are faced with the limitations of the existing infrastructure available to them for sharing and managing data derived from research on human health. We acknowledge these challenges and will continue to work with stakeholders towards better infrastructure to responsibly share sensitive data. 

While the EPA proposed rule recognizes the value of transparency in research, it is designed to apply beyond publishing research into that of regulatory decision-making. Rule making should be informed by the best available evidence, weighted appropriately on the basis of factors that may affect its level of uncertainty or readiness for generalization. Imposing a stringent standard of full data availability for scientific evidence before it can be considered by the EPA would have the consequence of limiting the pool of relevant, peer-reviewed scientific evidence that informs policy-making, both in its number and its diversity. This would have disastrous consequences for public health.

For these reasons, and while we reaffirm our commitment to championing open data sharing, we feel compelled to express our concerns about the EPA proposed rule in this Editorial and we encourage US-based readers to submit their own opinions through relevant legislative channels. 

Editorial: Thorp HH, Skipper M, Kiermer V, Berenbaum M, Sweet D, Horton R (2019) Joint statement on EPA proposed rule and public availability of data (2019). PLoS Med 16(11): e1003014. https://doi.org/10.1371/journal.pmed.1003014