It is a "gift" left by the former administrator of the US Environmental Protection Agency (EPA), which takes place in the US as a regulatory agency as well as a ministry of the environment. Three months before being disembarked in July 2018, after a series of scandals – exaggerated spending, favoritism and suspicion of corruption – Scott Pruitt had launched a project to reform the EPA for "Enhance the transparency of regulatory science" and modify the procedures for evaluating and regulating products (pesticides, etc.) or industrial practices (effluent discharges, etc.) that may affect public health or the environment.
On Tuesday, November 26, the editors of six of the most prestigious journals or family of scientific journals (Nature, Science, Cell, Proceedings of the National Academy of Sciences, PLOS, The Lancet) issue a joint appeal warning against this project. It aims to exclude from the information considered by the EPA in its analyzes studies whose data are not publicly available. The reform project, hailed by the American industrial circles, could be a "Disaster" according to the six signatories of the declaration, who fear the impending promulgation.
This one, they write, "Would be used as a means of removing relevant scientific evidence from the decision-making process, especially decisions affecting public health". Epidemiological studies, for example, associating an increased risk of disease in individuals more or less exposed to a particular product, could de facto be excluded from analyzes conducted by EPA scientists, on the grounds that they do not disclose all the personal data of the subjects enlisted in the study. Such a requirement of transparency would lead to the exclusion of a large part of the observational studies now widely used to guide or define actions or regulations for the protection of health and / or the environment.
"As leaders of peer-reviewed scientific journals, we support the sharing of research data, but we also recognize the validity of scientific studies of which absolutely all data can not be indistinctly shared, for reasons of confidentiality, write the signatories. Datasets that include personal identifiers – including genome-wide studies of thousands of people to identify genetic variants – are just one example. "
