Paris (AFP)
From simple error to characterized fraud, scientific work also has its mistakes, which the Covid pandemic has exposed to the light of day.
If serious problems remain rare, the race for publications favors breaches of "scientific integrity", explain its specialists.
On June 4, 2020, the Lancet announced the withdrawal of a study on hydroxychloroquine which claimed that this drug is ineffective against Covid and even dangerous.
Since its publication at the end of May, several scientists have publicly expressed their doubts about the reliability of this study.
The medical journal is prestigious, the molecule at the heart of heated debates: the scandal is global.
"The retraction of this article - rightly done - was an important moment, many people thought that scientific articles could not be trusted", deplores Elisabeth Bik, figurehead of scientific integrity, interviewed by AFP.
This microbiologist tracks down anomalies in publications (poorly designed tests, inconsistent figures, undeclared interests, retouched photos) likely to constitute "breaches of scientific integrity".
# photo1
These are all the rules that guarantee that research is conducted in an honest and rigorous manner, in strict compliance with the scientific process.
From the lab to the publication, "integrity control is the guarantee that science is carried out satisfactorily and therefore serves a purpose", notes Catherine Paradeise, sociologist, professor emeritus at the University of Marne-la-Vallée, in the Paris region.
- Funding -
Via Twitter, her blog or comments on the PubPeer platform, Elisabeth Bik makes her findings public, with the authors responsible for explaining themselves and / or for the publications to correct or even withdraw the articles.
Since 2013, she has reported nearly 5,000 articles and several hundred have resulted in corrections or retractions, says the scientist, heavily attacked on social networks since she pointed out what she considers to be anomalies in studies co-signed by Professor Didier Raoult.
# photo2
If the activity of Ms. Bik - and other scientists active on social networks - constitutes its most visible face, scientific integrity also has an institutional side, more discreet, within research establishments.
The concern for integrity, "it has existed since there have been researchers, in a certain way", notes Ghislaine Filliatreau, delegate for scientific integrity at the French Institute for Health and Medical Research ( Inserm).
But it was in the early 1990s, when the United States created the Office for Research Integrity (ORI) that it became institutionalized.
The concern was first of all financial, "because there had been enough scandals for the American legislator to decide not to put its financing anywhere", joined by private companies, also concerned about their funds, notes Ms. Filliatreau.
- Perverse effects -
France, for its part, set up a French Office for Scientific Integrity (OFIS) a few years ago.
In most research establishments, "integrity" officials deal with allegations of "scientific misconduct" against their researchers.
# photo3
It is in this context that the University of Aix-Marseille has been investigating since 2020 the publications of Didier Raoult and his team concerning the treatment of Covid by the very controversial hydroxychloroquine.
"When we do research, we must trace everything we do, it is fundamental in good practices (...), we must therefore be able to ask those concerned to show us their lab notebooks, their protocols, tell us who did what during the experiment, ”explains Ghislaine Filliatreau.
Serious cases (plagiarism, bogus experiments, invented results ...) are rare, underline the specialists, but the Covid crisis has clearly offered a new "sounding board to the difficulties of scientific integrity", underlined in March the senator French Pierre Ouzoulias, co-author of a report on the subject.
Because there is also a "systemic problem" of research conducive to shortcomings, underlines the report, like many researchers.
# photo4
Funding is very largely indexed to the number of articles published, which "produced perverse effects", explains Catherine Paradeise in particular.
With the risk of seeing the quantity take precedence over quality or "to work out a little with the experimental conditions" to save time, she says.
- From labs to courts?
-
A rate of publications further accelerated with the Covid with work made public "too quickly", notes Elisabeth Bik, who also advocates strengthening peer review.
If the Covid has made the general public aware of the importance of the scientific approach, it also makes many researchers fear that questions of integrity end up in the courts, since Didier Raoult filed a complaint against Elisabeth Bik for "moral harassment ".
# photo5
An investigation was opened on May 2, the Marseille prosecutor told AFP.
A "judicialization of criticism and scientific controversy" denounced in recent weeks by the National Center for Scientific Research (CNRS) and the Ecole normale supérieure (ENS) in France or in an open letter signed by hundreds of world scientists " committed to the transparency and integrity of research ".
© 2021 AFP