CHRONICLE. Retraction of an article more than 4 years after its publication!

This article is taken from the monthly Sciences et Avenir n°937, dated March 2025.

A controversial article, signed by Professor Didier Raoult, among others, was retracted on December 16. Published in March 2020, at the start of the Covid-19 pandemic, it focused on the supposed curative effects of hydroxychloroquine. At a time when everyone was waiting for an effective treatment, it opened up beneficial perspectives for the general public concerned about their health and unfamiliar with the scientific approach.

Read moreMycoprotein is about to get louder and more aggressive

Yet, very early on, researchers highlighted the multiple breaches of good practice it contained. Remember that the study involved 26 patients, which was clearly too small to provide a reliable answer in the case of a disease whose case fatality rate was, at the time, around 5 %. Indeed, 5 x 26 % = 1.3, which is a little more than one statistical death in the sample... Bad luck!

Furthermore, 6 of the 26 treated patients were arbitrarily withdrawn from the study, one of whom died and two of whom went into intensive care. Finally, the placement of subjects in the placebo group was neither random nor double-blind, meaning that the doctors knew who had been treated. In addition, conflicts of interest with the journal in question, International Journal of Antimicrobial Agents, were evident since two of the co-authors are members of its editorial board.

Read moreJay Bhattacharya takes office as 18th director of the National Institutes of Health

Read alsoHydroxychloroquine against Covid: a seminal study from the Raoult era retracted

An inability to convince the general public

Given these breaches of scientific integrity, the study should have been retracted early on. It should be noted that retraction procedures are initiated in three cases: an honest error, fraud, or improper appropriation of another's work, such as plagiarism.

Here, we are clearly – and you don't need to be an expert to be convinced! – in the second scenario. How is it that it took four years to realize this? This is regrettable. But even more so is the inability of the scientists invited to appear in the media to convince the general public.

By Jean-Gabriel Ganascia, professor at Sorbonne University, in Paris, researcher in artificial intelligence at LIP6 (Sorbonne University, CNRS), former president of the CNRS ethics committee. Latest published work: AI explained to humans, Threshold, 2024.