This caused an instant and global media hype, which is understandable given that there is not much else against COVID-19 at the moment and colchicine is a generic drug. safe and very inexpensive – around 25-30 ¢ per pill, I’m told. But, precisely, it is perhaps because of the colossal expectations which were then created that we have been witnessing for several hours now very mixed reactions from clinicians and / or researchers.
Some, and not the least, have praised the authors of the study as “encouraging” – this was the case of Dr Eric Topol, from the Scripps Institute in the United States, which is a big name in medical research. But others, like Oxford University epidemiologist Martin Landray, have spoken of “masquerade” (travesty), of “All that for this”, etc.
How to explain such divergent reactions? It is that part of the benefits that the press release dangled did not cross the threshold of “statistical relevance”, as they say. When we compare two groups (here, those who gave colchicine and those who received the placebo), there is always a chance that the differences observed were not caused by the treatment, but are simply due to chance. There are tools in stats to estimate this probability, and a difference is only considered “significant” if it has less than a 5% chance of being due to chance – in research jargon, we say that the ” P value ”must be less than 0.05. Otherwise, the two groups are considered to be identical.
However, for the main measure of effectiveness that was planned in the study, that is the percentage of patients who had to be hospitalized or who died, this famous P value does not reach precisely this threshold, even if it does not miss it. not by much (it is 0.08).
But there you have it, explained Dr Guy Boivin, infectious disease specialist at Laval University and co-principal investigator of COLCORONA, a rather significant unforeseen event occurred during the clinical trial: “During the first wave of the pandemic [et l’étude a débuté en plein dedans, dès mars dernier, ndlr], he said, we had major supply problems for diagnostic kits, so much so that the government of Quebec had said to stop taking samples. [pour les tests PCR] except in hospitalized patients [alors que les patients enrôlés dans l’étude ne devaient justement pas l’être]. The instruction was then that if you have had contact with infected people and you have such and such symptoms, then you are considered positive for COVID-19. And that’s why we enrolled around 300 patients [sur près de 4500 au total] whose diagnosis could not be confirmed by PCR. “
Dr Boivin and his colleagues believe that some of these patients could have respiratory viruses other than COVID-19, and therefore that colchicine could not have an effect in them. A large part of the complications of COVID-19 can indeed be explained by a disproportionate inflammatory reaction, and colchicine is precisely known for its anti-inflammatory effects, but not all other viruses behave in this way. Last spring, Dr. Boivin also reminds us, there were still a fair amount of respiratory infections other than COVID-19 circulating (other coronaviruses, influenza, parainfluenza, etc.).
For this reason, Drs Tardif, Boivin and their colleagues saw fit to exclude these 300 or so patients whose COVID-19 has not been confirmed by PCR. And when we only retain the 4200 cases whose diagnosis is perfectly certain, well the result becomes “statistically significant” (not by many, but still):