See. Hydroxychloroquine did prevent corona deaths. Countless lives could have been saved. That was the conclusion of cross-thinking medium The Other Newspaper from a Belgian study that appeared last week in the – somewhat controversial in 2020 – scientific journal New Microbes and New Infections. In it, researchers from the AZ Groeninge hospital in Kortrijk write that treatment with hydroxychloroquine led to less mortality among their corona patients than among patients who received usual care.
Also read
French ‘miracle doctor’ wrote his own magazine
Side effects
Does this Belgian study now show that those findings are incorrect? “No, we cannot draw that conclusion from this observational study,” responds Casper Albers, professor of applied statistics at the University of Groningen. “They compare all Covid patients who received hydroxychloroquine in one hospital in Kortrijk with all Covid patients in all other Belgian hospitals who did not receive that drug. So you cannot say whether it is because of the hydroxychloroquine that fewer patients died in Kortrijk. Maybe people are healthier in Kortrijk, maybe there is less particulate matter in the air, or there are better doctors.”
In addition, Albers notes, 13 of the 352 patients stopped treatment due to side effects. “So the researchers should have taken 339 patients into account when calculating the mortality rate. Then the mortality rate in that group becomes higher.”
In the study, the Belgians did control for some differences between the groups, such as gender and age. “But there are always unknown factors,” agrees Jelle Goeman, professor of biostatistics in Leiden.
Both professors emphasize that in order to draw firm conclusions, a well-controlled clinical study is required in which patients are randomly assigned to a treatment group or a control group. Those are done. And from there it emerged: the drug doesn’t work.
It is not that remarkable that the Belgians found a positive result, says Albers. “The probability that their conclusions are true has a certainty of 0.01. This means: 1 percent of the time you will still find a false positive result. So this could just be the one study that finds a false-positive result. This study alone cannot dismiss the findings of all other studies.”
