- An article, which appears this Thursday in Clinical Microbiology and Infection, shows, by analyzing 29 studies, that the famous hydroxychloroquine has no effect on Covid-19 patients.
- Moreover, for the first time, this work reveals that the cocktail hydroxychloroquine + azithromycin increases the risk of dying.
- Nathan Peiffer-Smadja, one of the authors of this meta-analysis, explains why this is an important step.
Never (doubtless) have the French developed such a passion for molecules with complex names. Should you take hydroxychloroquine with azithromycin in the face of coronavirus? After many twists and turns, a study, which appears this Thursday in Clinical Microbiology and Infection, could (again) reignite the debate. Or close it.
According to this meta-analysis, taking hydroxychloroquine against Covid is useless. Worse, taking the hydroxychloroquine + azithromycin cocktail increases the risk of dying. One of the authors of this article, Nathan Peiffer-Smadja, infectious disease specialist at Inserm, explains why this publication should attract attention.
Methodology question, how did you carry out this meta-analysis?
We followed a very precise method by launching keyword searches on several databases, to extract all the articles on the subject. We then selected 29 studies, with 12,000 control patients, 8,000 patients treated with hydroxychloroquine + azithromycin, and 11,000 treated with hydroxychloroquine.
We submitted it for publication in an international scientific journal; the editor-in-chief of the journal, as well as two "reviewers", reread our study. I am specifying this because, in recent months, there have been articles published in the media before being peer-reviewed and validated for publication.
What is the result of your study?
We compared groups of patients treated with hydroxychloroquine, hydroxychloroquine and azithromycin, and control groups, who had not received any of these molecules. We get two big results. The first is that hydroxychloroquine is not effective in treating Covid-19. Another major result: hydroxychloroquine associated with azithromycin increases mortality. In detail, the risk of dying increases by 7% in absolute terms, and it is increased by 27% in relative terms. That is to say that if a patient has a 10% risk of dying [depending on his pathologies, history, lifestyle, etc.], he will have a 12.7% risk of dying with this treatment.
Does this mean that hydroxychloroquine alone is not dangerous?
Hydroxychloroquine does not increase mortality. But three randomized controlled trials, the best way to assess effectiveness, show that the side effects increase: more diarrhea, heart problems or liver problems.
There have already been a lot of studies on this treatment, and even a meta-analysis. How is your work new?
The latest meta-analysis summarized 8 studies (compared to 29 here) and was early, with no randomized controlled trial. It was therefore less robust than this one. Above all, we are the first to prove that this combination significantly increases the mortality of Covid-19 patients.
What does this change for the search for a treatment for the coronavirus?
What this is changing is already for patients: we must not use hydroxychrloroquine to treat Covid-19. And we must absolutely avoid the combination of hydroxychloroquine and azithromycin. It's consistent with what we thought, but here we have scientific proof. These two treatments lengthen the QT interval (the time between Q and T waves) of the heart rhythm, thus increasing the risk of a serious disorder. Especially since the Covid causes heart attacks.
This study is final. This treatment has been the subject of a lot of debate ... It's pretty terrible, because it has hampered clinical research on other treatments and put people at risk. For the moment, there is no effective treatment against this pathology. We only know that in severe pulmonary forms, the use of corticosteroids is effective. And it's a randomized clinical trial that has shown it!
Do we know if this increased risk varies according to the age of the patient?
There is no major difference, neither according to the age and comorbidity of the patient, nor according to the dose of treatment used. But these data were not necessarily provided by the studies carried out.
A Belgian observational study has just revealed that low-dose hydroxychloroquine would have an effect on mortality. So the debate is not closed?
In this Belgian study, like those of Prof. Raoult's team, people treated with hydroxychloroquine are younger, have fewer comorbidities and received more corticosteroids than in the control group. In observational studies, data from treated patients is used to generate hypotheses. But what makes it possible to make a therapeutic decision are randomized trials and meta-analyzes, the top of the pyramid of evidence-based medicine.
A French study published in June suggested that hydroxychloroquine would reduce the time spent in hospital. Is this an aspect that you have explored?
No. But out of the seven randomized trials on hydroxychloroquine alone, none concludes with a decrease in the time to hospitalization, and an English study even concludes with an increase in the length of hospitalization.
Hydroxychloroquine has been the subject of intense debate in France, in particular around Professor Didier Raoult. What to say to the French who doubt?
We must trust authorities and quality studies more than a person. The problem was focused on a treatment, while the real question is the evaluation process. And it is the same for all molecules. Didier Raoult made a non-rigorous evaluation. To know if a treatment works or not, you have to give yourself the time to evaluate. Otherwise, we risk endangering people. So we need trials where we draw lots of patients who have the "candidate treatment" and others not.
Is this arm wrestling around a treatment exceptional?
What is truly exceptional is how much the public has embraced these issues. Some made them believe that it was up to them to choose their treatment. I took part in therapeutic trials in Guinea on Ebola, and nobody started to say "we need such a molecule".
What I had already observed, on the other hand, is a questioning of the scientific approach, even of clinical trials. Some people feel that you can't draw lots of patients. It is true that there are very fatal pathologies (like Ebola, with 80% mortality), for which we will not draw lots between one treatment and nothing, but between two treatments. For Covid-19, a diagnosed patient has a 1% risk of being hospitalized. It is therefore necessary to be very demanding on the tolerance and the absence of toxicity of the molecule.
Do researchers need to communicate better with the general public?
Obviously. The tools of a rigorous scientific approach are essential, but we absolutely need good communication, and this has been lacking in recent months. The coronavirus has revealed that many people do not understand what we are doing. Therapeutic choices in the hospital are made on the basis of these studies. We must work to forge a link between medical research and the general public.
Health
Coronavirus: Is a new scientific study rehabilitating hydroxychloroquine?
World
Coronavirus: Hydroxychloroquine and chloroquine not effective and even harmful, according to a large study
- Treatment
- Medicine
- Didier Raoult
- Research
- Coronavirus
- Covid 19
- Health
- Science