For once, I’ll have to react to the news. In normal times, this blog is devoted to evolutionary biology, but fortunately, I have some skills to speak to you about another very, very, VERY pressing subject: hydroxychloroquine, and COVID-19.
TL; DR: The author of this blog article disregards any data external to the field of science and medicine, that is to say would rather ignore the identity of the authors of studies on hydroxychloroquine for the benefit of data reviewed by the scientific community.
With the data from the latest studies that have been done on the subject, calling for a global rollout of hydroxychloroquine therapy (with or without azithromycin) risks killing more people than saving them.
To date, no studies have shown that hydroxychloroquine treats anything related to COVID-19. The studies done on the subject have still not given us proof that the treatment works, the methodological problems therein are much more serious than you think: we cannot draw any conclusions from the results, and this implies that we are in the process of making a mistake by defending this treatment at all costs.
As you are probably confined, you are aware that we are suffering from a pandemic which culprit is a big bad virus, SARS COV2, causing the disease COVID-19.
It managed to establish itself as the great destroyer of our global economy (and of our spring trips) thanks to its very high infectiousness rate (even more than the flu!), which therefore made it spread at a significantly high speed. Its mortality rate is still difficult to determine, but the most reliable data (to be nuanced of course) would give a rate of 0.25 to 0.8% of mortality. To put it in perspective, this is a small figure, we are very, very far from a fatal disease like tetanus (20 to 30%), smallpox (15 to 30%) or ebola (between 25 and 90%).
It is nonetheless worrying, because this mortality rate is to be reassessed when the health systems become engorged, which is inevitable in view of its overwhelming infectivity, hence the expression “Flatten the curve” that you will have seen everywhere (with in my opinion its most clear and graphic explanation here). It is therefore justified that all the nations of the world have opted for containment of the population to limit infections, so that hospitals have time to treat people at risk without finding themselves overwhelmed, while collective immunity takes place.
A treatment is currently being proposed against COVID-19: hydroxychloroquine combined with an antibiotic, azithromycin. Many make their decisions regarding this treatment, without necessarily having all the information.
It turns out that, with my background as a biologist, I am able to shed some light on the situation. I can read scientific studies, and I know how science works. I’m not going to expand on the humans who are behind this treatment, as I’m not qualified to make that kind of judgment, and I’m just not interested in it. However, I can tell you about truth and rational decision-making, and give you clues to pick your own way.
So let’s start.
I will focus on the first study first, then on their latest (and more precisely the reasons that the authors give to justify their method).
The first study can be found here: https://doi.org/10.1016/j.ijantimicag.2020.105949.
The question asked is: are hydroxychloroquine and azythromycin effective in improving the condition of patients with COVID-19?
The study had 26 people who were being treated with hydroxychloroquine, including 6 with the antibiotic (azithromycin), compared to a control group of 16 people.
We are immediately faced with a disturbing ethical problem: some of the patients were recruited after refusing to give consent to take the treatment. This means that those recruited were aware that they were taking the molecule, which of course had a placebo effect! As a reminder, the placebo effect is an improvement in the condition of a patient, who believes he is being treated, and this can have a very significant effect. The « treated » therefore have an advantage compared to the control group, which didn’t receive a molecule causing a placebo.
In addition, another part of the control group was neither treated in the same place nor under the same conditions as the « treated ». There are thousands of reasons why they would have different results, and thus cannot be compared.
The results therefore do not allow to draw a conclusion!
My second big problem with this study is the monitoring of the patients’ condition. Of the 26 patients, 6 were withdrawn from the study and final statistics. Of these 6 patients, 3 saw their condition worsen so much that they had to be moved to intensive care, 1 died (so his condition was serious). There were no explanations given other than « we could no longer follow them », which is not an acceptable answer.
This has an effect on the final statistics.
But I am not sure that we speak the same language.
I use the language of science. The vast majority of scientists I have heard speaking on radio, TV, in articles speak the language of science.
What we mean when we say “Based on this study, we cannot conclude that this treatment is effective on the disease” can actually mean “Mayhaps, if we search even more, we can find a true effect on the disease ”. And that’s true !
But.
This is actually the default position we have vis-à-vis everything.
And as there is nothing that says that hydroxychloroquine has an effect on COVID-19 (in other words, there is no evidence), taking this medicine is just as rational as the following treatments: taking a teaspoon of salt per take a glass of vinegar a day [1], taking a treatment of amoxycillin (another antibiotic)…
The possibilities of treatment are endless. And the treatment with hydroxychloroquine is at the same level of effectiveness as the above-mentioned treatments: there is no reason to prefer to treat yourself with one of these treatments compared to another.
What it means in the real world and in living humans is that:
Hydroxychloroquine (whether or not combined with azythromycin) does not work on COVID-19 (in the same way as a vinegar treatment) (until proven otherwise).
I’ll let you digest, and I’ll resume the article.
As for the 2nd and 3rd studies? They do not have a control group. They are therefore purely observational studies on the evolution of the disease.
Without this control group, the effectiveness of treatment still cannot be determined. The authors decided not to include a control group for 2 reasons:
- they are convinced of the effectiveness of the treatment
- they find it unethical not to treat patients with their treatment.
Their reaction is logical: if the treatment is effective, one has to use it for their patients.
These 2 studies starting from the principle that the treatment works, then they do not bring anything to this question. However, we have seen that we know nothing about it.
Fin.
In the end, this situation is very well known in the research world. It is defined by the concept of equipoise: at present, we do not know if the treatment is effective or not. So this is exactly the position in which it is necessary to do a clinical test, because the primary goal is to dispel this uncertainty. To find out more, I strongly recommend this article (in french) by Doctor Ferry-Danini, which explains this much better than I could at this address: https://medium.com/@ferry.danini/petite-introduction-%C3 % A0-l% C3% A9thique-des-trials-cliniques-d1b6d9f0bbb2
Several remarks you could make:
- « But we have to act, we don’t have time! »
Absolutely ! We must act ! But then why not try everything? Why is the vinegar treatment not widely used? Because it’s just a waste of time.
Again, for the same reason that you don’t deploy vinegar treatment, you don’t have to deploy hydroxychloroquine treatment. It is not logical.
2. “He at least proposes something…”
He is not the only one ! Researcher teams around the world are making assays right now to find a treatment! I can cite the DISCOVERY program, the SOLIDARITY program… but these do things so as to obtain a significant result! My greatest support to them.
3. “He knows! He is an infectious disease specialist with a great career! ”
And many other infectious disease specialists with a great career say the opposite of what he says. Who to believe?
It is an argument of authority. In science, the personality who proposes the hypotheses is not an argument to determine their veracity. We do tests, we do math, statistics, and we try to get as close as possible to the truth. We therefore operate by scientific consensus: if many scientists are convinced by the rational arguments given by someone, then this result falls into the realm of « plausible ».
4. “How can you rejoice at his failures? You are a monster and have no sense of priorities! There are people who die during this time! ”
I do not rejoice. In fact, I may be living in my bubble, but I don’t know anyone who is happy to see the current situation, where people are actually dying. Among doctors and researchers, I do not know anyone who is happy that there is no treatment for this disease.
5. « We must act, we have nothing to lose! You have to give this medicine, just in case! »
In reality, we have much to lose.
Remember the goal we want to get closer to: minimize the number of deaths.
It turns out that this drug is used by part of the population, including lupus patients. This extremely serious disease can be fatal when left untreated.
The craze around this molecule is creating a worldwide shortage for these patients, putting their health at risk.
Let us assume that the treatment does not work on the disease: the shortage would accelerate, the patients would not survive more (no life is saved), the lupus patients would see their conditions worsen and some would die (we actively kill). We arrive at a negative balance.
6. « But it’s a medicine, and we’ve known it for 40 years, we know it can cure »
Two points. First, amoxicillin has also been known for more than 40 years, but it has not been prescribed for the pandemic. Why ? Because there is no reason to believe that it works to cure the sick of this disease precisely. The same can be said about a treatment with hydroxychloroquine.
Second, this drug is actually known. It is however known for its effect on extremely different diseases, and has known, serious side effects, especially in certain populations. There are already deaths caused by self-medication. So again, it’s not trivial, even dangerous, to expose so many people to this treatment.
In the end, let’s make this last reminder: nobody wants to see more people die. Not the researchers, not the doctors, not even Big Pharma (we always forget that the people at Big Pharma have family too). We all have loved ones we are worried about. It is not by a whim that we do not want this treatment to be generalized. But truly because our goal is that as few people as possible die.
Sources and other prime choice reads :
The amazing Washington post article with incredible automatically generated graphs by the sheer magic of mathematical modeling: https://www.washingtonpost.com/graphics/2020/world/corona-simulator/
The truly beautiful statistical review on the first study:
Dahly, Darren, Gates, Simon, & Morris, Tim. (2020, March 23). Statistical review of Hydroxychloroquine and azithromycin as a treatment of COVID-19: results of an open-label non-randomized clinical trial . (Version 1.1). Zenodo. http://doi.org/10.5281/zenodo.3725560
[FR] L’article de Juliette Ferry-Danini sur le concept d’équipoise :
[FR] L’article de Florian Gouthière rappelant l’intérêt du groupe contrôle : https://www.liberation.fr/checknews/2020/03/30/pourquoi-la-nouvelle-etude-de-didier-raoult-fait-elle-debat_1783607?utm_medium=Social&utm_source=Twitter#Echobox=1585589025
A bonus article talking about a TIGER that caught the COVID-19 in a Bronx zoo, explanations from veterinarians and specialized researchers : https://www.wired.com/story/tiger-coronavirus-bronx-zoo/
[1] Plus, there apparently is anosmia and dysgeusia respectively in 86 and 88% of patients, it’s even easy to swallow ! [FR]https://www.lequotidiendumedecin.fr/actus-medicales/recherche-science/covid-19-une-anosmie-chez-86-des-patients-et-une-dysgueusie-chez-88-selon-une-premiere-enquete