Thread by @GertrudRey

Hydroxychloroquine (HCQ) Does not Work for Treating COVID-19. THE END.

***Warning! This post may be an emotional trigger for some! I’m just presenting the facts.***

Long thread. (1/n)

Sometime at the beginning of this never-ending year the president announced that HCQ was a “game-changer” for treating COVID-19. While I would be hesitant to embrace any drug as a game-changer without any real evidence, at that time, I did read a couple of studies... (2/n)

...done in Marseille, France, and felt cautiously optimistic about HCQ. When those studies came out, I posted about both of them and will summarize them briefly here. The first study showed that the drug can reduce viral load in a small number of COVID-19 patients. (3/n)

The second study was bigger (80 patients) and showed that treatment with HCQ not only led to reduction in viral loads but also to clinical improvement in all but two patients. (4/n)

Based on this, the FDA issued an Emergency Use Authorization for use of HCQ in the treatment of COVID-19.

Now the real test of whether a certain drug has a certain effect has to include a number of factors. (5/n)

Such a test would be a prospective, randomized, controlled, and preferably double-blinded trial using a large number of patients (large sample size). Prospective means that it looks forward to future results and is in contrast to a retrospective study... (6/n)

...which looks back at events that have already happened. In a retrospective study you can plan and control for various factors (described below), while in a retrospective analysis, everything has already happened, so you can’t apply those controls. (7/n)

Randomized means that participants are randomly assigned to experimental or control group, so that the only expected difference between the experimental and control groups will be the outcome variable that is studied. (8/n)

This eliminates unwanted effects that have nothing to do with the variable being analyzed. Double-blinded means that neither the investigators nor the subjects know who is receiving a particular treatment. (9/n)

This leads to more authentic conclusions because they reduce researcher bias. Large sample size is important for determining whether results are statistically significant. (10/n)

Since the Marseille studies, several of these prospective studies, including all the above-described bells and whistles have been done. They were done at different stages of illness and they analyzed the effects of HCQ both as a preventative measure (prophylactic)... (11/n)

...and as a treatment for existing infections. The definitive conclusion is that HCQ DOES NOT WORK for treatment of COVID-19. Because of this, the NIH even discontinued one such trial (nih.gov/news-events/ne…). (12/n)

Now, there are many prominent, well-respected scientists out there saying that HCQ is the key to treating COVID-19. I see how this can be confusing. (13/n)

One paper that is heavily quoted by some of these prominent scientists was published in the International Journal of Infectious Diseases (ijidonline.com/article/S1201-…). This paper is an observational, retrospective (i.e., not good) study. (14/n)

Basically, they gathered data from a bunch of hospitals from lots of patients who were treated in different ways. Some received HCQ, some received HCQ plus azithromycin, some received neither. (15/n)

However, being that the data were gathered retrospectively, they couldn’t control for the proper variables. They couldn’t control who was enrolled into which kind of treatment. The study was subject to all kinds of biases. (16/n)

In the conclusion, the authors more or less state that treatment with both HCQ and azithromycin had a modest effect on mortality in patients, but they admit that “prospective trials are needed to examine this impact”. (17/n)

Since then, multiple studies have emerged that show the opposite. A study in NEJM (nejm.org/doi/full/10.10…) shows that “the use of HCQ, alone or with azithromycin, did not improve clinical status at 15 days as compared with standard care.” (18/n)

A systematic review in the Annals of Internal Medicine (acpjournals.org/doi/10.7326/M2…) concludes that “evidence on the benefits and harms of using HCQ or chloroquine to treat COVID-19 is very weak and conflicting.” (19/n)

A recent study in Nature (nature.com/articles/s4158…)
shows that if you add HCQ to monkey cells and infect those cells with SARS-CoV-2, the drug inhibits viral replication. But if you do this with cells from the human airway epithelium (the REAL cells of interest)... (20/n)

...the drug doesn’t work. When they treated monkeys (the actual animals) with HCQ and infected them with SARS-CoV-2, the drug also did not protect the monkeys from infection. (21/n)

Another study in Nature (nature.com/articles/s4158…) shows pretty much the same thing (monkey cells vs human airway epithelium). But here’s the kicker: if you engineer the human cell protease needed for viral entry into those monkey cells... (22/n)

...(cells in which HCQ was previously shown to be effective), then HCQ has no effect against the virus.

Based on all of these data, it is now clear that the use of HCQ in treatment of COVID-19 provides no benefit and the FDA has since withdrawn authorization of use of HCQ. (23/

All the data floating around that allegedly show that HCQ works, are anecdotal – meaning that they are personal observations that were not based on the properly controlled experiments. (24/n)

A professor at Yale (Harvey Risch) wrote an unsubstantiated opinion piece where he bunched together a few cherry-picked studies to support the use of HCQ. Notably, none of these studies were controlled or randomized, but really, I shall say nothing else about this person. (25/n)

You can read a well-formulated blog post at Science-Based Medicine (sciencebasedmedicine.org/hydroxychloroq…), which summarizes all of the above very nicely and addresses all of Harvey Risch’s claims. I think I’m done here. (end)

In a *prospective study you can plan and control.... (Typo)