Hydroxychloroquine: What is going on?

You have likely heard of hydroxychloroquine, the anti-malarial drug which some people have suggested is useful in the prevention and treatment of COVID-19. President Trump has said he has taken the drug, despite limited evidence of effectiveness.

Within the last few days, a number of new pieces of information have come out. Let’s put them into context.

What is hydroxychloroquine?

Hydroxychloroquine (and the closely related chloroquine) are widely used anti-malarial medications. They are effective at both preventing and treating malaria, and are also used to treat some autoimmune diseases (e.g. arthritis, lupus). These drugs are cheap and widely available. This is part of what makes them appealing as a treatment.

Why would we think hydroxychloroquine could be effective against COVID-19?

There is some old virology evidence suggesting that these drugs can kill cells of SARS-CoV (the original SARS), which is closely related to SARS-CoV-2 (the virus that causes COVID-19). Similar cell-level effects were shown early on in SARS-CoV-2 in very small studies. In addition, there were some early claims made in France that use of hydroxychloroquine in COVID-19 patients improved outcomes.

What would you use it for?

This is where it gets a bit confusing. There are two uses for antimalarials such as hydroxychloroquine that have been suggested. One is treatment. The other is a form of post-exposure prophylaxis — basically, preventing infection in people who have been exposed to the virus. The President seems to be using the drug for the second purpose.

There is evidence on both uses, and it is worth separating the two.

Is this an Effective Treatment?

No. Or, at least the evidence we have seen so far does not support this. The best evidence on this comes from France and China, where data shows treatment with this drug does not reduce serious illness among sick patients. The data from China is a randomized controlled trial, the “gold standard” for evidence.

In general, there is really no evidence that this works for treating COVID-19. (If you’re curious about COVID-19 treatment, our explainer is here). (Is it dangerous or just neutral? See below).

Is this an Effective Prevention Measure?

Does hydroxychloroquine work to prevent COVID-19 development, as the President’s usage would suggest? Again, the best available evidence says no. There is new data on this from a large randomized trial, just released in the last few days. This study randomly divided about 800 people, nearly all with high risk exposure, into two groups: a group that received treatment with hydroxychloroquine and a group that received a placebo. The question was whether the treated group was less likely to develop disease. The answer was no. The study found that 11.8% of those who were treated developed the disease, versus 14.3% of those untreated. This was not a statistically significant difference.

There is some much less rigorous evidence suggesting a protective effect of hydroxychloroquine. This case-control study in India compares disease development among health care workers who took hydroxychloroquine versus those who did not. They find less disease development among those who took it. But the choice to take the drug is not assigned randomly here, and this case-control method has a lot of very serious issues. Given the availability of randomized trial evidence, this should probably be dismissed or seriously downweighted.

One thing to be very clear about: Even people who suggest that hydroxychloroquine should be used for prevention are suggesting it may be used for people with exposure. No one (serious) is suggesting that this be taken by everyone all the time. If you want to avoid COVID-19 your best bet is wearing a mask, washing your hands, and social distancing (see our explainer about avoiding infection here!)

Is it Dangerous?

Okay, so this doesn’t work. But is it dangerous? Some people have been touting use of this by basically saying “What’s the harm? Maybe it’s not really effective but also probably it doesn’t hurt.” But…does it?

The first thing to know: these drugs have known side effects. They are not a sugar pill or a vitamin. Consistent with this, randomized trial data suggest more adverse effects among those taking the drug than placebo. In the new randomized trial on prevention among exposed people, 40% of those taking hydroxychloroquine reported adverse effects, versus only 16% of the placebo group. None of these side effects were serious, but this study is consistent with what we generally know about the drug and its potential side effects.

Much more concerning was a paper in the Lancet which suggested that patients treated with hydroxychloroquine were more likely to have serious complications and were more likely to die. This paper used a very large sample of people from a registry of cases and the effects were quite drastic. It led to widespread concern about use of this drug because the mortality differences for treated and untreated patients were large.

But, just in the last couple of days this paper (and another paper with the same data) were retracted by the authors. The retraction seems to be a result of lack of data availability. The journal required that independent peer reviewers be able to access the data. One of the authors worked for the company which produced the data. However, they claimed for confidentiality reasons they were unable to provide the data. Since the journal required this, they chose to retract the paper.

It is difficult to really understand what is going on here, and whether there is more to the story than just inability to provide the data. We’ll need to wait for more information (which may actually never come) to understand this better.

The Bottom Line

Current evidence suggests that hydroxychloroquine is not effective at treating or preventing COVID-19. It is possible that will change, of course, but the evidence is increasingly persuasive.

On the flip side, these drugs have significant (known) side effects. It is not clear that they have additional risks for COVID-19 patients. More will need to be done to determine that.