Hacker News new | past | comments | ask | show | jobs | submit login

Chloroquine and its derivatives were given to thousands, if not millions, of people for years. We know and understand the safety profile of this drug well. To think that all of a sudden it became more dangerous was silly and unreasonable.

This study always smelled bad. The media was so quick to champion it everywhere though. Why might that be I wonder? Where are the retractions now? The whole thing is disgusting.

The media in the US is doing everyone a disservice. Because of this less people are going to believe "science" and more people are going to just retreat to their echo chambers and believe whatever they want. I know I'm not going to trust the Lancet or the NEJM ever again. It's all political now.

What a mess.

> We know and understand the safety profile of this drug well. To think that all of a sudden it became more dangerous was silly and unreasonable.

COVID patients are different from the patients that normally receive chloroquine and the chloroquine doses used for COVID are also higher. There is also the issue of how chloroquine interacts with other drugs such as azythromycin.

This particular study has its problems but other studies on chloroquine were also starting to suggest that chloroquine could have dangerous side effects.


> Chloroquine and its derivatives were given to thousands, if not millions, of people for years. We know and understand the safety profile of this drug well. To think that all of a sudden it became more dangerous was silly and unreasonable.

Aspirin is one of the most widely used drugs and generally considered safe. But you should never give it to a kid with chickenpox, because doing so is a risk factor for developing Reye's Syndrome, a serious and potentially fatal disease.

It's not that aspirin "becomes more dangerous" in general, but it is more dangerous for those particular patients. And we don't really know why, either; but the data is clear on the correlation.

We don't yet know all the ways that the novel coronavirus makes people sick. Heck we don't even have a reliable inventory of possible symptoms yet.

So it's really not fair to pretend that it would be silly or unreasonable to suspect the possibility of some sort of negative correlation with a particular drug or class of drugs. This sort of thing happens all the time in medicine.

> To think that all of a sudden it became more dangerous was silly and unreasonable.

This is a dangerous view and not a fair take. Drugs CAN become more dangerous when used in a different context. The sheer permutation of prior/current medical conditions, interacting drugs, demographics, genetics, medical procedures, and a bunch of other factors all play into the equation that determines a drug's safety and efficacy profile. This is why we continue to do research. This is why guidelines are constantly shifting. This is why medicine requires years of study.

> The media was so quick to champion it everywhere though.

I definitely agree with this point. The media has caused harm to the population by taking the results of a single observational study, and parading it around like it was some new concrete medical certainty. This is not how professionals operate. We don't flip our practice on the whims of a single OBSERVATIONAL study.

And I emphasize OBSERVATIONAL because most people here don't understand what that entails. Most people here who comment on clinical trials are not even trained to interpret them. Is it randomized/non-randomized? Double-blinded? What did data collection look like? How were results analyzed? What was the patient population? Timelines? Control arm? Placebos? Previous findings? Primary/secondary endpoint? Do you know the difference between a meta-analysis and a systematic review? NNT? Hazards ratios? Odds ratios?

Too many people outside of medicine think they know how to interpret a study, when in reality they have no idea what they're looking at and cherry-pick the interesting sentences that they're looking for.

> I know I'm not going to trust the Lancet or the NEJM ever again.

This is not a fair take. The journals are responsible for reviewing and publishing the most influential clinical research in the world. Occasionally a bad study makes its way in due to falsified data or other illegitimate factors. Health care professionals are well-aware of this, which is why we are trained to interpret studies, and be conservative in the face of radical findings like this.

If anyone has a problem with the Lancet or NEJM, they should see some of the mess found in lower-impact journals.

I did not agree with the conclusions of the paper, but I fell prey of the same bias that many others had: it was "published in Lancet" and "those journals have strict peer review".

The lesson I learned is that I will always scrutinize papers I read with a more critical eye, regardless of where they were published in.

I think we need to understand exactly what "peer review" is required for publication or publication certain places. Is it:

a) a grammar/spelling check

b) a gut check to make sure it "makes sense"

c) validation of the math/analysis

d) validation of the underlying experimental procedure used to collect the data

e) validation of the underlying selection criteria for inputs/candidates in the experiment

f) reproducibility of the experiment and results

g) something else?

h) all of the above

My impression is that "peer review" generally includes a & b and sometimes c & d.

Sadly, some of us are aware of the mess found in lower impact journals. It doesn't help with the trust issue. At all.

> I know I'm not going to trust the Lancet or the NEJM ever again.

Wait, what??

The fact that the lancet was able to determine that there were validity concerns with the data and issue an expression of concern in less than 2 weeks is proof that the system works.

Science is never 100% certain. The media has never been a good source of scientific information and probably never will be.

edit: also, if you "don't trust the Lancet or NEJM" then you don't trust a huge number of the treatments available to you at the hospital. These are two of the most important and eminent medical journals in the world and many life-saving treatments began as articles in these journals.

It is proof that the peer review failed, on a topic that the Lancet couldn't ignore was high profile. I have no idea what the motivation of the reviewers were, but given how political this molecule has become, I can't help having a terrible doubt.

Peer review is never intended to be an independent audit of the data for any journal. Reviewers check the methodology, look for obvious mistakes or oversights, provide suggestions for clarification or improvement in the text and for high-demand journals, weigh in on their opinion of the importance of the study. All of the feedback is based on the paper itself, so outright fraud usually is not caught at this stage unless the malicious author is exceptionally sloppy.

It is later when other scientists start working to replicate or expand on the study that they request access to the data, and dig into the details. This is when most instances of scientific misconduct are found, and this paper is no exception.

In science, publication is the start of the peer review process, not the end.

HCQ and its derivatives were given not to thousands or millions but tens of millions sometimes for years on end. In tropical countries like Nigeria it is customary for visitors and residents alike to be on long term HCQ dosages.

I was quite surprised at the newfound fatality. Given that it was the Lancet, I took it at face value. Their political biases made them overlook the sheer flimsiness of the study.

Rather than a mess, it is showing us the messy humans behind the veil of "science". There is no platonic scientist - just humans doing imperfect science.

We don't know as much about the safety profile of the drug in patients with COVID-19.

Aspirin has a long track record and well understood safety profile, but it can still cause life threatening complications like Reye syndrome in certain viral infections.

Philip Greenspun had some good articles on this. See below.

He also quotes the CDC website saying it is quite safe.

> “CDC has no limits on the use of hydroxychloroquine for the prevention of malaria. When hydroxychloroquine is used at higher doses for many years, a rare eye condition called retinopathy has occurred. People who take hydroxychloroquine for more than five years should get regular eye exams.”



What makes it even worse is that social media companies are now so paranoid about being labelled spreaders of misinformation, they are relying on sources who could easily be wrong about something as well (like in this case).

I guess this way at least they can point the responsibility somewhere else, and say it's not their fault.

Some media conglomerates, like CNN, were clearly using this study as a political attack. Not uncommon these days to cherry pick your favourite paper.

On top of that, there is no patent money to be made by big pharmaceuticals on this drug. It makes me wonder where the funding really came from in this Lancet study.

A lot of people on this forum have been aware with the issues in social science and medical research for a long time, it's only under real stress that it's become an issue that will get wide coverage.

The next fiasco-in-the-making to watch is the 'evidence-based' criminal justice reform sure to come to a lot more American cities soon.

> was silly and unreasonable / This study always smelled bad / The whole thing is disgusting / "science"[0] / I know I'm not going to trust the Lancet or the NEJM ever again / What a mess

You're throwing around an awful lot of poo. Nothing concrete at all but all in very emotive terms. Not constructive - but was that the point?

[0] (note scare quotes)

The really odd thing is the media were highlighting risks... but not giving them context. Aspirin is risky, but millions take it too, over the counter.

I’m all for being skeptical, but I think they went overboard, especially given the alternative of having no medicine.

Every medication has a side effects and you balance if the benefits are better than the side effects it causes. Chloroquine has tons of side effects[1] though (including things like causing trouble with breathing). It might still be worth it if you have Lupus.

When an existing drug is intended to treat some new disease it needs to be reevaluated for that disease again, because even if it can save lives for one disease it can be deadly for another.

So the fact that it is given to thousands or millions does not mean much in this context.

[1] https://www.drugs.com/sfx/chloroquine-side-effects.html

>> Because of this less people are going to believe "science" and more people are going to just retreat to their echo chambers and believe whatever they want.

That's true. There are people who want to discredit science as a whole. Reality tends to get in the way of most hidden agendas - that's why they're hidden.

>Why might that be I wonder?

Because our president was idiotically and irresponsibly championing a drug for which there was no good evidence of efficacy at all.

"I know I'm not going to trust the Lancet or the NEJM ever again. It's all political now."

This is an absurd reaction and I suspect you know it. There isn't a single scientific journal in the world not subject to some bias, trend, or failure of one form or another. To dismiss an entire platform because of an imperfect record amounts to an abdication of your own responsibility to form your own opinions about scientific matters. This is a responsibility for which journal reviewers can only ever take partial responsibility for both practical and intrinsic reasons.

Peer review is not a rubber stamp which blesses anything which passes through it as "TRUTH." It is a minimal standard and subject, at any rate, to all the uncertainties and biases implicit in any human endeavor. It is entirely possible this paper represents a political bias. Almost everything does. It still falls to us to make reasonable judgments about science. What doesn't make sense is to reject the entire process because it fails to meet some unreachable expectation of perfection.

And the Lancet and the NEJM have also been willing to publish lots of studies that don’t necessarily go with their politics. It’s one of the few places you can read about the facts of the healthcare system, not just the political narratives. So it seems unfair to take the anger at the media out on these journals specifically.

> lots of studies that don’t necessarily go with their politics.

Can you provide examples? I'm sure it'll be hard, if not impossible, to provide an example that's as emotionally satisfying as directly contradicting Trump, but I'd love to see which studies you're thinking of.

Random example: hospital mergers don’t have a negative impact on patient outcomes: https://www.nejm.org/doi/full/10.1056/NEJMsa1901383

Somehow people are able to turn Trump's sheer ridiculousness and lack of credibility into a positive - they can say "you're just irrationally attacking Trump" and try to drop things there, instead of having to make the argument that Trump is right.

Lancet is different than other scientific journals - it is the pinnacle. Publishing something so bad is reputationally harmful in ways that might not be so severe in other cases. Of course, this isn't the first incident Lancet has had (recalling the Wakefield incident in particular), but it should not be excused - just as Wakefield shouldn't - and should be judged rather harshly.

Edit: As OP mentioned, this is the "wakefield incident"

Don't forget, The Lancet is also the journal which published the paper that kicked off the anti-vax movement: https://en.wikipedia.org/wiki/Lancet_MMR_autism_fraud

This is what GP means by "Wakefield".

That is the Wakefield incident he was referring to.

It also published [0] which seemed to be the ultimate source for the 'ibuprofen kills COVID-19 patients' scare, every article I found on ibuprofen+coronavirus led back there, which matter-of-factly states:

> ACE2 can also be increased by thiazolidinediones and ibuprofen.

The closest thing I could find [2] to a source for that was [1]:

> NSAIDs [Non-Steroidal Anti-inflammatory Drugs, which includes ibuprofen] might affect how Covid-19 binds to human cells, according to Dr. Yogen Kanthi, assistant professor of cardiology at the University of Michigan, who studies inflammation.

> “There is data from basic science studies that have shown that Covid-19 itself binds to a protein at the surface of cells called ACE2,” he said. “There is a hypothetical risk that giving NSAIDs like ibuprofen could increase levels of ACE2 shown in animal models, but not in patients.”

[0] - https://www.thelancet.com/pdfs/journals/lanres/PIIS2213-2600...

[1] - https://kvia.com/health/2020/03/17/france-says-ibuprofen-may...

[2] - https://news.ycombinator.com/item?id=22611363

TIL about the Wakefield incident. TY.

> This is an absurd reaction and I suspect you know it.

Why? They knew the data was fake when they published it, but they published it anyway to boost sales of Remdesivir.

They're literally trying to make money by killing people. If you wouldn't trust someone if you saw them run over someone else with their car, why would you trust Lancet?

It's much easier to throw your arms up and say "Conspiracy" than actually assess things on a case by case basis.

Peer Review gone wrong -> https://www.youtube.com/watch?v=wLlA1w4OZWQ

On the other hand, we've seen so many cases of people more interested in contradicting Trump than being logical about their positions, that I don't think such theories qualify for "conspiracy" territory. ...just emotionally charged bad decisions.

> ...of people more interested in contradicting Trump than being logical...

Precisely the way trump himself behaves, but you don't point that out.

...because his stupidity isn't relevant to the point I'm making.

You illustrate my point by assuming that I'm supporting him, just because I'm not criticizing him.

Valid point. Upvoted.

> This is a responsibility for which journal reviewers can only ever take partial responsibility for both practical and intrinsic reasons.

So what you're saying is we should never fully trust these journals? We should always be skeptical?

That's what I'm saying too...

If you're healthcare, you're taught not to trust results because it was published under a brand-name journal.

YES, you should be skeptical, but mostly not at the journal-level. You need to be skeptical at the article-level. That is why it's so important to be actually TRAINED to interpret the studies.

The abstract of a trial is like an "advertisement" for the study. You quickly scan it to see if the study is worth reading. If it is, you make multiple passes of the article, identifying biases, understanding the study context, calculating ratios and numbers, reading through the lens of your own practice, and a bunch of other things.

Healthcare pros are minimally trained in research paper interpretation and are almost all unable to perform the most basic statistical or critical review work.

So yes, I'd argue that brand names are a big problem. You just have to see how proud people are when they are accepted in one of the major publication venues, and the prestige that results.

I think you are not being objective.

Except that it on the sole basis of this publication that the WHO and many other health organisations suspended ongoing trials of this molecule. So let's not pretend they would have done the same based on some random obscure journal.

i am in the process of compiling a list of ways to quickly spot flawed studies. do you mind sharing your best tips?

There is a course on this (taught Autumn 2019 at U Wash) that will also soon be published as a book:


Andrew Gelman put it very well:

"No, the real scandal is that the respected medical journal Lancet aids and abets in poor research practices by serving as a kind of shield for the authors of a questionable paper, by acting as if secret pre-publication review has more validity than open post-publication review." [1]

[1] https://statmodeling.stat.columbia.edu/2020/06/01/this-ones-...

No, you said you wouldn’t trust it at all and that it was 100% political.

HN isn’t the ideal venue for hyperbole.

But you argued as if you were always trustworthy of these Journals before but from now on, you plan to be skeptical. Your future intentions are similar but for the wrong reasons.

One should always be skeptical of research and have dicsussions on the reported data and to what extent and in what context it makes sense. The media outlets summarizing the research papers very often fail to do that so as to skip the details and keep the readers happy because they assume "readers just want to know the superficial abstract knowledge, they can't possibly be interested in the critical analysis of the figures and the numbers so let's exclude that".

But in this case the media is acting very dishonestly. I remember every major outlet posting articles about how now HCQ was debunked by this study. I have yet to see a single retraction of those articles or any new articles mentioning this. Only the science publications are covering this part of it so far.

But why should the actions of the media change your perspective on the trustworthiness of the journals?

Journals are, after all, a filter for high-correlation results using small data sets. By definition that includes every coincidental result out there. You can probably create a distribution of accuracy with enough historical data.

> So what you're saying is we should never fully trust these journals? We should always be skeptical?

I think that exactly how good scientists proceed. This isn't my field, but my understanding is that research groups don't "fully trust" peer reviewed research, so they usually attempt to replicate the key results they plan to cite in their new research.

I agree, but if you do this you are usually labeled a ‘denier’ of some sort in regards to the topic at hand.

No one in science places 100% blind trust in any paper, even if it is published in a prestigious journal.

In fact, scientists often say, somewhat tongue-in-cheek, "It's published in Nature, so it's probably wrong" (the more revolutionary a result, the more likely it is to be wrong, but also the more likely it is to get into one of the top journals).

Repeating what I've said in the past, anyone that has worked in scientific research for at least five years probably has a list of high-profile papers in their field that never should have made it past peer review. There's definitely at least one Nature paper in my personal hall of fame, and I'm still pissed off about it after six years.

Do you think this study would have been published if Trump had loudly proclaimed that Hydroxychlorine was dangerous and should be avoided?

I still believe it would have. The information regarding COVID-19 has been evolving since the beginning, 24/7 which has lead to the "truth" changing repeatedly. Also, Trump has huge support from his business buddies, including Big Pharma so I would have to assume they would have either protected him or they would have warned him.

If your implication is that the media and the science community conspired to make the president look stupid, I think you've got a big hill to climb to prove that.

No conspiracy is necessary. There is a mass emotional desire to contradict anything Trump says. Agreeing with something Trump says has become very taboo, particularly in academic social circles.

...and I think it's possible that's what we see here. An emotional decision to publish, and then no one wanted to contradict because no one wanted to be viewed as a Trump sympathizer.

It is an abdication of responsibility from Lancet / NEJM editors. Everyone not living under a rock is well aware hydroxychloroquine has, unfortunately, become a political hot potato. In this context, a paper purporting results counter to decades of clinical practice should warrant a modicum of additional scrutiny.

This incident comes in the middle of the replication crisis, with 50%-70% of studies failing replication, including in the medical sciences field. Furthermore, many suspect that BigPharma pushes inconclusive garbage as peer reviewed science to promote commercial interests. Then there is the response to covid19, which, justified or not, some regard as catastrophic overreaction. The BigScience enterprise is in peril of losing public trust. Which is a terrible loss.

Lancet / NEJM editors failed. There are no excuses. The right course of action is to apologize, not to further antagonize the public. Try humility for a while.

> I know I'm not going to trust the Lancet or the NEJM ever again. It's all political now.

Who can you trust these days?

This may have been asked rhetorically, but I will answer it seriously.

First, there are some really good science reporters out there, doing excellent work. I recommend Ed Yong at the Atlantic, Helen Branswell of STAT News, Jon Cohen at Science, Amy Maxmen at Nature. There are others (it's a team effort and those are all great publications), but if you just read all of the articles under their bylines, you will get a good overview of what's going on. They will not let you down.

If you have a little bit more time and energy for this stuff (as I do), then there are many experts on Twitter who freely share their analysis, and among them are some excellent communicators. When controversial stuff comes out, you can get a good sense very quickly what the flaws are and what seems to be the consensus. If you want to dip your toes into these waters, I recommend Angela Rasmussen, Trevor Bedford, and Marc Lipsitch. For critical analysis of bullshit published by seemingly reputable academics, I highly recommend Carl Bergstrom (his takedowns are highly entertaining as well as informative).

It's frustrating for me to see such pearls of knowledge and expertise so freely available, yet the vast majority of people snarfing down huge quantities of information swill.

Named editors. Peer review is recent, and has mostly proved itself a bad idea. Editors who have a personal reputation to maintain will do a better job, just as they did back in the day when, you know, guys like Einstein walked the earth.

Personally I do not think of trust as binary. If I ruled out every journal, publisher, or news property that published something that turned out to be wrong, I would read nothing at all and wallow in blissful ignorance. Actually that sounds kind of nice right now...

Couldn't agree more. People lament the attack on "experts", but when they are transparently lying to the people [0], you can forgive the breakdown of trust.

[0] human-to-human transmission doesn't happen; masks are ineffective; HCQ is horribly risky; a vaccine is coming;

It was interesting to see Facui so quickly endorse remdesivir but perhaps he did so because it's safer than choloroquine. It is convienent that remdesivir costs $1000 to 4000 per patient, whereas choloroquine is basically free compared to that.

And this: "Before the Covid outbreak, remdesivir was a drug for which Gilead had failed to find a use." (Barrons 6/2/2020)

> Why might that be I wonder? Where are the retractions now? The whole thing is disgusting.

It’s going to take a long time to recover from what the media has done here. HCQ isn’t even allowed as a treatment in a lot of places, or require hospitalization (funny for a drug on the market for 60 years and over the counter in some places).

Other countries are ramping up production and we’re still dealing with the yet another casualty of culture war.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact