Hacker News new | past | comments | ask | show | jobs | submit login
WHO launches global megatrial of the four most promising coronavirus treatments (sciencemag.org)
452 points by MichaelMoser123 7 days ago | hide | past | web | favorite | 161 comments





I am missing Fujifilm's Avigan/Favipiravir here - Chinese reported success using it:

"On 17 March 2020, Chinese officials suggested that Favipiravir seemed to be effective in treating COVID-19 in Wuhan and Shenzhen.[31][32][33]

A study on 80 patients comparing it to lopinavir/ritonavir found that it significantly reduced viral clearance time to 4 days, compared to 11 for the control group, and that 91.43% of patients had improved CT scans with few side effects."

https://en.wikipedia.org/wiki/Favipiravir


The four drugs they are testing seem to be already developed/approved in humans for other uses so expanding the label is easy, while avigan is still experimental (some limited approval in Japan, but worldwide acceptance not there yet) and would have to go through much more testing before being unleashed at coronavirus scale.

Also not produced at any sort of scale yet, which would limit its ability to be deployed quickly. These others are already in production and most of them cheap and common.

Surely if it's already being used in Japan it's been tested enough to use globally in an emergency.

It's very limited approval, it's not something that's just approved and available to be prescribed to anyone. According to the wiki it's only approved in an emergency (presumably meaning individual, not societal, emergency) because of concerns with toxicity and alll. And countries don't automatically trust other countries to approve drugs correctly - Europe did not buy what the FDA was saying about opioids for chronic pain which saved a lot of lives on Europe. The damage that would come from releasing a half baked drug into the world with terrible side effects would vastly outweigh the benefits of mitigating coronavirus, and that's a mistake you only get to make once.

I'm sure they'll test the drug in the future though, there are likely plans to test it as we speak, and if it works really well there are lots of ways to get expedited approval, especially in a state of emergency

https://en.m.wikipedia.org/wiki/Favipiravir


One guess: it's not a drug which they have the ability to crank up the production on quickly. Some drugs are easier to make than others.

Stage of disease also is important. In the first symptomatic stage, makes sense to try antivirals and HCQ to reduce or eliminate the virus, and prevent a cascade into the next stage. The problematic second stage - ARDS, a subset of patients who really go critical, seems to call for different approach because at that time, it’s the cytokine storm and our own immune reaction killing us.

Cytokine storm is generally observed with young patients. Their immune system is "too strong for their own good".

My understanding is that most deaths are in the elderly, and that those deaths are due to pneumonia.


source on cytokine storm age differences? all resources I can find indicate that it's more prevalent in older patients.

Either of you have links on this?

That's my understanding too, younger ones get it harder.

If this helps https://en.wikipedia.org/wiki/Cytokine_storm

"It is believed that cytokine storms were responsible for the disproportionate number of healthy young adult deaths during the 1918 influenza pandemic, which killed 50 to 100 million people.[16] In this case, a healthy immune system may have been a liability rather than an asset."


Do you have a source for the last part?

A good synopsis of current understanding (as of last Friday) was a UCSF Medicine Grand Rounds broadcast. I've bookmarked the most relevant slide to our discussion in particular: https://youtu.be/bt-BzEve46Y?t=2349

Significant lung and myocardial injury and papers have honed in on ARDS as a real problem.


This is a wonderful resource. Thank you for pointing it out.

Not OP, but here's a Lancet correspondence on the topic (includes references):

https://www.thelancet.com/journals/lancet/article/PIIS0140-6...


Last night I watched a YouTube video about COVID-19 symptoms and stages of the disease and it did touch on the points brought up by OP.

Could you share the url ?

Not GP but this video gives a good overview: https://www.youtube.com/watch?v=BtN-goy9VOY

Maybe OP is referring to this one?

https://www.youtube.com/watch?v=BtN-goy9VOY


This is specifically the video I watched yesterday.

https://www.youtube.com/watch?v=OOJqHPfG7pA


oh I should have guessed :)

if you want some interesting details from American MDs there's this podcast

http://www.microbe.tv/twiv/twiv-593/

lots of subtle details that you can't get on simplified stats


This doesn't need a source. It's common clinical knowledge and OP is right. Unfortunately, preliminary results using our usual immunomodulatory drugs have been catastrophic.

No, sorry, you still need a source. In fact cytokine storms aren't the only driver of pneumonia and we don't know yet exactly what the mechanism(s) behind ARDS is. Immunology is outrageously complicated, and not well suited to that kind of pronouncement.

You just have to do the science, there's no way around it. And realistically we may not have time.


I will reformulate, then.

In the current state of knowledge about both ARDS and Covid and given the time available, clinicians rely on traditional teachings to care for covid-related ARDS until solid evidence can be provided.


[flagged]


Regardless of whether the parent comment is wrong, you'd be much better received if you started your reply less dismissively.

Hence my comment, perhaps positioned less than optimally in my statement, about preventing the cascade to next stage. The degree to which these drugs address the various issues is still in flux. In the UCSF grand rounds recently, there was some question as to the role of immunomodulators.

The drugs on trial are:

1. Remdesivir (Ebola drug).

2. Chroloquine and Hydroxychroloquine (Malaria drug).

3. Ritonavir/lopinavir (AIDS drug).

4. Ritonavir/lopinavir + interferon beta (virus signaling protein).


Interesting to see interferon on here. I remember it being hailed as some sort of wonder drug back in the 80s, but then sort of fell off the radar.

https://books.google.com/ngrams/graph?content=interferon


Covid disables the innate immune response by preventing the expression of interferon alpha & beta.

This allows it to engage in uncontrolled replication for up to two weeks before the adaptive response brings it under control.

This massive viral load greatly contributes to transmission.

(Or at least SARS does this.)

https://cmr.asm.org/content/20/4/660.long


The side effects are awful.

I really thought Camostat mesilate was going to be looked at

I think it is only approved for treatment in Japan

But it has been around for at least 20 years. The safety profile is extremely well known.

I don't even want to know what role lobbyists play in selecting the drugs used for the trial.

Clearly if they played a big role they did a terrible job with how many of these drugs are off patent.

Probably not a lot. Most of these are out of patent.

I guarantee you that Gilead has at least considered filing a patent application on a new use of the old compound. And they will probably get a grant. Same for all the other treatments.

Can anyone comment about the worldwide production capacity and possible lower price limits for these drugs?

It should be possible to supply a large part of the worlds population with (hydroxy)chloroquinine if that is deemed a solution, but the others seem less ideal candidates for that, or am I wrong?


What I can personally comment on is that if people go on with hoarding on quinine derivatives, we're gonna have a massive problem whether it's found effective or not.

The hoarding is even happening among my lesser scientifically-reasonable md colleagues. This has to stop. Now.


Hoarding is only a part of the problem. A company in The Netherlands producing chloroquine has got 24/7 police protection as of this weekend because some people came to them and were very insistent that they would like to have some product.

Link in dutch: https://eenvandaag.avrotros.nl/item/fabrikant-van-coronamedi...


Norwegian doctors have also been hoarding this for themselves lately. Enough that the Norwegian government had to step in and put it on a list so those that actually need it day-to-day can get hold of it.

I hope when this is over, someone will look into this behaviour from the MDs. It's eroding trust in the system when they claim potentially vital resources for themselves.


I don't know what the situation is like in Norway, but I 100% don't blame MDs in the US for trying to find their own treatments, because they're generally being left high and dry by the extreme rationing of the PPE. Doctors themselves have taken to personally reaching out to the communities near me so that they can find masks.

Obviously I don't want anyone with lupus / malaria / etc to be adversely affected, but I think the people being forced to work the front lines do deserve to be somehow prioritized just from a "how do we keep the system operating efficiently" standpoint, and it doesn't really seem like the people who could manage supplies to make sure both parties needs are balanced, e.g. hospital administrations and regional governments, are going to take much responsibility for that.


Note that given the dismissive behaviour of Occidental medical leaders regarding covid, I wonder what's left of trust in the system.

Our bosses were still laughing at us for worrying on Monday evening, and now they're pissing their pants. Idiots.


Not to mention have a government enforced monopoly on it.

When this is over, what we absolutely have to look into is the generational conflict.

<50yo: all MDs understand and have been vocal about it for weeks.

>50yo: nah, this can't happen to us great minds. Let's start preparing when the first patients are already in the ICU.


Would drinking tonic water help?

Modern tonic water contains substantially less quinine than the tonic waters that were used to fight, say, cholera in the historical period.

So, yes, if we lived in India in the mid-to-late 1800s, drinking tonic water might help, at least for the things that quinine helps.

But modern tonic water isn't really anything like that.

https://en.wikipedia.org/wiki/Tonic_water


Before the virus, the global supply of Remdesivir was only enough to treat a few hundred people. Chloroquine and Hydroxychloroquine are apparently very easy to make, and can be scaled up very quickly. Not sure about the last two.

Regardless, WHO should be doing trials on more over-the-counter products. It's dumb that they put out a statement warning people that there was no evidence that Garlic could be used as an antiviral, despite the fact that literally two days ago the FDA granted emergency expanded access for some multi-million dollar inhaled nitric oxide device based on the fact that nitric oxide has been shown to block the replication of SARS-CoV in vitro. (Guess the mechanism of action by which eating raw garlic nearly instantly lowers your blood pressure.)

Sources here: https://www.reddit.com/r/covid19stack/


You're not telling the full story. That approval was for compassionate use, i.e. it can be used in patients who are already in grave danger of death, because at that point it can't hurt and there's nothing known to have a chance of helping anyway.

And the article you're referencing chooses to link press releases, but only vaguely mention studies not only without linking but without even providing enough information to identify them - which hardly inspires confidence.

And the subreddit from which you're getting it is...wow. Just an absolute heap of fear-driven misinformation, and while I'm not without compassion for the people there who feel the urge to do something to protect themselves in the face of a very frightening situation, it's important to try to keep a clear enough head to understand the difference between developing knowledge and multiplying tiny probabilities into meaningless, misleading rumors. That subreddit is doing the second one of those, not the first. For the sake of your own emotional wellbeing, please consider taking it less seriously.

edit: Oh, sorry, I hadn't noticed that the people posting in that subreddit are basically 90% you. So, I guess I'm warning everybody else off it, and asking you to consider doing less of this, in the cause of everyone else's emotional wellbeing.


> Just an absolute heap of fear-driven misinformation

The vast majority of links are just to metastudies and literature reviews. If you think the Cochrane review studies or whatever are fear-driven misinformation then you should complain to them for creating them, not me for posting them.

Literally today Trump put out a statement saying that he's considering just sending everyone back to work and letting everyone get it all at once.[1] If that happens then the medical system basically won't even exist for 90% of people. What exactly do you propose that people should do in that scenario, just sit around drinking Gatorade and waiting to die?

[1] https://www.nytimes.com/2020/03/23/us/politics/trump-coronav...


Anyone knows why serum is not sufficient to completely solve all our issues? I applaud this trial but this confuses me. It seems like we already have the capacity to do this on large scale, it's considered very safe and as far as I understand is highly effective. There was an article about it here a few days back, but besides that I don't hear much about it in the general discussion, why?

It can’t be done effectively during exponential growth at the current rate: there simply aren’t enough identified recovered patients to supply serum to the currently sick patients. If the exponential rate decreases a bit, maybe.

FWIW, it seems at least plausible to me that a recovered patient who was treated with serum may not generate as many protective antibodies as a naturally recovered patient. For example, Rh antibodies are used during pregnancy specifically to prevent an immune response. I don’t know if this would apply to COVID-19 or if giving serum only after symptoms become severe would prevent this outcome.


See my reply to your original comment. Exponential growth isn’t a valid reason for this to not work. The factor of recovered vs currently Ill patients is constant, exponential growth or not. This factor may be high and I assume this is what you mean but given the (relatively small) percentage of critically ill and the fact that some countries are further along this is not a clear disqualifying argument against serum.

Not a doctor, but AFAIK its not considered safe enough. There are quite a few possible side-effects from using some of the mentioned medication, ranging from diarrhea to QT interval changes that could induce cardiac arrest.

One thing I couldn't find a source for is how many donors could potentially help how many people. Is it 1 to 1? 1 donor can help multiple people? Or one patient needs the amount of multiple donors?

The article I read on serum suggested that one donor could help two people. No URL, I’m afraid.

That's exactly it. If the virus is exploding exponentially the number of recovered cases is a tiny, tiny fraction of the active ones. This treatment doesn't work until things are under control, basically.

But if the death rate is "only" 1 in 10 (approximately) then doesn't that mean there are ~9 survivors for every terminally ill patient? It doesn't seem like the number of survivors would be the limiting factor here.

No, because the virus takes time. Survivors aren't "survivors" until they've recovered, so they reflect the people who "would have been" critical cases maybe 10-15 days ago. But 10-15 days ago the number of cases was (depending on which doubling statistic you use) 4x-32x smaller than what it is today.

Exponential growth messes everything up, basically. All our intuition about how things work tends to be wrong in subtle ways.


e^(ct)/e^(c(t - 14days)) = e^(c*14days) = k = constant, with c the growth rate. While you are right that, in an exponential growth scenario, the number of recovered cases is smaller than the number of currently active cases, the factor is proportional to the growth rate which is worst case constant, but likely going down over time due to quarantine measures. As such we have theoretically enough recovered cases if n_infected/n_critical > k. For the most critical age group an estimated 10% of infected require ICU care [1]. Thus we need k < 10. With a 1:2 ratio for serum we get k < 20. Likely not everyone will donate though and constant factors in this equation will end up mattering but the exponential nature does not kill the idea in the manner you indicate. However, it still requires scaling production exponentially. This is true for all other drugs as well though.

[1] https://www.imperial.ac.uk/media/imperial-college/medicine/s...


You don’t need 80-year-old serum to treat an 80-year-old patient.

I think trials for this are beginning - I suspect it would have been a hard thing to include in the WHO trial, as a hospital needs to be fairly sophisticated to be able to coordinate plasma retrieval and cleaning locally (vs buying RBCs from a local supplier).

I’m curious /excited to rare how this dynamic changes once we get serological tests online. Maybe all of the mild cases amongst youth can give enough serum to treat the harder hit individuals.

Here’s a recent discussion about serum:

https://www.globalhealthnow.org/2020-03/covid-19s-stop-gap-s...

And the linked study:

https://www.jci.org/articles/view/138003


1. Too expensive 2. Safety unknown at large scale 3. Production facilities insufficient

And many more reasons...


Do you have a source for this? Interesting if true!

1. Where are the high costs coming from? Distribution? Storage? People? Most of these seem shared for other drugs.

2. Blood transfusion seems to be one of the most common treatments there is (https://en.wikipedia.org/wiki/Blood_transfusion#Frequency_of...). Safety concerns seem to stem from so far unknown pathogens that are not tested for in the blood of donors.

3. Production facilities are every hospital. Nothing needs to be produced or what do you refer to?


I'm referring to my own clinical and scientific perception, which seems to be shared among my peers, but I didn't run a comparative study.

Right now, chloroquine is cheap. I suspect that will change. As for plasma yes, every step from collection to administration is expensive if you want to keep with the current safety level. And other people need plasma too...

Plasma toxicity worries currently comes more from transfusion-related reactions than from infection.

Hospitals have very limited capacity in term of production. Just starting with the fact that you'd have to find enough volunteers in the first place, but even so a hospital is not a factory.


Dr. Peter Attia did a few podcasts with a virologist discussing this.

It’s helpful in the short term, but not effective for longer than a few weeks apparently. Also there may be scaling issues?


Was it this one? https://www.globalhealthnow.org/2020-03/covid-19s-stop-gap-s...

> It’s not a vaccine. Think about it as the administration of a protein, it’s a liquid that is given to people that gives them immunity.

> Right. Because the vaccine would provoke the recipient’s antibodies. You'll have the antibodies, but they won't be your antibodies—though it'll do the same thing.


Yes. What do you want to say with the quotes from the article? Yes it is not a vaccine, neither are antiviral treatments.

The antiviral treatments undergoing trial are (for the most part) very inexpensive drugs that have been known for years or even decades so they don't need additional testing once they're proven to work and can be manufactured at scale almost immediately (since the production chains already exist).

This is only true of the chloroquines, and those are antiparasitics, not antivirals.

The difference of serum to vaccination is that any resistence/immunity gained from serum will only last until the antibodies are eliminated from your blood.

A vaccine triggers an active immune response. That includes the differentiation of aptly-called "memory" B cells. Those can lie dormant for decades and spring into action when needed.


The comparison here is not vaccine vs serum. It’s antiviral vs serum.

Its interesting to note they did not mention antibiotics as something that was being tested. I know for a fact that a major bay area hospital is treating covid-19 patients with a course of antibiotics(unsure which type), which is strange as its a virus. The nurse I know said they are seeing good results, the antibiotics are administered right away(they didn't say if anything else is administered) and 5 of 6 patients are being sent home with none coming back with worse symptoms. The sample size is small, maybe 30-40 and I wonder if these patients are already healthy enough that the virus is not affecting them greatly. It looks like other countries are looking into it as well:

https://techcrunch.com/2020/03/19/french-study-finds-anti-ma...


It is almost certainly azithromycin. Interestingly, it's antiviral effects have be previously studied. However, people aren't really happy with the lack of error bars on that (French Study) chart or the small sample size. Studies are double blind for a reason.

I would recommend this post: https://blogs.sciencemag.org/pipeline/archives/2020/03/19/co...

... and most anything that Derek Lowe has to say about small molecule drugs. He's apparently had a bit of time at home recently to blog more, and is rather focused currently on Corona virus.


One theory I've heard is that it is not the virus itself that kills you, but your immune system overreaction causing secondary damage which leads to infections. So the antibiotic might be aimed at that part? Just a guess.

If that is true (it is theorised that was how the 1918 flu killed people) it would primarily kill people with strong immune systems, not people with weak immune systems. I saw an epidemiology specialist (Michael T. osterholm maybe?) say that luckily this isn't how COVID-19 behaves or we would see millions dead.


Would drinking a lot of alcohol have the same effect? Temporary immune system suppression

They are probably put on antibiotics when they are intubated, to prevent:

Ventilator-associated pneumonia (VAP) is a type of lung infection that occurs in people who are on mechanical ventilation breathing machines in hospitals. ... Between 8 and 28% of patients receiving mechanical ventilation are affected by VAP

https://en.wikipedia.org/wiki/Ventilator-associated_pneumoni...


this information might become relevant if anybody dear to you is hospitalized with Corona virus related symptoms: The WHO is running a randomized trial to check the effect of existing medications:

"Enrolling subjects in SOLIDARITY will be easy. When a person with a confirmed case of COVID-19 is deemed eligible, the physician can enter the patient’s data into a WHO website, including any underlying condition that could change the course of the disease, such as diabetes or HIV infection. The participant has to sign an informed consent form that is scanned and sent to WHO electronically. After the physician states which drugs are available at his or her hospital, the website will randomize the patient to one of the drugs available or to the local standard care for COVID-19."


I find it curious that favipiravir isn't in the list. Are they just leaving it for Japan to test, so it wasn't mentioned?

Drugs used in this trial should be already widely available and accepted as treatment for other illnesses in many countries.

Favilavir is accepted only in Japan and China.


It's mentioned separately, but not giving any further reasons:

> A global data safety monitoring board will look at interim results at regular intervals and decide whether any member of the quartet has a clear effect, or whether one can be dropped because it clearly does not. Several other drugs, including the influenza drug favipiravir, produced by Japan’s Toyama Chemical, may be added to the trial.

I'm not finding the original statement/source from WHO, I'm sure they gives more reasons there around why some drugs are not included (yet)


Most likely, yes. It's not generally available outside of Japan (and China where it has also been tested) so not much point in adding it to a global trial.

Is that due to licensing issues or manufacturing capacity?

more to do with the side effect. it causes birth defects in pretty alarming rate.

Japanese government only allows it to be used as a backup, Chinese licensed the design but the government only allows it to be used when all other avenues are exhausted.


Aren't most severe patients elderly? Why would birth defects be a problem at this point?

That isn't the issue; it's that this drug is not available at all outside of those two countries. Side effects are the reason it isn't. The WHO is prioritizing drugs with worldwide availability.

If it's effective, I'd hope it would become widely available for future pandemics.

How on earth is Remdesivir available worldwide?

Kuwait just recently bought a large amount of the drug .

Unlikely the reason, since the no 1. on the list is not globally available either. But I guess things can change quickly.

UCSF had a great video conference call with several of their experts essentially covering these drugs and the two different stages of the disease. I highly recommend watching the first half at least. https://youtu.be/bt-BzEve46Y

thanks man, very informative and without doubt well condensed and explained info on the presentation of the disease and treatment options.

I’m wondering what took so long? Shouldn’t we be ready to do something like this almost immediately?

Still seems chloroquine is required to meet a higher bar than the others which continues to be odd.


What took so long? It's clear you've never been involved in clinical trials. This is blisteringly fast. These things usually take many months of planning and review.

Guess when we knew we would have to get serious.

Fast since it has been taken seriously you mean.

Main researchers and leaders were still laughing at us for worrying on Monday evening.


I'm sorry that you seem to work for a crappy hospital. I'm at Mayo, and we've been serious about it for weeks now.

I'm at one of the top european hospitals. And I'm working ICU.

It's the same all across the western part of the continent.


Chloroquine has a host of potential, damaging side-effects. It's reasonable to hold it to a higher standard when the risks of taking it are also higher.

And do the others not?

The fact that this particular treatment has a booster club is odder still. People are acting like they have a personal stake in which treatment wins.

I mean...it seems likely that some of the boosters who have been astroturfing it do actually have a stake.


Look at this thread: everyone has their “one weird trick” (substance) which they believe is the silver bullet. Including garlic.

I doubt that these people have financial stake in any of these products. Sometimes, the substance might be relevant to some opinion they hold dear. The guy recommending garlic might consider all modern medicine a conspiracy of the pharmaceutical industry. A significant number of people on HN and reddit have also been really into cytokines for the last few years, the reason for which eludes me (maybe it’s the one bit of biology they heard of at some point and they are still enamored by their smartitude and want everyone to take notice).

Quinine and similar might have just been popular for the funny Gin&Tonic tie-in. Then, the US president read about, called it “the biggest discovery in medicine” or similar on Twitter, and now the alt-right has a stake in its success.


Now people are dying from taking that stuff:

http://bannerhealth.mediaroom.com/trending-now-at-banner?ite...


It is the one treatment that is cheap, already manufactured in large scale, and not in control of any single entity. Of course people are rooting for it.

"already manufactured in large scale"

People who need it for medical conditions are finding it impossible to find. Not sure how large the production is.

Any treatment that is found to actually be effective in real studies is pretty clearly going to be nationalized, recreated by countries around the globe, and mass produced overnight.


> Not sure how large the production is.

It's large, or can be easily scaled up. Novartis is pledging 130M tablets, Teva 16M, Mylan is planning on ramping production up to 50M.

0: https://www.fiercepharma.com/pharma/new-commitments-mylan-an...


I think the patents on chloroquine are long expired so won’t yield large profits for the pharmaceutical industry. Given the regulatory capture of the FDA in the US, my cynical side would not be surprised if lobbyists find a way to twist things toward the favor of more profitable drugs.

There are always vested interests and people with a financial stake, though I was talking about something much lower-level.

A number of zero-medical-experience personalities have become mini-celebrities with their boosting of it. For those who saw the movie "Contagion", their behavior is shockingly similar to the Jude Law character and its boosting of Forsythia. Some of these people are active on HN, and it's amazing how often absolute trash science is heralded like it's some suppressed truth.

A few days back Trump was retweeting one such huckster. A guy who has repeatedly tried to recreate his own Wikipedia page (to be deleted again for non-notability). Who has a pretty standard hustler-and-exploiter bio. Now he's claiming that he's at the forefront of "working with companies" to get this magical treatment to the streets. It is deplorable.


The WHO usually isn't in the business of sponsoring trials. The development of a single master protocol takes time, especially for one as large as this, and pulling one together within just a few weeks is already quite incredible, given the many, many partners that have to be involved, and how carefully you have to design it to balance a bunch of different factors.

> Still seems chloroquine is required to meet a higher bar than the others which continues to be odd.

Does the article say anything about different bars for the different treatments? I must have missed that. What are you referring to?


It seems quite a few countries in Europe are moving towards hydroxichloroquine/chloroquine and away from lopinavir-ritonavir for the recommended treatment of severe cases.

Top comment alludes to the stage of the disease as being important. So when we talk severity, is it severity by stage? Or overall severity? How will these drugs help when the disease has already attacked the lungs causing the large immuno response?

One common answer to all questions about covid: we don't know. All we have is extremely preliminary evidence, and we're acting on it in hope of luck and also just to provide people with something to cling to.

Isn't lopinavir-ritonavir already verified it doesn't help?

Chloroquine has shown promising results but the number of cases was small and the study was flawed. Lopinavir-ritonavir has shown disappointing results but the number of cases was small and the study was flawed. If you show faith in one you've got to show faith in the other for the sake of methodological consistency.

Read the article. Despite poor results from China, WHO feels larger testing against a wider variety of patients is warranted.

Mostly by "global interest". Reading between the lines, I wonder if they're being pressured.

Kinda makes you wonder why they aren't using methods from the multi armed bandit literature.

Multi armed bandit methods work best with immediate success-fail metrics. This one has time delays.

An example of how machine learning goes wrong is if a treatment slows down the progression but increases the death rate. Given exponential ramp up in the incoming cases, it will look good until the final horrifying numbers are in. You need to slice and dice the numbers by cohort to detect/react to this.


I decided that some numbers on how things go wrong would help.

Suppose that the treatment increased deaths by 50% but delayed death by a week. And we have a doubling rate for the disease of 1 week.

Back of the envelope that means that the treatment will have 1.5x the deaths from when the disease happened 0.5 times as much for 0.75 of the deaths at any point in time. It looks like it saves 25% of lives when in fact it kills 50% more people. The raw numbers will look good until you look at a cohort over time.

Current doubling time for deaths has been about 3 days. My assumption of a week is therefore optimistic. Perhaps we get there with social distancing.


"Multi armed bandit methods work best with immediate success-fail metrics. This one has time delays."

Well, sure, but everything works best with immediate success-fail metrics. That's one of the most basic results from learning theory is that the longer the latency between stimulus and response the slower the learning rate can be. I'm not sure how multi-armed bandit is special in this regard in any particular dimension. All learning techniques are going to be susceptible to the problem you outline in your second paragraph.

This is one of those "there is no perfect solution" situations. It's really easy to say that out loud. It's quite difficult to internalize it.

(Also, just as a note to your other post, bear in mind that our hard-core "social distancing" efforts in the US are just about to reach approx. 1 incubation period. It is only just this week that we're going to start seeing the results of that, and it'll phase in as slowly as our efforts 1-2 weeks ago did. My state just went to full lockdown today, though we've been on a looser lockdown for a week before that.)


Everything works better with immediate success/fail metrics. However the simplest approach is easiest to analyze, and is easiest to analyze after the fact in as many ways as you want. The more complex the decision making, the less we should be willing to put it under the control of a computer program. (Unless that program has been well-studied for our exact problem so that we trust it more.)

Which medicine looks effective? Which medicine gets people out of the hospital faster? What underlying conditions interacted badly with given medicines? These questions do not have to be asked up front. But they can be answered afterwards. And knowing the answers, matters.

Here is an example. Suppose that we find one medication that gets people out of bed faster but kills some. In areas with overwhelmed hospitals, cycling people through the bed may save net lives. If your hospital is not overwhelmed, you wouldn't want to give that medicine. Now I'm not saying that any of these medicines will come to a conclusion like that. But they could. And if one did, I definitely want human judgement to be applied about when to use it


I don't think anyone is proposing actually removing all humans from the loop, so I think that's an argument against a strawman.

Even if they were proposing it, there's no realistic chance of it happening.

I don't want people blindly copying "standard" scientific procedures either, where we run high-stastistical-power studies for months with double-blind scenarios then carefully peer-review it and come up with some result somewhere in 2022.


So, hopefully there will be blinded researchers who analyse the data.

They'll probably use sequential stopping rules to take samples of incoming data.

If one of the treatments works much much better, then they'll almost certainly recommend that (but doctors will probably figure this out first, anyway).


For someone unfamiliar with machine learning literature, can you please briefly explain how it helps here?

in a world where you have many options and have to figure out which is best by repeated experimentation, but where experimentation itself has some cost, you have a multi-armed bandit problem. (the name is supposed to evoke a room full of slot machines -- you want to find the one with the highest payouts by repeatedly playing them, while losing as little money as possible before you find it.)

for example, if you have a few medications, you might start by trying them all equally at random and then as data comes in, use a bandit algorithm to gradually shift more and more new patients onto the ones that prove most effective, in a way that optimally trades off accurately estimating the effects with wasting time testing the less effective drugs.

interestingly, the first formulation of the problem is due to Dr. Thompson at the Yale Pathology Department in the 1930s; he came up with Thompson sampling. So these are techniques that were originally designed for medical trials.

I think that designers of medical trials probably do have a good grasp of this stuff (some statistical estimators that originated in the medical world have even been successfully imported into reinforcement learning/MAB research) so probably they would be using a bandit-like technique if they felt it made sense.


Coordination is complicated. Bandit problems also assume clear/instant payout per levee pull. I suppose WHO could do this in the back end.

Every patient would be treated with random drug/treatment. With accumulated treatment results, a multi armed bandit algorithm would adjust probabilities so, that most effective treatment would be used more often.

For example, in Thompson sampling probability of choosing option is equal to probability of that option being the best option given evidence so far.

Aim is to maximize reward (successful treatments), while spending little as possible time on exploration (testing less effective treatments).


Delay is a real problem here.

Trial would take longer.

Is the expectation that SARS-CoV-2 will become a seasonal virus, permanently with us, like the other 4 coronaviridae (which we don't have a vaccine for)?

https://www.cdc.gov/coronavirus/general-information.html

It seems that those keep coming back each year either because they mutate, or the antibodies produced by our bodies are no longer produced/effective after a year or so.

So even if we had a vaccine for SARS-CoV-2, would it lose its' potency after a year, and people need annual booster shots? Or would the virus mutate so that the specific vaccine no longer works?

Also, is the expectation that the mortality rate of SARS-CoV-2 will reduce over time because of evolutionary pressure? Is that really the case, given that much of the spreading happens in the first 1-2 weeks, before the host is potentially dead?


> So even if we had a vaccine for SARS-CoV-2, would it lose its' potency after a year, and people need annual booster shots? Or would the virus mutate so that the specific vaccine no longer works?

The answer is probably yes to both questions.

RNA viruses, like influenza, HIV or coronaviruses do mutate a lot. This is what makes them hard to eradicate, plus in the case of influenza at least we've got an endless reservoir from wild animals.

But a lot of people are already vaccinated for influenza every season and the availability of a vaccine that works for 6 months would make this very manageable, even for people that forget or refuse to vaccinate, due to herd immunity.


There is research going into their more fundamental structure to try and defeat their whole family at once

So out of this mess, we might end up with a cure for the common cold?

Not really. The majority of cases of the common cold are caused by rhinoviruses, not coronaviruses. Then again I would gladly take a 10% cut in the common cold...

Quite! Though at first the goal is to create vaccines that affect multiple strands and are likely to be vulnerable to in the future

any link? I'm curious and searching brings current crisis.

When are the results expected?

From what i've heard in the news about this trial, not until 6 weeks minimum

Combinations need to be tested.

Seems strange that they're not going to use a control group that receives a placebo.

Would this new global mechanism be useful for other diseases like cancer or AIDS?

Forbes: Coronavirus Patient Dodged A Bullet With Hydroxychloroquine.

Mar 22, 2020,01:43pm EDT

"Two scientists at major university centers reviewed the French trial for me. They agreed, separately, that while the study is preliminary, small, and not without flaws, its findings were strong enough, given the drugs’ known safety records, to guide treatment decisions in a crisis.

“Despite the limitations of this study, in the absence of any effective treatment, in this urgent situation, this Plaquenil and Azithromycin combination therapy should be given to patients with COVID-19 as a treatment option,” Ying Zhang, a professor of microbiology at Johns Hopkins Bloomberg School of Public Health, wrote in an email. “For now, there is no time to wait."

Brian Fallon, a research scientist and clinical trials investigator at the Columbia University Irving Medical Center, agreed on the study’s overall merit despite the patients who dropped out. After analyzing the data and counting all six dropouts as treatment failures, he said the overall rate of improvement was still statistically significant for the entire group, though not for the hydroxychloroquine group alone."

https://www.forbes.com/sites/marybethpfeiffer/2020/03/22/one...


We detached this subthread from https://news.ycombinator.com/item?id=22663274.

Not sure if you replied to the right message, but yeah there are safety risks that's why we evaluate them in trials. Favipiravir also has pretty bad side effects (teratogenic effects -thalidomide like). That's why we run these trials, and using already approved drugs or more widely developed drugs will mean we're dealing with known evils as opposed to unknown evils

It's not just hydroxychlroquine, the best results in the (albeit small) study were when it was paired with a Z-pack. A WSJ editorial had also talked about the same thing, but the Forbes article you posted has more details:

"The just-released French study reported that 70 percent of hydroxychloroquine-treated patients, or 14 of 20, were negative for the virus at day 6, as were all six patients who were treated with hydroxychloroquine and the antibiotic azithromycin (which Novins also received). But the study was small – 20 treated patients and 16 controls – and had other serious limitations."


Agree. This is interesting.

Covid-19: India Recommends Hydroxychloroquine As Prophylaxis For Healthcare Providers, Patient Family Members

https://swarajyamag.com/insta/covid-19-india-recommends-hydr...


the biggest limitation was that several pts in the treatment group went to the ICU and didn't continue to receive treatment, so we don't know how it affects the worst-case disease progression.

[flagged]


Please keep partisan flamebait out of your HN posts. It leads to partisan flamewars, which are the last thing we need here.

https://news.ycombinator.com/newsguidelines.html

matsemann 6 days ago [flagged]

> I also wonder if millions will die around the world because nobody wants to admit that a thing Donald Trump tweeted about is the solution

I'm sorry, but why the hell would that happen? That's gotta be the most insane premise I've ever read on HN.


Please don't take HN threads further into flamewar.

https://news.ycombinator.com/newsguidelines.html


A quick web search answers your question:

https://www.cnn.com/2020/03/23/africa/chloroquine-trump-nige...

https://www.usatoday.com/story/opinion/todaysdebate/2020/03/...

The media went into full gear attacking chloroquine as dangerous and attacking Trump for mentioning it, pretty much the moment the tweet was posted.

Still the “most insane comment you’ve ever read on HN”? (rhetorical)


I'm not rooting against chloroquine (or any other therapy, and I don't think anyone else is either: you'd need to be a monster to want people to keep dying, keep getting sick, etc.

I am, however, vehemently opposed to sloppiness. It's tempting to say "The stakes are too high to be careful--let's just go with what've got" Nothing could be further from the truth--things need to be done carefully because the stakes are so high. Suppose we go all in on chloroquine and it doesn't actually work. Now we've got a big pile of useless stuff AND fewer resources, fewer staff (since they get sick too), and so on.


I’m sure it was just a typo and they wanted to call your comment “inane”.

[flagged]


Personal attacks are particularly not ok, so please edit them out of your HN posts.

https://news.ycombinator.com/newsguidelines.html


You know your comment would've been very reasonable if you stopped after the first sentence, omitting the personal attack.

Ending with an incendiary sentence while simultaneously claiming to be above internet flame wars is also likely not what you intended.


There are plenty of hateful and sociopathic people across the political spectrum. There are also people across the entire political spectrum that care about helping others. It's narrow minded and bigoted to suggest otherwise.

I dislike Trump, but even I can see that there is a subset of the media that has a conscious or unconscious principle of attacking him for almost everything he says and does. If he does X, attack him for doing X. If he does the opposite of X, attack him for doing the opposite of X. If he first does X and then does the opposite of X, first attack him for doing X and then switch and attack him for doing the opposite of X. If he makes peace, call him soft on our enemies. If he makes war, call him bloodthirsty. Etc.

This subset of the media also tends to lie about what Trump says. So, for example, a few years ago they insinuated that Trump had called all Mexicans criminals and rapists. Trump never had. A couple of days ago, they insinuated that Trump blew up on a reporter because the reporter had asked Trump what he had to say to Americans who were scared. In context and without selective editing, though, it seems a lot more likely to me that Trump blew up on the reporter because the reporter had just a few moments previously asked Trump if Trump's tendency to put a positive spin on things was giving people false hope.

This subset of the media has been accusing Trump of exaggerating the effects of hydroxychloroquine. But I have not been able to find any statements from Trump in which he promoted it one-sidedly as any sort of miracle cure.

Trump lies all the time, but that doesn't mean I don't see that a certain fraction of the media lies about Trump too.

I don't think it's a total collective effort on the part of the media to smear Trump, although I'm pretty sure that there are pockets of collective effort there. There are subsets of the media that uncritically praise Trump, so it's certainly not one-sided. There's probably also a clickbait profit motive behind some of the deceptive media coverage. It's a lot easier to take a Trump quote out of context and write a quick article making it seem like he said something horrendous than it is to actually do investigative journalism and thoughtful commentary.

As for whether such media people would care about millions dying? Well, did they care enough about the fate of Iraqis to be adequately cynical about Bush administration hogwash about WMDs that any intelligent teenager with an interest in war could have poked holes in? No. To add insult to injury, many of them actually started to fawn over the Bushes once Trump was in office.

>Note that those opposing Trump are on the left

"Left" is not well-defined. If you mean "left" as in socialist-leaning, then I would say that I doubt that a clear majority of Trump's opponents are on the left. Many are.

>and actually care about helping others

No matter how you define "left", the "left" has no monopoly on actually caring about helping others. And the "left" has plenty of people who talk a big talk about wanting to help others but who actually, whether they realize it about themselves or not, only care about power and use the talk of help as a disguise. All parts of the political spectrum have people like that.


[flagged]


Please don't post any more flamewar comments to HN.

https://news.ycombinator.com/newsguidelines.html




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

Search: