The effectiveness of cloth masks has been misrepresented | Hacker News Reader

The effectiveness of cloth masks has been misrepresented

rssdss.design.blog

39 pointsmartingoodson36 comments

I'm part of a mask-providing charity. [1] This article is misleading in three ways:

1) It acts as if a single person speaks for #Masks4All. But #Masks4All is a hashtag, not a person.

2) It says that #Masks4All claims cloth masks are 96% effective. I've never heard that. We always claim cloth masks are 30% effective, surgical masks are 60-80%, and N95 masks are 95% effective.

3) It says that cloth masks are 70% effective, and then exaggerates this to say that cloth masks are "not effective."

Who in their right mind would say that "70% effective" means "not effective?"

[1] https://maskedprotectors.org/

> 1) It acts as if a single person speaks for #Masks4All. But #Masks4All is a hashtag, not a person.

This is always a big indicator that someone has no idea what's going on on the internet. A similar trend goes for almost anyone who talks about "anonymous".

There are a bunch of variables here that we need to know to be able to say whether any type of mask reduces the likelihood of infecting others, and the percentage of viral particles filtered by the mask is just one of them.

Another key factor is the rate at which the risk of infection changes depending on the percentage of particles filtered.

For instance, it's conceivable that you can be equally likely to be infected whether 5% or 95% of particles are filtered, depending on the "potency" of the particles that are not filtered. If contact with just N particles increases a person's likelihood of infection by some percentage, and the average cough transmits 100N particles, then even if you filter out 95% of those, you're still at 5N.

But we know that in most cases, increasing the percentage of filtered particles maps in some way (though maybe not linearly) with reducing the number of infections.

All that is to say: I think we need to be careful about using this one variable -- percentage of particles filtered -- when talking about effectiveness of masks. Effectiveness is more of a holistic calculation, taking into account viral potency, the distance unfiltered viral particles travel through the mask, the degree to which they linger in the air, etc.

yes, please keep adding level-headed reasoning to the conversation despite the continuing public panic over the topic.

another major confounding variable that makes those effectiveness claims misleading is the environment in which they're used. ENT surgeons (and their patients) are absolutely positively at heightened risk by literally being in the face of patients for prolonged periods. they should use all sorts of countermeasures, including surgical masks.

the typical person on the street gains no added benefit from wearing a mask, regardless of the effectiveness of a given mask at blocking particles, as the ambient viral concentration is effectively non-existent basically everywhere outside, assuming most folks aren't congregating (closely and face-to-face) regularly. even in most indoor situations, mask effectiveness only comes into play in very specific situations.

if no one shoots the ball, you can't block their shot, no matter how great a shot-blocker you are.

latching onto miraculous prophylactics without skepticism and thought is a danger in itself.

> even in most indoor situations, mask effectiveness only comes into play in very specific situations.

I think you might be underestimating how transmissible this particular virus is. If it were that hard to catch it, people wouldn't be catching it at the rate they are.

no, note that the confirmed infection rate as of this morning in CA is 0.1% (~40K out of ~40M). because of the roughly 2-3 week lag in effects, the moderation of that rate is mostly attributable to modest distancing rules, before masks were widely encouraged.

21% here in NYC. [0]

Note that this finding is materially different from zero, unlike the Stanford study. [1]

Big picture : Don't be misled by the confirmed case count.

[0] https://www.reddit.com/r/COVID19/comments/g6pqsr/nysnyc_anti...

[1] https://statmodeling.stat.columbia.edu/2020/04/19/fatal-flaw...

Macintyre (2009): Parents wearing masks reduced the risk of contracting influenza from their sick children by 74% (95% CI: 33%-91%). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2662657/pdf/08-...

Jefferson 2008: Wearing masks reduced the risk of contracting SARS by 68% (95% CI: 60%-75%). Wearing N95 masks reduced the risk by 91% (70%-97%). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2190272/

MacIntyre (2015): Use of cloth masks instead of surgical masks increased the risk of infection by 13x (95% CI: 1.69x to 100x). Cloth masks filtered out only 3% of particles, versus 56% for surgical masks. https://bmjopen.bmj.com/content/5/4/e006577.short

Overall, these findings are suggestive that the infection protection factors are broadly similar to the total filtration efficiencies.

> It says that cloth masks are 70% effective, and then exaggerates this to say that cloth masks are "not effective."

No. It says that the main South Korean study "found" that masks were 70% effective, but that that study is not significant, because it was based on only 4 people, "one of whom could not produce detectable virus particles even without a mask" -- so really, 3 people.

It also says that a much more important study [1], based on 1607 hospital workers found that penetration of cloth masks by particles was almost 97%.

That other study was made in 2015, so before covid, but there is no reason to believe its results would be different with the covid virus; it's certainly not disproved by a study based on 3 people.

[1] https://bmjopen.bmj.com/content/5/4/e006577

Also, the 4 person South Korean study was very clear in their conclusion:

https://annals.org/aim/fullarticle/2764367/effectiveness-sur...

> In conclusion, both surgical and cotton masks seem to be ineffective in preventing the dissemination of SARS–CoV-2 from the coughs of patients with COVID-19 to the environment and external mask surface.

That study is n=4. They didn't have enough statistical power to make any conclusions of that sort.

Incidentally, if you look at the SK study's actual data, it appears that the cloth and surgical masks reduced the concentration of the virus by about 0.5 to 1.0 log units -- that is, by 66% to 90%. However, those results results were not statistically significant, because n=4.

Nancy Leung, from Benjamin Cowling's lab in Hong Kong, did a much better study, n=114. They found that surgical masks did confer significant protective benefits.

https://www.nature.com/articles/s41591-020-0843-2.pdf

First, each patient coughed directly upon a petri dish 20 cm away, not once but 5 times.

I don't think anyone would rationally believe that a non-fitted mask would have a high probability of providing protection to a person directly coughed on by an infectious mask-wearer, all at a distance of 20cm. If they were going to test under those conditions, they should have used an N95 mask. Doing it with a surgical mask under such conditions was a waste of time and no surprise at all.

The argument has never been that non-fitted masks prevent 100% of all viral particles emitted by an infectious person from escaping into the environment (which is what that tests, since it only takes a few virions to start a colony in a culture).

The argument has been that wearing a non-fitted mask can significant decrease the amount of virus-laden droplets floating around in the air, and falling onto surfaces.

The more we find out about asymptomatic carriers, the more this makes sense. These carriers aren't going around coughing their lungs out, they transmitting disease by talking and breathing. A mask likely helps a lot with that.

I've really like to see the same experiment repeated, but with the patient wearing the mask and talking, and the petri disk held ~2 meters away. These are the conditions under which the disease is being transmitted now, outside of households and healthcare facilities.

> a much more important study [1], based on 1607 hospital workers found that penetration of cloth masks by particles was almost 97%.

Amusingly (or horrifyingly, if you prefer), I'm fairly sure that's the study that #Masks4All posts were referring to when they claimed that cloth masks are 96% effective. I recognize the 1607 number at least.

> 1) It acts as if a single person speaks for #Masks4All. But #Masks4All is a hashtag, not a person.

The article is referring to Jeremy Howard who set up Masks 4 All, and coordinated a little bit of research around the effectiveness of masks.

https://masks4all.co/in-the-media/

https://www.fast.ai/2020/04/13/masks-summary/

His messaging is a bit frustrating because he misrepresents the science, he says stuff that's anti-science, and he's pretty dogmatic.

He owns the domain masks4all.co, and set up that one particular website, but he does not own the #masks4all subreddit [1], youtube tag [2], twitter tag [3], or the variety of other #masks4all websites such as maskssavelives.org [4] and maskedprotectors.org [5].

[1] https://old.reddit.com/r/Masks4All/

[2] https://www.youtube.com/results?search_query=%23masks4all

[3] https://twitter.com/search?q=%23Masks4All

[4] https://maskssavelives.org

[5] https://maskedprotectors.org

This article is falsely presuming that Jeremy Howard is the leader of the masks4all movement in order to lambast the movement. If it wanted to disagree with Jeremy Howard, or masks4all.co, it should put that in the title. Not #Masks4All.

'It says that #Masks4All claims cloth masks are 96% effective. I've never heard that. We always claim...'

Author of the article here: this is from the masks4all.co website: 'If you have COVID-19 and cough on someone from 8 inches away, wearing a cotton mask will reduce the amount of virus you transmit to that person by 36 times...'

A 36-fold reduction is equivalent to a 96% reduction.

> It says that cloth masks are 70% effective, and then exaggerates this to say that cloth masks are "not effective."

This caught my eyes too. If we had a simple way to reduce the virus dissemination by 70%, we would be able to finish the entire pandemic in a couple of months.

> Who in their right mind would say that "70% effective" means "not effective?"

People who want to keep being scared into indefinite quarantine (which unfortunately does require the same from everyone else). If questioned they’ll mention some really undefined requirements so they can keep moving the goalposts.

I get the importance of following science in weird times like these.

But, if the author of articles agrees people should still wear masks, what is the end game of this article?

The people who want to open back up economies using masks instead of appropriate other precautions, will not read this. The people who will read this and understand it will not change their views. The people who will read this and misunderstand it will just add to the negativity and confusion by claiming that no one should wear masks.

Maybe I'm just going into a depression spiral, but what's the point?

If it were discovered that people systematically refuse to correct false claims about the efficacy of masks, that would hurt confidence in mask recommendations much more than any individual article could.

I agree that a headline like "Cloth masks are effective, but Masks4All is doing bad science" would have been better.

This is how I feel at work sometimes; like someone will say something kind of wrong but mostly right or their rationale for something is a little bit off. Is it really worth derailing a meeting over?

This is the worst. If you correct them you'll derail the meeting for a short time and look bad. If you don't correct them, there is a risk somebody will follow their premises into some conclusion that leads to very bad actions, and you won't be able to explain why, derailing much more than a meeting.

The author of the article did not say that "people should still wear masks", he said "I’m not claiming that people shouldn’t wear masks".

If you want something more actionable, "behave as if nobody was wearing a mask" which is absolutely not the position the #Masks4All guy is taking; he is promoting people in physical proximity but wearing masks as appropriate.

While I don't know much about anyone involved, the danger of telling people "wearing masks is really effective" if you can't back it up is that it may encourage changes to peoples' behaviour. Even if you also tell them to socially distance anyway.

"Wear a mask, but it probably doesn't do much" is much more reasonable and responsible messaging.

Totally anecdotally, but when I'm walking to the supermarket, I find I have to step into the bus lane (the roads around here are basically empty right now) to avoid the mask people way more often than the non-mask people; a lot of people behave as if wearing a mask makes it okay for them to walk in the middle of the footpath in such a way that keeping 2m distance from them is impossible without stepping into the road.

Author here. Would your behaviour be different if you knew that a cloth mask lets through 30 to 97% of COVID-19 particles rather than thinking they can filter almost all infective particles? If so, this article is for you.

(For example, perhaps you would lobby your government to ramp up production of surgical masks for the general population.)

So IF I read this correctly, cloth masks still do a pretty good job? The post ends with...

"I have reanalysed the Korean data using the suggested replacement value of half the LOD and the results don’t change very much, suggesting a reduction of 70% of virus particles when using cloth masks."

And it starts with...

"His key claim is that “cotton masks reduce virus emitted duringcoughing by 96%”, citing a recent South Korean study."

So 70% isn't as good as 96%, but it seems like that's still a decent number. Or am I reading this wrong?

When giving out numbers scientifically, it's important to be precise, or have the uncertainty known. In that case, 96% is widely different than 70%. As Andy Slavitt says in his April 22 podcast, when someone claims to know something too precisely in regards to this pandemic, and they don't say "we don't know" enough, run the other direction.

You also want to have someone doing a scientific study to be as unbiased as possible. As I understand it, (most of?) his family is immunocompromised. This doesn't mean he shouldn't do it, nor that what he's doing is not saving millions of lives, but it's a little easier to question the "salesmanship" aspect of it when he hasn't spent a lot of time in the field.

"Cloth masks reduces virus from coughing by 96%"

"Cloth masks tested on 4 people reduces virus from coughing by 70%"

If that is the only study being citied to go against public health officers I'm be skeptical. If cloth masks work it should be feasible to do a larger study and prove it with a high degree of confidence.

The study that finds 70% effectiveness was based on just 4 people, one of whom couldn't produce virus particles even without a mask.

Another most serious study based on over 1500 hospital workers found that cloth masks let through over 95% of particles.

That more serious study was done in 2015 but there's no reason to think it would not still be correct.

No, 70% isn't nearly enough. A single contamination event (like a cough) is in the neighborhood of a million virus particles to stop.

Think about hand sanitizers: some advertise 99.9% reduction, which is "millions become thousands", bare-minimum levels of actually reducing risk of infection from an exposure. 70% is like "millions become fewer millions"

70% is "a 2m safety distance becomes some X, that isn't very easy to calculate". There's very likely an exponential fall of virus with distance, so it's probably something visibly smaller than 2m, but not by a lot.

Besides, that's hardly the only variable at play. The virus doesn't spread on the atmosphere only by dispersion, so the particles energy is relevant. It doesn't spread in only in dry particles, so that ratio is relevant, as is the efficiency in blocking the larger wet particles. And it doesn't spread only by coughing, so the efficiency on normal breathing and speaking is important too.

Also, we don't need completely effective barriers for people at the street, this is not ebola. We just need to reduce the virus spread by ~60%.

The one thing we know is that the countries where people use masks are having an easier time fighting the virus. We don't really know if it's because of the masks, but it's a good bet.

> The researchers found that 97% of particles penetrated through cotton masks. Why would a ‘review of the evidence’ neglect this key finding?

This statement in the article is completely disingenuous. The study in question[1] was looking at whether wearing a mask protects the wearer from getting infected. The primary purpose of wearing the mask is to prevent the wearer from infecting others. These are unrelated!

[1] Text of the study: https://bmjopen.bmj.com/content/5/4/e006577.full

This is pure clickbait and here is why. In the debate over masks, the anti-mask side always forgets that this epidemic is multiplicative i.e. it each person spreads it to multiple people (R0 > 1).

Even a marginally effective mask that cuts infection by 30% will greatly bend the curve. How? A 30% effective mask (where 100% is full hazmat outfit) does not mean 30% fewer cases. It could mean greater than 30% or even 90% fewer cases. The same non-linearities (read: exponential) that work for virus spread also help in the other direction. [1]

Social distancing is expensive and we need it. Social distancing requires the shutdown of large parts of our economy and changing the way we live. Masks are a relatively minor inconvenience and one whose effectiveness, no matter how marginal, we can not afford to ignore.

Do science all you want, and that blog post is hardly science, but do it with an ethical and moral responsibility to not turn people away from masks. Turning people away from masks is akin to telling them the virus is not contagious or fatal.

References:

[1] https://twitter.com/nntaleb/status/1249296844712218624

the point of the article wasn't to say don't wear a mask. the point was that it is dangerous to tell people a mask is 90% effective when its not. lying about the effectiveness WILL cause people will take risks they shouldn't making the situation worse. being honest about the masks effectiveness and explain why it still helps is the best thing we can do.

I wonder if the author of this article is aware that evolutionary biologist Bret Weinstein uses and promotes using a bandana as being far more effective than nothing for those who are unable to get an N95 mask.

Reply on news.ycombinator.com