MIT study finds indoor mask wearing much more effective than social distancing | Hacker News Reader

MIT study finds indoor mask wearing much more effective than social distancing

pnas.org

63 pointsantattack37 comments

"By assuming that the respiratory droplets are mixed uniformly through an indoor space..."

... Is this a reasonable starting point? Social distancing assumes the distribution is not uniform. You would think the researchers would try measuring the distribution first instead of just assuming that a priori.

Edit: Haha, I'm struggling to understand the point of this study when they state something like this: "These models are all based on the premise that the space of interest is well mixed; thus, the pathogen is distributed uniformly throughout. In such well-mixed spaces, one is no safer from airborne pathogens at 60 ft than 6 ft."

No not at all.

Airflow in buildings is a complex topics and needs to be modeled with complex computer simulations to get right as "intuitive" assumptions can sometimes be very far of the reality.

Any single source of heat, like heaters, laptops or even your body heat is influencing this.

There are companies which do such analysis for hospitals to make sure/reduce the risk that Covid from Covid patients infect non-Covid patients over air flow. (Best would be an airlock between areas, but lets be realistic especially when they get many Covid patients this is often not possible).

Either way it's completely unrealistic assumption.

Still from all what I know about air (subtile) flow (which is not much) the end conclusion of: "wearing mask indoors is quite effective, maybe more then social distancing". Might still be true, assuming you don't strictly social distance etc.. Because not being ever in the same apartment as an affected person is definitively better then being there with a mask ;=).

A example of how waring a mask indoors might unexpectedly help is that you body heat creates a minimal updraft, minimal but then the virus water droplets are also supper light wight. So if you combine a air conditioning/filter system which pulls out air (not uncommon for hospitals) with a mask and that effect the mask will cause the remaining airosoles to be "closer" to you as they where at least slowed down even if they escaped and the minimal updraft + vacuuming air filter system will with that have a higher chance to "pull out" the virus before any one can get affected by it. Or at least that is what I heard from some people of which I forgot the name of which do indoor air flow simulations.

Borders on fraud, when such a strong unproven assumption gets transformed into a "guideline", which will then be disseminated and reported to a wider audience which isn't inclined to look into the details.

In an enclosed space mixing will eventually occur. The question is; how fast? So if social distancing initially works then there would come a time where it does not. If the mixing is fast enough in practice then this model might not be effected all that much if it predicts significant risk later.

This is a mathematical model, not a study of empirical data. It's worth distinguishing, because models are only as good as their assumptions.

The model presumes a "filtration constant" for masks, which only apply for pressure-sealed masks, like properly-fitted N95s. Simple face coverings don't do much to protect you against aerosols, for the same reason that they don't do much to mitigate the effects of attempting to breathe in a smoke-filled room - the airflow is just redirected through the sides and top of the mask, rather than being filtered. So, while the results may have some value for N95s, I seriously question their value for "masks" as they are currently colloquially understood.

A filtration constant can apply to poorly-fitted masks, too. They just have a smaller constant.

The filtration constant for a well-sealed N95 can be 99.99%. The filtration constant for a thin cotton "face covering" can be 20%.

The point isn't that the model doesn't apply to poor masks -- it's that they have poor constants. We will do much better to include them in the model, and then mathematically predict how important it is to improve your filtration constant, than to say "those masks don't apply" and throw up our hands at quantifying the improvement that could be gained by better masks.

It seems we are increasingly leaving the scientific method out of science.

The scientific method is the most powerful tool we have to determine causation. It's also very strict, which means that it's limited in applicability.

It's strict because you have to intervene in the real world after forming a hypothesis about what change might be caused by that intervention. The experiment should be constructed to try to refute the hypothesis.

Increasingly, what I'm seeing is observational "science", where you keep forming hypothesis until you find one that fits the data you have.

Another thing I see, which this MIT study appears to be, is where the scientists completely make up a fictional world (called a model), and intervene in that fictional world.

(Note: models are more likely to work when they aren't completely made up by scientists. For instance, mouse models are very important to life sciences. But the reason they are useful is because scientists can't make up a complete fiction to support their preferred conclusion -- they still have to face the realities of mammals in all their complexity.)

I wish we could have a different word for this, so that people can't use the vaneer of science to add authority to something as though it were a law of physics.

I mean respiratory virus meet respiratory system. You cover your respiratory system and you are most of the way there for protecting yourself. Social distancing indoors wouldn't be effective without ventilation as you'd be breathing in the same air that others are exhaling. Touching on the other hand, it is very effective for.

You still wouldn't want to be in a room for extended time if someone is exhaling the virus as the mask can only do so much.

Just more evidence for what I think is obvious.

The mask also "slows down" the exhaust making airosoles which escaped the mask of the infected person "be" more around that person, instead of further spread.

Which combined with e.g. a air filter system which pulls out air (instead of pushing it in) can reduce the risk by quite a bit (especially if the system is installed at the ceiling). Or so I have heard.

Which makes me wonder if you could "invert" and AC to act as such system in a improvising manner.

I think there are also AC, which do both: Pushing in fresh air and pulling out old air. Which probably have the best chances to reduce infection, especially if combined with wearing at least some kind of mask.

I hope we can get to a place by the Fall that everyone leaves windows open in public buildings, workplaces, etc. if they can be opened. It's the simplest mitigation strategy. The next easiest is changing the HVAC system to pull outdoor air in rather than recirculating the air, and putting in a higher MERV filter. Mask wearing is good too, but should be seen as a bonus on top of these simpler strategies. Less surface cleaning and more focus on air.

Mask wearing is incredibly effective with N95-equivalent masks, which are now easily and plentifully available.

If everyone wore masks with proper filters, we might be done by now.

The world seems too dumb to notice the difference between a piece of cloth and a proper filter.

If everyone wears N95 masks, the virus basically doesn't seem to spread.

Agree of all the stupidity in this pandemic, the false equivalence implied by leadership between n95s and cloth masks has probably killed the most people counterfactually. Particularly if you assume it has led to a false sense of safety by cloth-wearers, leading them to take risky behavior like going indoors they would have avoided otherwise if their cheap masks were explained to be insufficient protection.

I tend to agree with you on this. I've got some anecdata and commentary. So obviously I'm not the CDC or a doctor so don't decide anything based on my story here.

My wife works in the hospital and employees are only given one mask a month or so and must use the hospital provided equipment. For staff that are not in the covid units, incidents of transmission and overall infection rates have been very low. For those in the covid units themselves, the infection rates are 60-70%. The PPE situation is still quite poor, gowns and face shields are being reused where in ordinary times they would be disposable.

For ordinary people without regular contact with the infected and who take reasonable precautions (modeled in my anecdata by the staff outside the covid units) I believe that the masks combined with frequent hand washing are probably pretty effective. If you're in suboptimal circumstances and in constant contact with the infected I don't know. It's hard to tell with all of the PPE reuse.

The story being told is that the cloth masks reduce spread from you to others but do not overtly protect you. I can see circumstances where improper doffing techniques could cause these to be counterproductive but overall I tend to believe this. Properly fitted N95 masks have similar protection of others but also protect you from some, but probably not all exposures.

> My wife works in the hospital and employees are only given one mask a month or so and must use the hospital provided equipment.

Really? I mean really?

N95 masks are not reusable. They have an electrostatic layer that degrades after hours of moisture from breath.

In hospitals in Asia, where this is done right, infection rates have been close to zero.

US companies have a glut of N95 mask stock right now because no one will buy them due to cheaper alternatives. In addition, tons of places produce tested N95-equivalent masks which have not gone through certification procedures.

Where is your wife?

That seems like a class action waiting to happen.

Central Florida.

I believe the issue is two-fold. The Hospitals have contracts etc with GPOs that often require the hospital to make most if not all of their purchases (PPE and Drugs, there is some discussion about GPOs being partially to blame for drugs shortages). If those organizations are unable to source, or unwilling to source PPE (speculation - due to market value being over contract value?) then the hospitals receive a restricted supply e.g. a shortage.

The second axis is that this is not a situation that is generally known and hospital staff are generally underrepresented politically. To whom is the Nurse Assistant or LPN to complain to about their company practices? If you complain, you risk termination etc. Given that most took the vaccine the issue is partially mooted but still remains.

Well, no. The issue isn't mooted.

1) The vaccine doesn't prevent you from catching COVID19; it prevents symptoms. Your body fights it off. We know nothing about how it impacts long COVID or spread. I mean, we know it reduces them, but is it a 10% effect? 90%? If hospital staff are spreading COVID to patients, we've got a problem.

2) Vaccinated people in contact with non-vaccinated is how you get vaccine-resistant mutations.

We need vaccines and proper masks, especially in hospitals.

If there's an issue like that, it seems like an anonymous complaint might help. Letters to newspapers, regulators, and hospital leadership, all cc:ed so they know others have received them, outlining that:

- Proper masks are available: links to US plants having overstock, new technologies (e.g. Korean nanofiber)

- Used N95 masks don't work

- Reuse of N95 at the institution

That puts a lot of pressure on leadership to address the issue. Enforcement actions usually go nowhere if an issue was already resolved by the time the agency gets there, and it's a lot cheaper to deal with early rather than late.

> If everyone wears N95 masks, the virus basically doesn't seem to spread.

I don't know about elsewhere, but at least here in Germany, it would be literally illegal for an employer to require their staff to wear N95 masks at all times. Worker protection laws mandate 30 minutes of break time without a mask for each 90 minutes worked with a mask on.

(Disclaimer: IANAL.)

Really? That seems like it would make a lot of things exceedingly difficult (e.g., certain kinds of medical procedures or scientific experiments).

The world was told by the public health apparatus, up to and including the U.S. Surgeon General, that masks were not effective. The circle was then squared by telling people to wear "face coverings" and, well, here we are.

I would have hoped that in our current age of misinformation, that sources people trusted would not actively lie, but there you go.

> I would have hoped that in our current age of misinformation, that sources people trusted would not actively lie, but there you go.

Find a meta-analysis of mask wearing to prevent infection that was published at the time. Or find some RCTs of mask wearing to prevent infection. There are lots of papers looking at infection control - it's an important part of healthcare. Have a look at how they describe the quality of evidence.

Here's a well-known paper, published in a high impact reputable journal in June 2020. https://www.thelancet.com/journals/lancet/article/PIIS0140-6...

Distancing is rated at "moderate certainty"

> Transmission of viruses was lower with physical distancing of 1 m or more, compared with a distance of less than 1 m (n=10 736, pooled adjusted odds ratio [aOR] 0·18, 95% CI 0·09 to 0·38; risk difference [RD] −10·2%, 95% CI −11·5 to −7·5; moderate certainty); protection was increased as distance was lengthened (change in relative risk [RR] 2·02 per m; pinteraction=0·041; moderate certainty).

Face masks only achieve "low certainty"

> Face mask use could result in a large reduction in risk of infection (n=2647; aOR 0·15, 95% CI 0·07 to 0·34, RD −14·3%, −15·9 to −10·7; low certainty), with stronger associations with N95 or similar respirators compared with disposable surgical masks or similar (eg, reusable 12–16-layer cotton masks; pinteraction=0·090; posterior probability >95%, low certainty). Eye protection also was associated with less infection (n=3713; aOR 0·22, 95% CI 0·12 to 0·39, RD −10·6%, 95% CI −12·5 to −7·7; low certainty). Unadjusted studies and subgroup and sensitivity analyses showed similar findings.

These certainty ratings come from GRADE. https://bestpractice.bmj.com/info/toolkit/learn-ebm/what-is-...

Low certainty means "The true effect might be markedly different from the estimated effect".

This doesn't feel like lying. It feels like making the best of a bunch of not very good data.

If it was so well established that mask wearing's effectiveness has low certainty, then the following tweet from the US Surgeon General was definitely a lie:

"Seriously people – STOP BUYING MASKS! They are NOT effective in preventing general public from catching #Coronavirus, but if healthcare providers can’t get them to care for sick patients, it puts them and our communities at risk!"

Effectiveness doesn't just mean "can this mask trap particles?" -- because it's clear that FFP3 masks can do this. Effectiveness also means "In a particular situation with a thousand people, how many if them will be infected if none of them wear a mask, vs if all of them wear a mask?"

Healthcare professionals have to shave facial hair; they use good quality masks that are tested for conformity; they have sessions where their masks are fitted for them; they have training about how to put on, take off, and wear the mask; they have support to help them put on PPE; and they use the mask as part of a bundle of PPE that includes gowns, gloves, hand-sanitising, and eye protection.

In a healthcare situation we see that this package of measures does reduce infection. (And then we struggle to understand what's doing what because it's all a big confounded mess)

But for the public, going about day to day activity, it was really hard to see that masks would do much good.

And it's hard to do that even today: people in this thread are saying that it's obvious cloth masks do nothing but p95s are super effective. No-one can post any good quality evidence for that. All we get a some pretty rough studies that are trying to disentangle a bunch of measures to see which had most effect.

What we know is that:

* Cloth masks reduce viral loads by somewhere around 30-70%.

* High-quality masks reduce them by e.g. 70-95% depending on how they're fitted.

* We know risk of infection (and severity) is related to viral load, but we don't quite know how.

* We have a lot of well-documented evidence from Asia about reduced infections when high-quality masks are used consistently (e.g. buses in China, hospitals in Singapore, etc.).

* We know masks significantly reduce R0, but we don't know by how much; error bars are huge, and dependent on a slew of other variables.

The key difference between public health settings and hospital settings is opportunities to be infected. As a doctor, I might have 10 opportunities per day, and if I don't have properly-fitted PPE one of those times, and I'm susceptible, I'll catch COVID19. This means half-measures do very little.

In a public health setting, it's a numbers game:

* Personal: If I have 1 opportunity per 2 weeks to be infected, and I reduce odds by 50%, I'm half as likely to catch COVID personally.

* Public health: If I reduce infection odds by 50%, R0 is cut in half. That's huge. Our exponent is very different.

So cloth masks are definitely worth the $2, but won't stop you from catching COVID if you take other risks. If everyone wore N95 or even surgical or nanofiber masks, COVID would probably be gone pretty quickly.

This is disingenuous. The study you cite has a large effect size with low certainty ("Face mask use could result in a large reduction in risk of infection"). Best available data was that they probably help a lot, but we don't know yet, based on data from COVID19.

At the time COVID19 broke out, we had a ton of data from flu viruses and masks. That data clearly showed masks reduced viral loads a lot.

Was it perfect data? No data is perfect, especially 3 months into a new infection. But when the anti-mask recommendations came out, they were lying.

We did have a ton of data from flu viruses and masks, and those studies struggled to show any benefit to mask wearing.

You only see clear benefits when you reduce the quality of the research.

> But when the anti-mask recommendations came out, they were lying.

You keep saying this, but you've never posted the link to the meta analysis or RCTs that you think they should have been using to form their opinion.

Please show any study which "struggled to show any benefit."

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3591312/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5153448/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4699551/

Etc.

Studies range in effect sizes, but all the ones I read in March 2020 showed significant reductions in flu infection rates with masks. There were nice studies on cloth versus surgical versus N95 at the time too (summary: surgical and N95 were similar in performance, and much better than cloth, when used without full hazmat).

We know a lot more, specific to COVID, today. But that was best available evidence in March.

> If everyone wears N95 masks, the virus basically doesn't seem to spread.

Most of the very many dead healthcare professionals were wearing N95s that had been properly fitted.

1. I haven't seen that idea claimed anywhere. 'Most','very many'???

2. There was a great shortage of N95s for hospitals for a long time. (Let alone 'nursing'-home death-traps, where 33% of US deaths - 179,000 - have occurred. [0]) Sterilization methods took months to emerge.

3. Consider how dense viruses get in IC units, compared to retail locations.

[0] https://www.nytimes.com/interactive/2020/us/coronavirus-nurs...

In the UK at least I don't think that's true.

I saw an article the other day saying they'd given the FFP3 (out N95 equivalent) masks to the nurses in the ICU where patients are on ventilators, but the nurses dealing with the less sick patients just had surgical masks. Unfortunately, the less sick ones were up and about, coughing everywhere, not connected to machines to regulate their breathing and so those coughs are going into the air. Sadly lots of NHS staff have died.

These "studies" are stupid. Airflow is the [likely only] thing that matters. Social distancing in a 0 airflow environment isn't going to work, as droplets stay suspended. Masks prevent large droplet spread, but aren't fully effective (which is well known).

What is effective is physics. It's impossible for a droplets to spread upwind, and often downwind as well, since they don't hang around in suspension.

Get airflow certifications/requirements updated would not only help small businesses recover quicker but also pay dividends during normal flu season, let alone a pandemic.

This is the first that I’m reading about “airflow certifications/requirements”.

A brief search led me to this article explaining the concept of HVAC balancing. [0]

[0] https://www.primexvents.com/certification-matters-balancing-...

Actual article title: “A guideline to limit indoor airborne transmission of COVID-19”

The actual article doesn’t seem to have the spin described in the title.

You cannot get an infectious disease from someone who is not there. Stay home if you can.

Social distancing is good, even if it doesn't help with covid. In the past year, I've seen people respect personal space a bit more. For example, while shopping. That is a positive thing in my book.

"Study finds not driving a car much safer than driving a car", lets all stop driving and walk everywhere

well that's unironically a good idea and it's why many of the most livable cities on earth aren't designed around cars, let's do it like Pontevedra, reinvigorated small business, improved air quality, noise pollution and people's level of physical activity:

https://www.theguardian.com/cities/2018/sep/18/paradise-life...

Yes! Best idea on here. Cars do so much damage to humans, cities and the planet.

Reply on news.ycombinator.com