Comment by timr | Hacker News Reader

Comment by timr | Hacker News Reader

Parent Full thread

We have no good evidence for this claim. The only studies involving n95 masks and respiratory illness were conducted in hospital settings (i.e. don't reflect "normal" life), have high risk of bias (i.e. were small and were not randomized, controlled trials), and have shown mixed results.

This is the WHO review that summarized the existing evidence for masks at the start of the pandemic. There were 4 studies on n95 masks in hospitals:

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

Additionally, this is the most comprehensive review of mask literature I am aware of. It explicitly states that:

> At least ten studies evaluate the clinical efficacy of different types of masks compared to one another, but without a no-mask control group most provide little insight into mask efficacy.

> Four RCTs, four meta-analyses, and one prospective cohort study found surgical masks were non-inferior to N95s for protection against respiratory infections, and one found evidence that N95s provide greater protection than medical masks against self-reported clinical respiratory illness but not ILI. However, a recent review found that evidence that N95s protect healthcare workers from clinical respiratory infections at all is “low quality". One meta-analysis of particular note, an April 2020 preprint of a Cochrane review of clinical evidence for both surgical and N95 masks, “did not find any differences in the clinical effectiveness of either type of mask in the setting of respiratory viral infection transmission to healthcare workers,” although the review’s final November version omitted this language.

https://www.cato.org/sites/cato.org/files/2021-11/working-pa...

We simply don't know if n95 masks provide any protection -- personal or otherwise -- in the real world. That is the only honest answer to the parent's question. Everyone who claims otherwise is extrapolating from laboratory experiments, low-quality observational data, or (this is probably the most common) hearsay.

This is like saying we don't know if washing your hands before preparing food helps hygeine in the real world there have been no good studies.

1. Of course it does, it doesn't need to be studied;

2. Yes fine sure other behaviour may negate any benefit.

The question was about the performance of the masks, not whether the average person will grab a bus rail then get home whip off the mask and pick their nose before washing their hands.

To which the answer is simple, N95/FFP3 rates particle filtering effectiveness at a certain size; coronavirus particles are within that window.

> Of course it does, it doesn't need to be studied;

No, public health mandates are not something you derive from first principles, but from actual observational data about the effectiveness of interventions.

It's like the difference between a drug being effective in vitro rather than in vivo. There are many confounding factors and side effects that need to be understood, so it is irresponsible to skip directly from the in vitro study to mass distribution of an intervention.

To help illustrate this point, let's assume that wearing the N95 masks reduces the likelihood of getting infected by some fraction, but not enough to drive the reproduction rate down below 1 (which is certainly the case). Say you only get from 2.0 to 1.8. Therefore unless this difference is the key to overwhelming hospitals, it should make no difference to death rates. However:

Those who have lung disease (14% of the population) are less likely to wear these masks because it makes breathing more difficult for them, so they will become differentially more likely to catch the disease as a result of the healthy population wearing the masks than if no one was wearing the masks. But since herd immunity will be reached anyway (just at a later time), this intervention may result in an increase in deaths as the population of infected people will have a greater share of those with lung disease than if no one wore the masks.

Now, is the scenario I laid out true? Well, we don't really know what herd immunity is here, and we don't know whether differential rates of compliance are big enough, which is why you have to study it. You always have to study. You don't go from a thought experiment to an intervention without study, because there are always confounding factors that can reverse your thought experiment once the intervention is applied in the real world.

This is actually historically interesting, because it was the ineffectiveness of masks during the 1918 flu epidemic that was one of the major findings of public health, distinguishing it as a separate field from immunology. E.g. everyone assumed you needed to wear masks, but ex-post studies showed this made no difference for death rates in the 1918 pandemic.

Now, I'm not saying that the 1918 data is decisive here, because they didn't have N95 masks then, but still use of the masks they did have was considered a best practice by immunologists, so it was still a shocking result how something that worked in the thought experiment could fail when deployed as an intervention. All of a sudden, people began to realize that "Yes, you need to study these things. First principles and logical deduction is not enough." That realization that public health is not immunology plus Aristotelian logic but is instead its own field -- that is the bedrock axiom of public health.

The question wasn't about public health mandates.

> We simply don't know if n95 masks provide any protection -- personal or otherwise -- in the real world.

That is simply not true.

First, doctors and researchers consistently say that N95s are known to provide protection against Covid.

Second, this is the first google result on that topic:

https://www.cdc.gov/mmwr/volumes/71/wr/mm7106e1.htm

Quoting from that CDC page:

> What is already known about this topic?

> Face masks or respirators (N95/KN95s) effectively filter virus-sized particles in laboratory settings. The real-world effectiveness of face coverings to prevent acquisition of SARS-CoV-2 infection has not been widely studied.

> What is added by this report?

> Consistent use of a face mask or respirator in indoor public settings was associated with lower odds of a positive SARS-CoV-2 test result (adjusted odds ratio = 0.44). Use of respirators with higher filtration capacity was associated with the most protection, compared with no mask use.

> What are the implications for public health practice?

> In addition to being up to date with recommended COVID-19 vaccinations, consistently wearing a comfortable, well-fitting face mask or respirator in indoor public settings protects against acquisition of SARS-CoV-2 infection; a respirator offers the best protection.

You can say that that might be just correlation, but as those quotes mention, we do know how N95s work: they filter viral particles. That is easy to prove in lab settings. Real-world settings involve more factors, like how well-fitting the mask is, but it would be very surprising if a good N95 properly used did not offer any protection!

N95 masks work. It's a good idea to wear one when relevant.

> That is simply not true.

> First, doctors and researchers consistently say that N95s are known to provide protection against Covid.

I just provided you with the sum total of medical evidence supporting the claim. Citing unspecified "doctors and researchers" is not a rebuttal. You are repeating hearsay, which was exactly my point.

Also, you should know that that MMWR study you've chosen to emphasize here was widely panned by scientists on both sides of the mask debate. It's based on a telephone survey with a 13% response rate. The statistics are hopelessly biased due to the test-seeking behavior of the study participants. It is not credible:

https://vinayprasadmdmph.substack.com/p/mask-studies-reach-a...

It's good that you pointed it out, however, because it's a perfect illustration of the abysmal level of "evidence" that has been used to support similar claims throughout the pandemic. Unfortunately, a many terrible "papers" were published and widely distributed on masks, including many in "good" journals. If you don't understand how to appraise medical evidence, you will be easily misled.

> N95 masks work. It's a good idea to wear one when relevant.

Unfortunately, there is no good evidence supporting this assertion, and "when relevant" is essentially unknown.

It's not unspecified "doctors and researchers". It's the overwheming consensus among medical professionals. The CDC recommends masks, as I linked. So does the WHO,

https://www.who.int/news-room/questions-and-answers/item/cor...

And so does the European CDC,

https://www.ecdc.europa.eu/en/publications-data/using-face-m...

etc. etc. The picture is very clear here.

And, again, the lab evidence is incredibly supportive. We know that N95s filter viral particles. It's possible in theory that that doesn't end up mattering - say if people never wear them right, or if they end up taking more risks because they think they are safer - but that seems unlikely when other evidence says otherwise, and the experts concur.

> The picture is very clear here.

I wish it were.

It should not be difficult to verify that N95 masks filter COVID virus particles effectively.

If such a study had been performed, nobody could reasonably hypothesize that, owing to the properties of the viral envelope, an N95 filter could stop nanometer black carbon particles but not COVID viruses of the same size.

It is baffling that the very basic experiment has not been done: DOES N95 FILTER COVID?

The advancement of human knowledge absolutely depends on VERIFYING assumptions. Institutionalizing common beliefs without empirical data is a crime against science.

Why are you certain such studies have not been done? Plenty of things have been published in that space. For example,

https://academic.oup.com/jid/article/226/2/199/6582941

https://www.frontiersin.org/articles/10.3389/fmed.2021.65431...

You might respond with some specific skepticism about some particular detail in either study. But e.g. the first is in the respected Journal of Infectious Diseases. It passed peer review. And plenty of other studies exist.

The bigger picture here is that every major health agency in the world recommends good masks such as N95s, as mentioned earlier. That is the standard for relevant healthcare workers for good reason. The experts in those agencies have looked at the studies and made that conclusion, which is the general consensus in the healthcare field.

There is some chance all the experts are wrong on this. That happens sometimes. But it's rare.

I gave you a comprehensive review of all literature related to masks, used on people, to prevent actual illness. You don't like the conclusion of that data (there's little clinical evidence supporting their use), but digging up irrelevant laboratory papers doesn't change it.

The first link here is a lab study that doesn't demonstrate anything of relevance to the question: fit-tested masks, in a laboratory, where the outcome is "virus in the air" doesn't tell you if same masks, in the real world will protect against illness in people.

Second link is more of the same: demonstrating that a mask can filter particles, in an ideal setting, does not tell you anything about what actually happens in the real world. For that, you need clinical evidence (which I've already shown you is weak).

Medical science is filled with stories of things that work in theory, but don't work in practice. Literally every drug that failed a clinical trial (i.e. >99% of all drugs tested) worked in the lab. You must do controlled tests on real people to see what really works.

> the first is in the respected Journal of Infectious Diseases. It passed peer review.

Journal of Infectious Diseases is actually not a great journal, and "peer review" is not a magical talisman that prevents publication of bad research. See also: citing the WHO, the CDC, etc., when you've been shown that the evidence for the recommendations they're making is not robust.

> It's the overwheming consensus among medical professionals. The CDC recommends masks, as I linked.

"Consensus", without evidence, is just opinion. Tragically for the credibility of science-based policy, the evidence backing these mask recommendations is incredibly weak. That's just a fact. Repeating to me that "scientists say" something-or-other to be true does not change the fact that the evidence backing the claim is poor.

Medical science is riddled with examples of interventions that "medical professionals" recommended without evidence, only to find that they were ineffective -- or worse, harmful -- when a controlled study was finally performed. Just a few recent examples: hormone replacement therapy; aspirin for heart disease; widespread opioid use for pain; mammograms under age 50. All were "overwhelming consensus". All have been since found to be useless. Some have led to tragic outcomes.

The CDC and the WHO could have run randomized controlled trials to find out of n95 masks are effective at preventing illness. They did not.

nonsense, N95s are highly effective and if you want to make a good faith argument that they're not try citing the new england journal of medicine and not the fucking cato institute, good lord

Why do surgeons wear masks?

Not an expert, but as I understand it, surgeons could be exposed to much more than airborne viruses (bacteria, bodily fluids, etc). So masks would be worn for more possible reasons. It’s a good question though, I hope someone knowledgeable answers it.

Surgical masks and face shields are indeed designed specifically to protect from fluids. That is very important too.

Separately from that, when there is a risk of airborne transmission as there is with Covid, doctors and nurses will wear appropriate masks like N95s.

Reply on news.ycombinator.com