Hacker News Reader

I'm not sure if I should interpret this as masks are an ineffective tool, or more narrowly that combined with other efforts it's unclear what the impact was since there's a large degree of opinion on the matter compared to data with a higher standard of a peer reviewed double blind study.

With masks being such a hot topic in certain regions, it would seem unreasonable to conclude that there is only correlation available.

The opinion I'm referring to and cited is in line with this one (just did a quick google search, I don't remember the original source of my assertion): https://www.sciencedaily.com/releases/2020/06/200624082657.h...

There appear to be a number of smaller studies available as well: https://en.wikipedia.org/wiki/Face_masks_during_the_COVID-19... which if cloth masks are effective it would seem to me to be reasonable to extend this to rate of adoption and a contributing factor.

Note that I've edited my comment to better reflect my thoughts (probably since you wrote this).

I don't think you should take this one paper as some kind of definitive rebuttal on masks. You should interpret it as a reasonable summary of current evidence, as well as a not-terrible analysis of US states that poses some serious challenges to the effectiveness of mask mandates.

That said, an editorial/letter in science daily is not a rebuttal, and wikipedia is useless for politically sensitive questions. There are many papers on both sides of the issue, and it's easy to cherry-pick a set that support any particular position.

Is it the end of the discussion? No. But It's not as cut-and-dried as people make it seem. The evidence base for masks is very low-quality. The issue became political, and as with all things political today, nuance and reason were tossed out the window.

> Regions with an early interest in face masks had milder COVID-19 epidemics, according to a new letter-to-the-editor published in the American Journal of Respiratory and Critical Care Medicine. (the source of the Science Daily article: https://www.atsjournals.org/doi/pdf/10.1164/rccm.202004-1188...)

I don't see how this is much less scientific than the non-peer-reviewed paper you posted.

A letter to an editor is not peer reviewed. It is a letter from a reader.

Both of the papers I linked to are either peer reviewed, or a pre-print in the process of review.

> a pre-print in the process of review.

So not reviewed yet.

My point is that both a preprint and a letter to the editors have not gone through a rigorous review. Preprints might be submitted for peer-review but that's no guarantee that it will be published in a reputable venue.

Equating a pre-print with a letter because they have both not gone through a review is like equating a skateboard with a car because they both have wheels.

Yes, both are text, written in the english language, shown on a webpage. Beyond that, they share no similarities.

If you think a pre-print is a better source than an academic letter (published in a reputable scientific journal, which implies at least editorial review), then lets agree to disagree.

A pre-print from legitimate researchers is a much higher standard of evidence than a letter. It's not really a matter of opinion.

If you cannot reliably judge the legitimacy of a pre-print in an area, then you aren't qualified to judge the legitimacy of any scientific comment in that area.

You keep calling it a "letter" but it is not just a "letter" in the sense that I can write a letter to you, it's a letter written by experts in the field to a top ranked scientific journal that has been editorially reviewed (by experts).

> If you cannot reliably judge the legitimacy of a pre-print in an area, then you aren't qualified to judge the legitimacy of any scientific comment in that area.

I agree, but now we are moving into the content and not just judging the book by its cover.

Trials on masks universally fail to take into account the field results. And that’s where they fall down. Mask effectiveness relies on purity - everybody doing things with military (or medical) precision.

The short of it is that people don’t stick rigidly to the effectiveness requirement. Hence why states and countries that stop compulsion don’t see any deviation in the rate of change.

Like HCQ it works in the lab, but not in real life.

What’s disturbing is why so many people still believe they do - and why they desperately want them to work - which moves into the political and religious (“I’m pure - it’s everybody else that isn’t”) rather than the scientific.

At this stage it’s become like carrying posies to ward off the Black Death.

> (and before anyone says it: yes, it's entirely possible to have conducted clinical trials of masks. a cluster randomized trial, for example, could have been used to establish efficacy early in the pandemic. That we didn't even consider doing these "because it's a pandemic" is one of the biggest scientific lost opportunities of our lifetime.)

There were studies done; all the way back in March 2020 there was an infographic making the rounds showing a rough estimate of percentage of particles filtered/let through by different types of cloth masks and N95/etc. If I remember right, all the cloth masks let through >50% of particles.

There have been a number of laboratory studies performed that show varying levels of effectiveness for different kinds of masks and materials (where "effectiveness" is defined in all sorts of different ways that may or may not be relevant). Prior to Covid-19, there was a low-quality evidence base that didn't provide much signal either way:

https://www.cebm.net/covid-19/masking-lack-of-evidence-with-...

Strapping a mask on a mannequin or to a nozzle is, of course, very different than asking people to wear them everywhere. For evidence of real-world effectiveness of a medical intervention, you do a controlled trial. We could have done some, but we did not. It is notable -- but not at all perfect -- that the only RCT conducted with masks and Covid-19 in 2020 found no significant personal protective effect:

https://www.acpjournals.org/doi/10.7326/M20-6817

It is a well-done study, and certainly limits the claims that can be made for masks, but doesn't answer every question. A good cluster RCT might have been able to address the question of whether or not masks protect others, for example.

This good cluster RCT is in the midst of being done.

https://www.nber.org/papers/w28734

In this paper they demonstrated an intervention in Bangladesh that increased mask usage by 30 percentage points (from ~10% to ~40%). In the follow-up paper they are doing random antibody testing between the two groups.

It probably won't be as satisfying for you, but in the US mask-wearing got split along political lines. You can compare red areas (haven't worn masks for months to a year) to blue areas (were still mostly wearing them up to this month when the CDC guidelines changed). Overall, no noticeable effect.

There are confounders though that make such data not useful (I'm not sure how much statistics can find data in the noise or not though - but it won't be trivial). Red areas are in general more rural and spread out, while blue areas are denser population. Thus blue areas are naturally a different risk level. Also masks and political stance are not a perfect correlation. I personally know democrats who don't wears masks and republicans who do.