[timr]

> We cannot do the science, because such a randomized test would be deeply unethical.

Thank you for perfectly illustrating the problem. I was actually sort of worried that people wouldn't take me seriously when I said that science has become so political that we actually can't do any. It sounds conspiratorial, doesn't it? But, alas:

1) I post results of RCTs showing that masks do little, if anything.

2) Someone replies that the evidence is insufficient for reasons X, Y, Z.

3) I say "great, let's gather more evidence to resolve those uncertainties".

4) Someone replies "we can't, it's unethical".

Quite a fine castle you've built on that cloud, good sir. Very safe indeed.

[Terretta]

And yet, it’s been done:

In a First, Randomized Study Shows That Masks Reduce COVID-19 Infections

A large study co-authored by Yale SOM’s Jason Abaluck and Mushfiq Mobarak tested the effectiveness of a mask-promotion program in Bangladesh in increasing mask use and preventing symptomatic infections. The study found that masks significantly lower symptomatic infections, especially among older people and when surgical masks [instead of cloth] are used.

https://insights.som.yale.edu/insights/in-first-randomized-s...

[timr]

That's literally the study this thread is talking about. It's great that they did it -- it should have been done in early 2020, all over the world. We should have dozens of other studies just like it.

The fact that we don't is an indication of how hard it's been to pursue any sort of science in this area. Almost nobody funds it, you can't get it published if it doesn't fit the public health narrative, and even if you do, the the news media won't report on it, and it might just get censored on social media.

[angelzen]

Make that zero. To their credit, the authors of the Bangladesh mask study released the raw data. "The difference between the two groups was small: only 20 cases out of over 340,000 individuals over a span of 8 weeks." Drawing any conclusions from such small amount of data is deeply unserious.

http://www.argmin.net/2021/11/23/mask-rct-revisited

[cthalupa]

It isn't a political issue to not do studies that require doing things that actively and intentionally put people at things we know are highly likely to put them at risk of physical harm. The world decided this was unethical long before the current political climate.

[angelzen]

In masks + covid context it feels like "unethical" is used as a thought terminating cliche.

* The pandemic is affecting 8,000,000,000 people all around the world. A large size RCT enrolls X0,000 participants. For example, the Pfizer covid mRNA vaccine clinical trial had 21,728 placebo participants. For every RCT participant, the RCT results are going to inform the course of action for hundreds of thousands of people. This is to say that such RCTs are extremely valuable given the high infectiousness of the virus.

* The risk to the participants is at worst a moderate increase in the chance of being infected with covid. The covid fatality risk to a healthy adult is small. In the US there have been about 50,000 deaths with covid in <50yo age group. An RCT with 20,000 <50yo people on the placebo arm may see about 3 covid deaths assuming all the placebo participants are going to be infected. Realistically, only a fraction of participants are going to actually be infected with covid, thus there is a good chance every single one of the RCT participants will continue their lives just fine, especially if the study designers are careful to not include people with serious comorbidities.

* Vaccines are available to reduce the death risk by another order of magnitude if deemed necessary. Alas, while the vaccines have strong effects in preventing serious covid, they only have a middling effect on preventing infections [edit: after a few months]. The mask/no mask infection effect remains measurable.

* Covid is endemic. Everyone is at risk to to be infected with covid sooner or later. Wearing a mask may decrease the daily risk by a moderate margin, unfortunately integrated over many days the infection probability approaches 1. The RCT is merely speeding the risk by a moderate margin for the placebo arm participants.

* There is a large pool of potential volunteers that don't (want to) wear masks anyways. Adults have the right to volunteer for risky activities, including activities that may result in death.

[cthalupa]

This is a well written argument, and I don't disagree with the majority of it. However, there is one key point that I do disagree with, and makes the rest of it moot:

>* There is a large pool of potential volunteers that don't (want to) wear masks anyways. Adults have the right to volunteer for risky activities, including activities that may result in death.

This is certainly true. But the difference is that once we begin performing medical studies that ask this of people, the medical industry is now complicit in adults performing risky activities that may result in death and is asking people to do so, or if doing so, to do so for the benefit of medical research. This is a line that has not been crossed by the modern medical research industry as it has been long decided that this extra pressure, however small, is a lever they do not want to pull because it is fundamentally incompatible with 'Do no harm.'

It isn't 'do no harm, except a little when we think it might outweigh the downsides'

It's a line I don't think we should cross. I understand why someone would disagree with that.

[angelzen]

The null hypothesis is the natural state of humanity, without medical intervention.

[lern_too_spel]

The hypothesis for why masks work is due to the laboratory-measured reduction in transmission from the host. The only way to test this is to have a large population where most do not wear masks or most do wear masks. You will not find enough volunteer to make such a test possible. Individual mask wearing does not test this hypothesis.

[timr]

This is a fantastic comment. Thank you for breaking it down.

[GekkePrutser]

But what if those people volunteer?

[vishho]

I feel like I am going insane. Pure psychopathy or scientism to want randomized controls for studying infection of a deadly virus. Absolutely disgusted. No wonder we had a lab-leak...

[cedilla]

It is unethical. All your accusations of narrow-mindedness, but you can't see the obvious, glaring ethical problems with such a trial.

Not the least because it would be completely infeasible. You may pretend that we can "simply" do a "controlled experiment" but you ignore how impossible it is to tell any sizeable group of people to behave in a certain way and report honestly about it.

"Randomised control trials" are the platonic ideal but in the real world, you can't endanger people for your curiosity, and you will have greatest problems to actually enforce your test protocol. It's more than "tricky". Unless you have access to some spherical people in a vacuum.