Thread by @ProfDFrancis

Both John Mandrola and Vinay Prasad are angry about this study in JAMA.

Source: jamanetwork.com/journals/jama/…

I thought I might join in and give the study a bashing.

Those are indeed the claims against masking.

The problem facing the not-very-bright opponents of masking is NOT that they are not good at distinguishing TRUE from FALSE.

Their problem is being poor at QUANTIFYING the importance of things. This is what we do in science.

All TRUE:

"Uncomfortable"
"Loses a human connection"
"Unpleasant if exerting"

And that is indeed why we have evolved faces, rather than masks, to go on the front of our head.

Probably UNTRUE:

"Does not slow spread of virus."

If the virus were mobile like a gas molecule, the allegation would be true: no slowing of spread.

But it floats in liquid water, not air or water vapour.

SOME of that water, if even a little, is captured in the mask. So viral spread is slowed, at least a little.

TRUE, but to a trivial extent:

Lowers O2 sats
Increases pCO2

Here is my thought experiment to prove it without getting out of your armchair.

IMAGINE you did an experiment of putting 1000 facemasks on a person's face.

Like this, but for facemasks:

With a 1000 facemasks on the face, and the face as well covered as this person's torso is ...

... what is your best estimate of would likely happen, ON AVERAGE in a large pool of subjects, say n=100?

Effect on pO2

And effect on pCO2?

Hopefully you all agree that the pO2 will go down.

A bit or a lot, I have no idea which is more likely, but I think it is inevitable that the result of such a mega trial of mega masks is an decrease in pO2.

Likewise the pCO2 would go up a bit.

This is all because of an increase in anatomical dead space and work of breathing etc.

Obviously people would compensate with hyperpnoea, but it would never be quite enough to neutralise the pO2 and pCO2 changes.

So 1000 masks would affect pO2 and pCO2, at least a little bit, if measured across a large group and averaged to minimise noise?

What about 1 mask, instead of 1000?

Because if you say "No effect when you just put one on", I am going to ask,

"How about when you add the second?"

EVENTUALLY there is gonna be an effect.

Will that effect

So now, in retrospect, I hope everyone agrees that it is impossible for a single mask to have NO effect on pO2 and pCO2.

What we mean as doctors when we say "no effect" is "no effect beyond ridiculously trivial to an extent so small that it is vastly overwhelmed by random noise in any individual person".

The latter part is taken as obvious by all intelligent people.

What we do instinctively, although we don't verbalise it, is estimate that the effect is ludicrously trivial, and therefore "for practical purposes 0", and stop thinking about it.

However some people do not have that level of sophistication.

It's a pretty undemanding level of sophistication, but some people don't have it.

Some are hopelessly dim, but others are "floating voters" of mask fear. This paper may tip the balance for a few of them.

Deciding what to worry about, or what to think about, isn't easy.

People don't teach us that at school (correct me if I am wrong).

We have to work it out for ourselves. I think it is not easy to express how to do it.

Masks upset pO2 and pCO2 in the same way that mobile phone radio waves heat my brain. It must be true, a tiny little bit, but not to an extent that I would bother to think about for more than 1 nanosecond.

I just add it to the vast list of things that I don't bother to worry about.

Pushing things onto that list is the way that sensible people trim down the things that they DO worry about.

It's the same as long-covid myocarditis. Theoretically possible, but way off the bottom of MY PERSONAL list of things to worry about.

I'm glad that others do the studies, so I don't have to: I think it would be terribly dull, but that's why it is them and not me who are doing it

Likewise the putting-the-mask-on-some-guy study in JAMA is a study that neither I nor John nor Vinay would put effort into doing, but we should be happy that someone is doing it, and doing it properly to a high scientific standard.

It confirms our suspicion, but is not useless.

Vinay is also technically correct in all of these criticisms.

In particular he is right to point out not everyone is persuadable by science. But I think there is a spectrum.

* Accept scientific approach
* Floating voter
* Trumpist - mind made up, reject data against beliefs

I would argue that the "floating voter" is a useful group to try to help. Think of it like Triage in a battlefield.

> Gonna be fine, do not spend resources
> Badly injured, worth treating
> Too badly injured, gonna die anyway

Something like this (although it has 4 branches)

And yes, the trial did not include people "who were unable to wear a mask for medical reasons".

Maybe they had an O2 mask already on them, or a terrible painful rash on the face etc?

If they couldn't wear a mask, then it is reasonable (indeed essential) to exclude them.

It won't satisfy stupid people, who will say it "only studied people agreed to be studied".

No intelligent person can complain that an RCT only studies people who agree to be studied, unless you are some sort of Nazi concentration camp doctor.

So the people who are complaining are just unintelligent.

People can complain this, as they have just done a few minutes ago.

The confidence with which they claim that it has had no impact is an index of hour poorly they understand the scientific process of experimentation, i.e. having a control arm.

Speaking of the scientific method, I hope everyone has kept up to date with our little Live Quizorama on elementary statistics concepts for medics?

Click→inspirion.org/sm/aca~-MDQ9cX…

Currently people are scoring ~80% which is great! A couple of questions tripping people up though