Thread by @Jabaluck

To frame this as "mask advocates" vs. "vaccine advocates" is to thoroughly miss the point.

Everyone on earth with access to vaccines should get vaccinated.

Masks are also a powerful tool against COVID, and you've misunderstood what our study says about them.
@MartinKuldorff

https://twitter.com/MartinKulldorff/status/1433202651106201607

Firstly, our study does not say that masks can only prevent 11% of COVID. Our study says that our intervention -- which raised surgical mask-wearing from 13% to 43% -- prevented 11% of COVID cases, and 35% among age 60+.

To put the point on your own terms -- if you vaccinated 30% of the population, would you prevent 35% of cases age 60+?

I am not sure. In an apples to apples comparison, I suspect vaccines are more effective than masks. When available, vaccines are also lower cost (I'd rather get a jab than wear a mask for a year).

But we don't have to decide to use either masks or vaccines. Vaccines are enormously cost-effective and every dollar invested in getting more people vaccinated has enormous returns.

In many settings, masks are also enormously cost effective given status quo vaccination levels, especially where we still see cases growing and large numbers of people dying from COVID (including settings where vaccines are not available).

Secondly, you fixate on the fact that the p-value for our primary outcome is p = 0.043. And you suggest, given this, maybe our intervention did nothing.

This would be rhetorically irresponsible framing if you hadn't read the paper and only looked at a single regression result (which is what I suspect happened), but given the full set of results, it is indefensible.

In the very next table, we see that the impact on WHO-Defined COVID-19 symptoms has a p-value of 0.000. Why is the p-value larger for symptomatic seropositivity? Because we don't have blood test results for most of the symptomatic individuals (yes, we checked balance).

If you look further at the breakdown of results, you see a consistent pattern which is extremely unlikely to have arisen by chance.

Surgical masks tend to have bigger impacts than cloth masks, as you would expect given their higher filtration efficiency, and they are more impactful at older ages.

There are theoretical reasons we might see bigger effects at older ages. For example, the elderly tend to have fewer contacts, meaning that reducing the probability that any given contact transmits reduces the likelihood of having any transmitting contact by more (ht: @VC31415).

@VC31415 In summary, if you read our paper and your takeaway is that this is a disappointing result for masks rather than, "This is a cost-effective way to prevent many deaths", then you have misread the paper and you are not thinking about policy in a coherent way.

Tagging @MartinKulldorff correctly.