To use a randomized trial to measure the effect on the spread of the disease, researchers would have to find a community, divide it randomly into two and cut off all interaction between the two groups. If the communities of mask-wearers interacted with non-mask-wearers, the non-mask-wearers could infect people who would otherwise be protected. The results would be uninterpretable.
There’s more: Scientists would also have to mandate cloth masks for one group, while forbidding cloth masks for the other. But when people see others wearing masks distributed by researchers, they will improvise and make some masks for themselves. Researchers would have to mandate and ensure that people in the mask group wore masks everywhere, while surveilling the other group to ensure they didn’t wear a mask at all. This is not only grossly unethical; it also is just isn’t possible in any experimental conditions I can imagine.
It gets thornier: This pathogen has a very skewed distribution in which super-spreading events account for much of the transmissions while many people — upward of 70 percent in multiple studies — barely transmit at all. That means luck plays an oversized role in how some places will have outbreaks just because of misfortune, while others will be spared. Such imbalanced phenomena are hard to nail down with just a few studies, even if truly randomized. Was it the mask, or serendipity, that prevented an individual from spreading the disease? For such a skewed phenomenon, we’d need many, many studies to establish causality with confidence.
What about preventing the mask-wearer from being infected? That secondary benefit of mask-wearing is definitely measurable in randomized trials, because the mask-wearers in the trial would be the beneficiaries. We already have substantial evidence for that, which is why health care workers wear masks.
In fact, that’s where the study Senator Paul and others have cited (or mis-cited) comes into play.
First, it didn’t compare the effectiveness of cloth masks with the effectiveness of masklessness. It was a study in a health care setting, where preventing the workers from getting infected is a priority, and the “control arm,” the comparison group, was mostly or completely wearing surgical masks. Since it would be unethical, obviously, to prohibit health workers from wearing masks, there was no “no masks” group.
The study merely examined whether cloth masks were an acceptable substitute when surgical masks were scarce. Unsurprisingly, they found that surgical masks — which are generally used once and then disposed of — are much better than cloth masks.
If we can’t feasibly do randomized studies of how well masks protect the community, are we totally without evidence? We are not.