There are two main routes for catching flu. One is via your hands and your eyes and nose. Your hands pick up germs from the surfaces you touch, and when you touch your eyes or nose passages, the germs infect you. This was thought to be the main route for infection, and I still think it is. I’d been pushing iodine hand sanitizer for some time, the stuff used in hospitals, saying that that the alcohol hand sanitizer doesn’t work well, that it evaporates.

The other route, the one touted by the press these days is via direct cough droplets, breathing them in or getting them in your eyes. Masks and face shields are the preferred protection from this route, and the claim is that masks will stop 63% of the spread. The 63% number has an interesting history, it comes from this test with infected hamsters. Hamsters are 63% less likely to infect other hamsters when they wear a mask. Of course, the comparison has some weaknesses: hamsters don’t put their fingers in their noses, nor do they rub their eyes with their hands, and hamsters can be forced to keep the mask barrier all the time — read the study to see how.

A more realistic study, or more relevant to people, in my opinion showed a far lower effect for masks, about 20%. During the HiNi flu pandemic of 2009 a group of 1437 college students at a single university were divided into three randomized groups, see the original report here. Students at a few chosen residence halls were instructed to wash their hands regularly, use sanitizer, and wear masks. Students at other halls were either told to wear masks only, or told to go on as they pleased. This was the largest group, the control. They included students of the the largest residence hall on campus. The main results appear as the graph below, Figure 1 of the report. It shows a difference of 6% or 20%, depending on how you look at things, with the mask plus hand-health group, MPHH, doing the best.

After 6 weeks of monitoring, approximately 36% of the control group had gotten the flu or some collection of flu symptoms. The remaining 64% of the residents remained symptom free. This is he darkest line above.

Of the FM Only group, the medium line above, those instructed to wear face masks only. 30% of this group showed flu symptoms, with 70% remaining symptom free. Clearly masks do help with humans, but far less than what you’d expect from the news reports.

The group that did best was FMHH, the group who both wore facemarks and used hand health, regular hand washing plus hand sanitizer. This group reported an average of 3.5 hours per day of mask use above the control group average. This is about as good or better than I see in Michigan. Adding the hand health provided an additional 1% improvement, or a 3% improvement, depending on how you look at these things. The press claims hand health is wasted effort, but I’m not so sure. I argue that the effect was significant, and that the hand sanitizer was bad. I argue that iodine hand wash would have done better at far less social cost.

I also note that doing nothing was not that much worse than mask use. This matches with the observation of COVID-19 in Sweden. With no enforced social distancing, Sweden did about the same as Michigan — slightly better, despite Michigan closing the schools and restaurants, and imposing some of the toughest requirements for social distancing and mask use.

There were other observations from the university study that i found isignificant. There are racial differences and social differences. The authors didn’t highlight these, but they are at least as large as the effect of mask use. Asians got the flu only 70% as often as others, while black students got it 8% more often. This matches what has been seen in the US with COVID-19. Also interesting, those with a recent flu shot got flu more often; those with physical activity 13% more often. Smokers got the flu less than non-smokers and women got it 22% more often than men. The last two are the reverse with COVID-19. I could speculate on the reasons, but clearly there is a lot going on.

Why did Asians do better than others? Perhaps Asians have had prior exposure to some similar virus, and are thus slightly immune, or perhaps they used the masks more, being more socially acceptable. Why were smokers protected? It’s likely that smoke kills germs; was that the cause. These are speculations, and as for the rest I don’t know.

I am not that bothered that the students probably didn’t wear their masks 100% of the time. Better would be better, but even with mask use 100% of the time, there are other known routes that are almost impossible to remove: clothing, food, touching your face. I still think there is a big advantage to iodine hand wash, and I suspect we would be better off opening up a bit too.

Robert Buxbaum August 7, 2020.