I agreed that face masks reduce viral load, and I don't say face mask is completely useless but even in good controlled conditions, the efficacy is not much different but in real life, how many of people use face mask properly. Many people wear face mask improperly (touch/ adjust mask and face too often) as same as how they wash their hands.
Face touching: a frequent habit that has implications for hand hygiene. People tend to touch their faces 23 times per hour, on average.
There are more outliers in without-mask group, both droplets and aerosols. If we exclude outliers, two groups are nearly the same. In the
Supplementary Table 2, there are not many statistically differences of viral load, and maybe (just maybe) they are
false positives. The more hypotheses are tested, the higher odds for false positive. For example: if we test only one hypothesis, the chance of at least one false positive is 5% (if we choose the statistically significant threshold is at 95% or p-value = 0.05). However, if we test 2 hypotheses at the same time, the probability to have two true positives will be: 0.95*0.95 = 0.9025. It means, the probability of at least one false positive will be 1 - (0.95*0.95) = 0.0975 > 0.05. It will increase to about 10% (with 2 hypotheses), not only 5% (with only one hypothesis).
To reduce the probability of false positive, there are suggestions to apply stricter threshold of p-value to 0.005
[1] or 0.001
[2].
See:
If we apply the threshold of p-value at 0.005, there is no statistically significance from the above research.
From the table, we can see results are very inconsistency between groups. Honestly, I think results satisfy both sides, support or not support and this probably is a reason why governments have very different recommendations on face masks. One can say "Governments recommend don't wear masks if you are not infected and leave them for health care workers because the lack of face mask supplies". That statement is not completely true. The perspective on it was there before the pandemic, based on scientific results, I guess.
As I said, I am fine with face mask, just don't think it as a magical stuff that can protect you from the virus and don't pay attention on how you wear it.
I used some misleading words above, so I am sorry but my opinion still remains:
Even with minor statistically significance + many biases (bad practices, is.) ~ no difference in real life (not lab conditions).
I know the pandemic put all of us under uncertainty situation but I believe in science.
At the ends, within the uncertainty about face mask efficacy, we have to base on local government recommendations and personal choice to wear face masks or not. I am not in a position to make any recommendation here.
