The biggest flaw with the latest study is that it noted less than 50% compliance in mask wearing. Unless the results ignored those who failed to wear their masks when they had been advised to, then the results are always going to show limited benefits.
If those who didn't adhere to mask guidance were infected at the same rate as the control group but weren't excluded from the data that's going to change the numbers.
Let's say there's 4,000 in each group, 2.1% of non mask wearers are infected then that's 84 from the control group, but also 42 from the 50% who didn't wear masks when they were told to. That then means that to get to 72 (1.8% of the masked group) you need 30 from the other 50%. However if you look at the percentage infection rate for that (30/2000) then it falls to 1.5%. Not a big fall in it's own right, but probably enough to mean that statistically there's some protection even allowing for margin of error, even though it's just for those wearing a mask and the infection coming from others.
Although by excluding half your group by then not wearing masks you make your sample size quite a bit smaller.
Other limiting factors on the benefits:
- it appears that not everyone from a household were advised to wear a mask
- the study was looking for people who were out of their home for over 3 hours a day, with the average within the study being 4.5 hours, that would imply some going to work and mask wearing. Well if they are doing that but not in a well ventilated area then masks do have limited benefits after several hours in those settings.
- out of the home for over the hours, could include someone doing dog walking as a job and so maybe outside a lot with little personal contact with others, or a delivery driver in their own van. Dependent on how the groups were considered with such people it could result in skewing of the results
Having said that, it's likely to be that any such study is going to have problems with determining if masks do provide much protection.
Also whilst it should be accepted that there's a risk that there could be a small increase in risk from mask wearing, this is still within settings where people are likely to be exposed for hours at a time and so the benefit is likely to be small.
However the likelihood of all the studies to date all being "unlucky" with their data showing lower than expected figures for those wearing masks and higher for those not then it could be argued that there's starting to be a case that mask wearing at worse is unlikely to be any worse than not wearing one but with the potential for it providing (especially in settings where time is limited) some limited benefits. That is unless there's any studies which shows otherwise then risk in advising mask wearing is likely to be small and probably has some limited benefits.
The problem is that even within the body of evidence that masks are statistically insignificant in providing protection or that they do is that neither those who strongly oppose mask wearing our strongly support mask wearing there's no killer evidence to conclude either way. Mostly because to ensure that the risk of infection is high enough that there's got to be long duration of exposure.
As the risk to someone who goes to the shops once a week for 20 minutes is always going to be very low regardless of if they do or don't wear a mask, even if there's several other shoppers who are infected. Likewise working in an office with 3 other people for a 7 hour working day then wearing a mask is going to provide limited benefits of the air circulation is poor.
Therefore, whilst there's no evidence that those who support mask wearing can provide that mask wearing is 95% certain that it provides benefits, likewise there's no evidence that those who oppose masks which is 95% certain that it does cause harm.
As I've said before high and growing case numbers could be down to other factors and if mask wearing is brought in because cases are rising then the impact on total numbers is likely to be small but important to those who (if there's a small benefit) who would otherwise have died.
Yes other factors are likely to have a bigger impact (such as not mixing with several different people, working from home, closing schools, closing pubs and the like), however such interventions are likely to harm more people's mental health and/or the economy.
Unless there's evidence that there's 95% certainty that masks cause harm, then it's probably best to try mask wearing where there's the most chance of benefit (i.e. those inside settings where there's likely to be limited time duration) alongside keeping of the other best practice measures.