A new mask study has been getting a lot of attention. It tested thousands of people in Bangladeshi villages and came to two conclusions: (a) masks work, and (b) surgical masks work better than cloth masks. A 28 percentage point increase in surgical mask wearing within a village reduced COVID-19 incidence from 0.76% to 0.67%.
But there was also this odd result:
An increase in mask wearing within a village had zero effect on subjects under the age of 50. The entire effect is for those 50 and older. That deserves a whole lot of attention since it's such an unexpected result. What is it about older age that makes mask wearing so effective?
Generally speaking, I have to confess that I've given up on masks as a way of reducing COVID-19 spread. The people who hate masks are simply never going to give in on this, especially since it's a very public behavior and, it's worth acknowledging, a genuinely annoying one. Some people may not mind masks, but a lot of us, including me, find them uncomfortable.
I don't mind if other people want to continue going after mask policies, but I no longer think they have much chance of working. At this point, our energies are better used on vaccination and nothing else.
Kevin Drum-
Look at the study dates. Up until 4 of 2021 PRIOR to Delta taking off
What is unique about Delta?
It affects younger people and is more contagious
The study is totally understandable given the time frame
Your conclusion is totally understandable given the time frame - but wrong
One reason is that people over 50 are less active, and less likely to move vigorously to allow the mask to be leaky. They are probably less careful because they have less risk from Covid also.
Re the mu variant, it appears from actual cases being found, that it is not more transmissible than delta and should not be able to outcompete it, at least with current population. If mu is baseline less transmissible than delta, but is a bit more able to avoid immunity of one or both types ( and so is more transmissible than delta among those with immunity) , it might become more transmissible then delta eventually as our immunity rises and make the herd immunity threshold a touch higher. But that would not be a huge deal - it would just increase the relatively small number of annual endemic cases a little.
But, if it is also more deadly , it will eventually become dominant with the endemic cases and cause more death.
So if mu is a problem, it is one will see later when the pandemic recedes, not now.
OK, I have a number of thoughts re this study and will break it into a number of posts .
1) First, the issue of the huge difference in effects they found for over 50 vs others. THAT needs to be explained somehow before I can give a great deal of credence to the results.
Some of the possible explanations commenters have given above are explanations for absolute differences in getting symptomatic covid , not relative differences due to mask wearing. Thus are completely irrelevant . Others might explain a small difference but not the huge discrepancy showing almost no differences due to masks under age 50.
What possible process could cause masks to reduce cases so significantly over age 50 but almost zero below age 50?
The problem is none is provided that could explain that. But some possible biases might and that could mean there was no real effect over 50 either. Or could mean there was a larger effect for all ages, which then could be easier to explain by some physical process. I.e a difference of a 20 % effect and a 0% effect almost impossible. 40% vs 20% much easier.
2) but want to point out they did find a little reduction for lower ages too, just not statistically significant at 95% confidence. But many miss that this means nothing. Still means reality is more likely a positive effect than negative.
3) the actual reductions in cases found were not all that large. The reason they could show it is significant was the huge sample size. Having such a large sample means that you lower the chance that a small difference is caused by ransom luck. But you do not lower the chance that some other thing correlated with your variable caused it and not what you think
4) it is just flat incorrect to report that this study even purports to show that mask wearing reduced cases. It reports that the intervention they did which increased mask wearing reduced cases. What other things did that intervention do?
They report they also studied social distancing and it also increased in the study group vs. the control group .
That itself might be an important finding if it indicates that wearing masks does not have a counterproductive effect in making people feel safer and thus take more risks ( does not actually prove that as the education intervention might have increased social distancing even more just offsetting a decrease due to masks itself).
But why would they not have done a regression analysis with all the data measuring both social distancing and masking ( they had to differ across villages) to see how much of the reduction was masking and how much was increased social distancing? Maybe 100% of the reduction was due to the increased social distancing and 0% due to masks.
Why did they not attempt to find out? Makes me suspicious as to motives.
The idea of drinking bleach to cure covid is actually some kind of freakish international conspiracy,
https://news.yahoo.com/leader-movement-touting-toxic-bleach-081052771.html
5) OK, assume that what they find really is a 10% reduction in cases across the population. Ignore the age differences for now.
Masking could have two effects relevant for cases of symptomatic covid . One is to reduce the spread to others by reducing your chance of getting covid ( symptomatic or not ) and /or your chance of infecting another . This effect is the entire reason why masking might be a good thing for the community.
The other effect could be a reduction in symptoms for those who do get covid ( since they only looked for symptomatic covid). This only benefits the mask wearer and thus is not a reason to mandate masks.
But let us assume that only the first effect is relevant- reduction in spread.
The correct effect there to measure is NOT reduction in cases , it is reduction is infection rate- I.e. R . Any reduction in cases is only a result of the reduced R and the reduction in cases is only the result of the lower R by the exponent of number of covid spread cycles in the period measured.
If the period measured had around 10 spread cycles ( time from your infection to when you might infect someone else) a 1% reduction in spread rate will result in 10% fewer cases approximately. And 27% fewer with double the period. And increasing over time.
If everything stays the same.
6) again want to hit this point.
If everything stayed constant ( masking, immunity , etc. ) and masking caused just a 1% reduction in spread each cycle, and a cycle was a week, you would have a 41% reduction in cases in a year. Wow!
But one thing will not stay constant- immunity. If you are reducing cases then less natural immunity which increases R. Over time that should mean a higher spread rate in the masked group and eventually later higher cases.
Until you get to herd immunity and, unless the masked population stays masked forever, both the masked and unmasked population will end up with the SAME number of total cases , just that it will take longer for the masked group.
To the extent that masks reduce deaths in the end, can only do so by
a) spreading out the cases to avoid overwhelming health system and thus lowering the chance of death per case
b) giving time by delay to get vaccinated and thus substituting some vaccine immunity for natural ( now need fewer actual cases to get to herd immunity) and substituting a less serious breakthrough case for a unvacinated case )
c) by making it more likely if infected it will be less symptomatic by reducing viral load
7) in their analysis of reduced deaths and cost pet death, looks like they just took their reduction in cases numbers over the period and used that .
THAT IS STUPID. it shows to me they just do not get it and are either biased to show a result or just ignorant. And makes me leery to trust anything they find.
Dear lord, KD is turning into some kind of weirdly complacent "yeah whatever" columnist.
I think kevin may just be understanding the dynamics more than you.
Let me illustrate.
Assume a virus for which we are never going to have a vaccine. And the R is say 4. And the only effective measure is masks and they work superbly - reduce the chance to the wearer of getting covid by 50% ! And the death rate is 1% and not affected by hospital crowding etc. And immunity from getting it is 100% effective and never wanes.
Yes, I know all of that is not true here and I will deal with that . But just assume so for now to illustrate a point .
If 100% of people will wear masks for two years, but not for lifetime , what does that accomplish? Absolutely nothing.
Because the R only decreases initially from 4 to 2, a doubling every week. And that is going through the population rapidly anyway. Not as quick as with no masks, but you will still have to get to herd immunity. And, since masks come off in 2 years anyway, 75% of the population is going to get covid and .75% are going to die. And masking is not going to change that unless it is for the rest of everyone's lives.
Actually more as you will overshoot herd immunity by trailing cases after you get there. Masking can play a role there by timing- the point to have everyone wear masks is right when you get to herd immunity, not before. So as cases go up and up - no point to masks. As you see them start to peak, everyone mask and keep it on until cases very low.
And, in that scenario, who is masking does effect who gets covid. Who are the lucky few - the 25% or less who never get it.
And then those that do mask are the selfish ones helping themselves and hurting others because any reduction in them getting it is an inevitable increase in someone else getting it.
In this scenario , you should thank those who are not masking as they might be taking the bullet for you.
This is weird
4 days ago we were hearing about Mu from Israeli labs and UK health ministry
2 days ago I posted that a case had been found in LA county
yesterday Faucci said is was a concern........yet
Now this:
https://www.insider.com/who-is-tracking-covid-19-mu-variant-evade-vaccine-immunity-2021-9?utm_source=yahoo.com&utm_medium=referral
{snip}
Business Insider
A new COVID-19 variant called Mu that might be able to evade immunity from vaccines has been detected in almost every US state
Joshua Zitser
Sat, September 4, 2021, 11:18 AM
A map of where the Mu COVID-19 variant has not been detected
The Mu variant has not been detected in Nebraska, South Dakota, and Vermont. Data from: Outbreak.info
A new COVID-19 variant called Mu might be able to evade the immunity people get from vaccines, Insider reported.
The Mu variant has been detected in 47 US states and the District of Columbia, according to data from Outbreak.info.
Only Nebraska, Vermont, and South Dakota are yet to detect a case, the data says.
See more stories on Insider's business page.
A new variant of COVID-19, which the World Health Organization (WHO) says could have the ability to evade the immunity people get from vaccines and previous infections, has been detected in almost every US state, according to data from epidemiology and genomic database Outbreak.info.
Called Mu, the B.1.621 variant was first detected in Colombia in January this year. As of September 4, cases of the strain have been reported in 47 US states and the District of Columbia, Newsweek was first to report.
The only states without reported cases are Nebraska, Vermont, and South Dakota, according to Outbreak.info's data.
The strain is currently most prevalent in Alaska, where data suggests that 139 reported cases account for 4 percent of the total 3,837 sequenced samples.
California has the highest number in terms of raw numbers, the data show, with 232 reported cases of the B.1.621 variant out of 139,930 sequenced. This accounts for less than one percent of those sampled.
The Mu variant accounts for fewer than one percent of total COVID-19 cases, with the Delta variant dominant in the US.
Earlier this week, Mu was added to the WHO's "of interest" list of variants.
Anthony Fauci, President Joe Biden's chief medical officer, said Thursday that the Mu variant did not represent an "immediate threat" to the US.
"This variant has a constellation of mutations that suggests that it would evade certain antibodies, not only monoclonal antibodies but vaccine- and convalescent serum-induced antibodies," Fauci said during a COVID-19 press briefing on Thursday. "But there isn't a lot of clinical data to suggest that. It is mostly laboratory in-vitro data," he added.
{snip}
From nothing to worry about to 48 states with this variant? In the space of days?
Are we being snookered?
And think about this. They have only sequence tested a small portion of overall COVID cases - which COULD mean under reporting of this strain.
Vog, re your last paragraph, no, it should not.
Yes, if you mean we are " underreporting" the absolute numbers of the strain as they are not sequencing all of them. But in percentage of cases, which is all they are trying to measure, you will have sampling error just like in a political poll. But the percentage should not be that far off from reality and the bias can be either low or high. And they are sequencing enough to have a large sample at least in big states.
What would be something to watch is if you see it starting to increase as a %age, even if in low numbers. If ca is finding it in .2% one week and. 5% next week, that is troubling even if the numbers are still real low in absolute terms. That is especially so if the relative %age increase is not slowing. If it is, say going from. 2% to .5% then to .8%, 1.0%, etc. That could indicate it has some small niche where it has am advantage and will slow down once it exhausts that niche.
Have you seen anything showing how fast the relative %ages are changing yet? I did see one thing from uk seeming to show that mu % age was not growing at all.