HomeMy WebLinkAbout2023_03_16 SSchumacher_Cochrane reviewFrom:Stephen Schumacher
To:Board of Health
Cc:Board@olympus.net of County Commissioners " <JeffBoCC@co.jefferson.wa.us>, " Operations KPTZ "
<contactus@kptz.org>," Jefferson County
Subject:Rebuttal of false statements re Cochrane review of masking evidence
Date:Thursday, March 16, 2023 6:15:57 PM
________________________________
ALERT: BE CAUTIOUS This email originated outside the organization. Do not open attachments or click on linksif you are not expecting them.________________________________
Dear Board of Health,
Health Officer Berry made multiple falsestatements at today's Board of Health meetingregarding the gold-standard Cochrane reviewfinding no evidence of mask effectiveness forpreventing viral transmission. Shockingly, sheasserted it is "no longer ethical" to performrandomized controlled studies on the subject.
Berry's statements echoed an evidence-free NewYork Times opinion column "The Science Is ClearThat Masks Work", which is carefully refuted in the article below.
Yours truly,Stephen SchumacherPort Townsend, WA
===https://www.eugyppius.com/p/zeynep-tufekci-sociologist-and-noxious
New York Times columnist insists "The Science is
clear that masks work," fails after many words toactually locate any of this mysterious Science
eugyppius - Mar 12
Zeynep Tufekci is an academic sociologist, a NewYork Times columnist ... Her latest piece is anopinion column screaming against all reason andevidence that The Science Is Clear That Masks Work.https://www.nytimes.com/2023/03/10/opinion/masks-work-cochrane-study.html
Mostly, Tufekci is very mad about the Cochrane
mask review, which found no evidence that masksdo anything; and about statements by its leadauthor, Tom Jefferson, who has explained invarious interviews that indeed there is no goodevidence that masks do anything.
Tufekci is happy to report that - after what wecan only presume is extensive harassment bylegions of people like herself - Cochrane haveagreed to reword part of their summarised findings.
When you are facing absolute defeat, small victories become very important.
... Cochrane ... says that the way it summarized thereview was unclear and imprecise, and that theway some people interpreted it was wrong.
"Many commentators have claimed that a recentlyupdated Cochrane review shows that 'masks don'twork,' which is an inaccurate and misleadinginterpretation," Karla Soares-Weiser, the editorin chief of the Cochrane Library, said in a statement.
"The review examined whether interventions topromote mask wearing help to slow the spread ofrespiratory viruses," Soares-Weiser said, adding,"Given the limitations in the primary evidence,the review is not able to address the question ofwhether mask wearing itself reduces people's riskof contracting or spreading respiratory viruses."
She said that "this wording was open tomisinterpretation, for which we apologize," andthat Cochrane would revise the summary.
Soares-Weiser also said, though, that one of thelead authors of the review [Tom Jefferson] evenmore seriously misinterpreted its finding onmasks by saying in an interview that it proved"there is just no evidence that they make anydifference." In fact, Soares-Weiser said, "thatstatement is not an accurate representation of what the review found."
It is, in fact, an accurate representation ofwhat the review found, and it's also functionallyidentical to Soares-Weiser's much more verboserepresentation about "limitations in the primaryevidence" and those things which "the review isnot able to address," but all those syllablesgive the masktard faithful more places to seek solace.
While the review assessed 78 studies, only 10 ofthose focused on what happens when people wearmasks versus when they don't, and a further fivelooked at how effective different types of maskswere at blocking transmission, usually for healthcare workers. ... Of those 10 studies that lookedat masking, the two done since the start of theCovid pandemic both found that masks helped.
The calculations the review used to reach aconclusion were dominated by prepandemic studiesthat were not very informative about how wellmasks blocked the transmission of respiratory viruses.
As I've said before, and as all thinking peopleshould readily affirm, absence of evidence is
evidence of absence, particularly given the
massive publication bias favouring positive
results. Masking has been around in some form forover a century now, and there's been ampleopportunity to investigate how well masks work.Particularly since 2020, there's also beennear-infinite funding flowing to all mattersCovidian. The reason there aren't very manystudies on community masking is not that nobodyhas looked into it, but that nobody can find it does anything.
...[<https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0240287>I]n
one study of hajj pilgrims in Mecca, only 24.7percent of those assigned to wear masks reportedusing one daily, but not all the time (while 14.3percent in the no-mask group wore one anyway).The pilgrims then slept together, generally intents with 50 or 100 people. Not surprisingly,given there was little difference between the twogroups, researchers found no difference from maskwearing and declared their results "inconclusive."
In<https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0029744>another
prepandemic study, college students were asked towear masks for at least six hours a day while intheir dormitories, but they were not obligated towear them elsewhere. Researchers found nodifference in infection rates between those whowore masks and those who did not. The authorsnoted this might be because "the amount of timemasks were worn was not sufficient" obviously,college students also go to classes and socializewhere they may not wear masks.
Yet despite their inconclusiveness, the data fromjust these two studies accounted for roughly halfof the calculations for evaluating the impact ofmask wearing on transmission. The other sixprepandemic studies similarly suffered from lowmasking adherence, limited time wearing them and, often, small sample sizes.
I don't know why this is hard: Low qualitystudies failing to find evidence that masks workstill add up to no evidence that masks work.
The only prepandemic study reviewed by Cochranereporting<https://bmcinfectdis.biomedcentral.com/articles/10.1186/1471-2334-12-26>high
rates of mask adherence started during theworrying H1N1 season in 2009 in Germany, andfound mask wearing reduced spread if startedquickly after diagnosis and if a mask was wornconsistently (though its sample size, too, was small).
For "small" read 218 participants. Note also that
these influenza masking studies can be pretty
crazy. This one, for example, wanted all
participants in the masking arm "to wear[surgical] masks at all times when the indexpatient and/or any other household member withrespiratory symptoms were together in one roomwith healthy household members," and they furtherdemanded that these masks "be changed regularlyduring the day." These strict at-home maskingregimens have very little to do with Corona-eracommunity mask mandates, and they're also highlylikely to discourage social interaction ingeneral, so the effects are hard to disentangle.Despite all of that, the authors still found nosignificant difference in attack rates betweenthe masked and unmasked arms of their study. Thishappens all the time in these papers, and theauthors generally respond by rooting around inother variables to find a correlation somewhereelse. This time, they discovered that, of the 136participants in the masked arm, 60 who reportedmasking "no later than 36 hours after symptomonset" of the index patient enjoyed "a borderlinesignificant protective effect."
This is super high-level scientific work here. Ifeel super confident wearing masks at home aroundyour family members prevents influenza.
To use randomized trials to study whether masksreduce a virus's spread by keeping infectedpeople from transmitting a pathogen, we needrandomized comparisons of large groups, likehaving people in one city assigned to wear masksand those in another to not wear them. Asethically and logistically difficult as thatmight seem, there was one study during thepandemic in which masks were distributed, but notmandated, in some Bangladeshi villages and notothers before masks were widely used in thecountry. Mask use increased to 40 percent from 10percent over a two-month period in the villageswhere free masks were distributed. Researchers<https://www.nytimes.com/2021/09/26/opinion/do-masks-work-for-covid-prevention.html>found
an 11 percent reduction in Covid cases in thevillages given surgical masks, with a 35 percentreduction for people over age 60.
As we all know, the<https://www.wmbriggs.com/post/37320/>Bangladesh
mask study was<https://boriquagato.substack.com/p/bangladesh-mask-study-do-not-believe>garbage.
Another pandemic study randomly distributed masks
to people in Denmark over a month. About half the
participants wore the masks as recommended. Of
those assigned to wear masks, 1.8 percent became
infected, compared with 2.1 percent in the
no-mask group a 14 percent reduction. Butresearchers could not reach a firm conclusionabout whether masks were protective because therewere few infections in either group and fewerthan half the people assigned masks wore them.
The Danish study found no statisticallysignificant effect from masking; whining aboutadherence is pointless because once again, noevidence is no evidence. Also too, the authors ofthe Danish study had incredible trouble finding ajournal that would publish their results, becausethey weren't able to show masks preventinfection. Thus we see, as I said above, that<https://www.eugyppius.com/p/most-mask-studies-are-garbage>this
entire field is corrupted with massivepublication bias in favour of any study at allfinding find that masks work, and despite thisfact, all anybody can come up with is this weak tea.
Scientists routinely use other kinds of databesides randomized reviews, including labstudies, natural experiments, real-life data andobservational studies. All these should be takeninto account to evaluate masks.
Lab studies, many of which were done during thepandemic, show that masks, particularly N95respirators, can block viral particles. LinseyMarr, an aerosol scientist who has long studiedairborne viral transmission, told me even clothmasks that fit well and use appropriate materials can help.
Lab studies like this form a great part of the"evidence" that masks are effective at preventinginfection. They are also incredibly, boundlesslystupid - the equivalent of showing that wood canfloat, when called upon to demonstrate the seaworthiness of a sailboat.
Real-life data can be complicated by variablesthat aren't controlled for, but it's worthexamining even if studying it isn't conclusive.
Japan, which emphasized wearing masks andmitigating airborne transmission, had aremarkably low death rate in 2020 even though itdid not have any shutdowns and rarely tested andtraced widely outside of clusters.
Across the entire Asia-Pacific, regardless ofmasking practices, Covid case counts andmortality were much lower than in the West. Thebehaviour of SARS-2 varies massively acrossregions and seasons, but not across masked and unmasked populations.
David Lazer, a political scientist at
Northeastern University, calculated that before
vaccines were available, U.S. states without maskmandates had 30 percent higher Covid death rates than those with mandates.
This Lazer guy is a crank who seems to havechurned out various pro-containment Covid surveysthroughout the pandemic. I can't find thepublication containing this specific 30-percentfigure, but I really hope Lazer controlled forthings like age, obesity and socioeconomicstatus, because statewide mask mandates areheavily politicised in the United States, withmaskless Republican-governed states being inaggregate substantially older, sicker and poorerthan masked Democratic-governed states.
Perhaps the best evidence comes from naturalexperiments, which study how things change after an event or intervention.
Researchers at Mass General Brigham, one ofHarvard's teaching hospital groups,<https://jamanetwork.com/journals/jama/fullarticle/2768533>found
that in early 2020, before mask mandates wereintroduced, the infection rate among health careworkers doubled every 3.6 days and rose to 21.3percent. After universal masking was required,the rate stopped increasing, and then quickly declined to 11.4 percent.
Mass General Brigham instituted universal patientmasking from 25 March; allowing for a "lag periodto allow for manifestations of symptoms" from6-10 April 2020, the authors arrive at an"intervention period between 11-30 April 2020."The study is therefore worthless, because casedata shows that infections stopped rising acrossthe state of Massachusetts on 17 April. In otherwords, these geniuses started masking theirpatients just as Covid achieved a natural peakand then credited their masks with the ensuingseasonal decline. Identical reasoning processesare literally how things like rainmaking rituals arise.
In Germany, 401 regions introduced mask mandatesat various times over three months in the springof 2020. By carefully comparing otherwise similarplaces before and after mask mandates,<https://www.pnas.org/doi/full/10.1073/pnas.2015954117>researchers
concluded that "face masks reduce the dailygrowth rate of reported infections by around 47percent," with the effect more pronounced inlarge cities and among older people.
This study actually looks at just six Germanjurisdictions that introduced mask mandatesbefore the federal masking rules. It's gotvarious problems, like the use of "synthetic
controls", the failure to take into account test
frequency, and so on, but I just want you to
notice the insane, utterly impossible effectsize. There's no way masks cut infections by 47percent. The highest quality studies, like theDanish randomised controlled trial, find nostatistically significant effect at all, but asyou wander outwards to ever worse observationalstudies, masks look better and better. This isnot what the literature looks like when an intervention actually works.
[Cochrane editorial board member Michael D.]Brown, who led the Cochrane review's approvalprocess, told me that mask mandates may not betenable now, but he has a starkly differentfeeling about their effects in the first year of a pandemic.
"Mask mandates, social distancing, the othershutdowns we had in terms of even restaurants andthings like that if places like New York Citydidn't do that, the number of deaths would havebeen much higher," he told me. "I'm very confident of that statement."
So the evidence is relatively straightforward:Consistently wearing a mask, preferably ahigh-quality, well-fitting one, provides protection against the coronavirus.
Although even Tufekci's highly biased andselective review of available studies fails tomarshal any evidence showing that masks areeffective, she finds somebody on the Cochraneeditorial board who will tell her masks savedlives, and that's enough to declare that maskswork as far as she's concerned. And the sad thingis, for her readers, that will be enough.