What’s remarkable about this
review is that the authors discredit their own work almost immediately.
The poor quality writing is readily apparent, both from their repeated phrasing of “facemasks” as one word, and the fact that of 1,732 studies considered for inclusion, only 13 actually met the criteria.
That’s correct, a mere 0.75% of the studies they apparently examined were actually used to generate their conclusions.
So how many people did this substantial, important, influential, seminal work actually cover?
Facemasks have become a symbol of disease prevention in the context of COVID-19; yet, there still exists a paucity of collected scientific evidence surrounding their epidemiological efficacy in the prevention of SARS-CoV-2 transmission. This systematic review sought to analyze the efficacy of facemasks, regardless of type, on the prevention of SARS-CoV-2 transmission in both healthcare and community settings. The initial review yielded 1732 studies, which were reviewed by three study team members. Sixty-one full text studies were found to meet entry criteria, and 13 studies yielded data that was used in the final analysis. In all, 243 subjects were infected with COVID-19, of whom 97 had been wearing masks and 146 had not. The probability of getting COVID-19 for mask wearers was 7% (97/1463, p=0.002), for non-mask wearers, probability was 52% (158/303, p=0.94). The Relative Risk of getting COVID-19 for mask wearers was 0.13 (95% CI: 0.10-0.16). Based on these results, we determined that across healthcare and community settings, those who wore masks were less likely to contact COVID-19. Future investigations are warranted as more information becomes available.
243 people.
Yeah, 243. There have been 583,211,225 reported COVID cases in the world to date, with many millions more undetected, and this evidence review covered 243 of them.
A comparison between the two numbers illustrates the absurdity of suggesting that 243 is a remotely representative sample:
Beyond the incomprehensibly small sample size used to generate these percentages, the most absurd part of their conclusion was ignoring that inescapable reality that everyone will get COVID.
There can be no reduction in likelihood of getting COVID from mask wearing because everyone will eventually get COVID. The absolute reduction is 0. The relative reduction is 0. End of study.
Of course, that’s not at all what happened in this instance, and the details make it look even worse.
The “Evidence”
The studies they collected to include in their “evidence” review were a combination of embarrassingly bad to unbelievably useless.
But before going into the studies they collected, it’s worth pointing out that the conclusions conflated healthcare and community settings.
Sixty-one full text studies were found to meet the criteria, and 13 studies were used in the final analysis. (Figure 1) Frequencies, relative risk, confidence intervals and t-tests were calculated where appropriate, to measure differences between groups who reported wearing masks vs. not wearing masks for the overall study group, as well as health care, and community settings.
It’s the height of absurdity to compare the two while pretending that your results are some kind of definitive, universally applicable data driven exercise.
But the studies included are where it gets extremely bad.
One of them, in an evidence review designed to supposedly determine the effectiveness of mask wearing to prevent COVID, was conducted in 2004.
Yes, you read that correctly. 2004.
To no one’s surprise, there were no COVID infections in 2004 in Thailand when a full PPE policy was in place.
Better yet, they decided that the CDC’s embarrassing attempt at science, the infamous hairdresser study, should qualify for this exercise:
The audacity of including a study from 2004 and presenting two hairstylists as some kind of useful evidence should be immediately disqualifying.
Although based on their demonstrated standards, it’s no surprise that only testing half of the supposedly “exposed” individuals was enough to meet the criteria for entry.
It gets better.
Another study included used such high quality methods as asking long term care facilities to fill out a questionnaire describing their compliance with “preventative measures:”
How is it possible that not one person involved in this examination stopped to wonder if a questionnaire such as this could be prone to bias, especially in July 2020 at the height of COVID panic when masking suddenly became the single most important intervention to stop the spread of the virus.
Shockingly, several poorly designed studies from China were included that demonstrated the benefits of mask wearing, with one example apparently using “social network analysis.”
As pointed out on Twitter, one included piece of “evidence” had nothing to do with mask wearing at all:
You can clearly see the relevance of a paper on protection provided by gas masks while performing a tracheotomy is to preventing the spread of COVID.
It’s notable that the evidence review references two charter flight studies, although as mentioned in the tweet, this had nothing to do with a charter flight.
What appears to have happened is that the authors copy-pasted the same explanation onto two different studies. It’s easy to see how thorough and well considered their work was; not at all sloppy or shoddy.
Not to mention that the actual charter flight examination involved 11 people
who were all wearing masks.
It’s impossible to judge how effective masks are or aren’t, when you’re not comparing to people who aren’t wearing masks.
Absolutely ludicrous.
Also ludicrous is the fact that there are literally zero included studies after July 2020.
They also lumped together community examinations with individual ones.
From every possible angle, this is a disgraceful, demonstrably incompetent attempt to promote masking, with zero merit.
So it’s been ignored by “experts” and other prominent media figures, right?