Contrary to claims by anti-maskers, face masks do help reduce COVID spread

A review of 38 separate “studies” upon which anti-maskers falsely base their claim that “masks are useless”

covid-19_cultured_isolated_microscope
This scanning electron microscope image shows SARS-CoV-2 (yellow)—also known as 2019-nCoV, the virus that causes COVID-19—isolated from a patient in the U.S., emerging from the surface of cells (blue/pink) cultured in the lab. (Image credit: NIAID-RML)

[Last updated 11/25/2021. For a description of the updates to date, see the bottom of the post.]

Ever since the start of COVID, there has been a loud minority of people who claim that face masks don’t reduce the risk of COVID. Some anti-maskers claim that no masks are effective at protecting anyone while others claim that only N95 or better masks are effective. And both groups claim that the science supports their claims.

When confronted about these claims, most anti-vaxxers I’ve interacted with don’t bother to point to specific scientific studies on the subject, but a few do. And the single most common reference I’ve been sent is a link to an anti-abortion site, LifeSiteNews.com. The article claims to have 47 studies that show that masks are ineffective at protecting people from COVID.

I’ve reviewed all 47 links over the last few days and read all the referenced studies that I could get my hands on to assess whether or not the studies actually showed that “masks are useless.” What follows is a summary of my assessment of each of the studies, by reference number in the LifeSiteNews article. I will update this article with the details of how I reached my conclusions for each paper at a later date.

Summary of findings:

  • Of the 47 referenced “studies,” 9 are duplicates, leaving 38 total links.
  • Of the remaining 38 references, only four three had results that supported the claim that masks were ineffective/useless.
  • 27 28 demonstrated that masks were effective (to various degrees of effectiveness) at filtering particles of the size of the SARS-CoV-2 virus and larger.
  • I couldn’t draw any conclusions on four of them due to not being able to locate a free version of the study online.
  • The last three were studies that were not applicable, opinion not backed up by data, and/or mathematical models.

The table below details the links by reference number and whether the study supports, rejects, et al that masks are ineffective/useless.

Claim Ref. # Conclusion Claim Ref. # Conclusion
1
Study does not support claim
25
Study does not support claim
2
Study does not support claim
26 Mathematical model, not a study. Misleading
3
Study supports claim
27
Study does not support claim
4
Study supports claim
28
Study does not support claim

(duplicate of #38)

5 Abstract does not support claim. Full study unavailable (duplicate of #32 & #41) 29
Study does not support claim

(duplicate of #39)

6
Study does not support claim
30 Contamination study; does not address mask effectiveness at all. Misleading (duplicate of #40)
7 Study unavailable – no conclusion 31
Study does not support claim
8
Study does not support claim
32 duplicate of #5 & #41
9
Study does not support claim
33
Study supports claim Study does not support claim

(duplicate with #43)

10
Study does not support claim
34
Study supports claim
11
Study does not support claim
35
Study does not support claim

(duplicate with #44)

12 Study unavailable – no conclusion 36
Study does not support claim
13
Study does not support claim
37
News article does not support claim
14
Study does not support claim
38 duplicate of #28
15
Study does not support claim
39 duplicate of #29
16 Study unavailable – no conclusion 40 duplicate of #30
17
Blog, references do not support claim
41 duplicate of #5 & #32
18 Duplicate of #17 42
Study does not support claim
19
Study does not support claim
43 duplicate of #33
20 Editorial on exercise effects of masking is not a study. Editorial is not applicable.
44 duplicate of #35
21
Study does not support claim
45 Abstract does not support claim. Full study unavailable
22
Study does not support claim
46
Study does not support claim

(duplicate of #47)

23 Study unavailable – no conclusion 47 Duplicate of #46
24
Study does not support claim

I also assessed each link for whether it was original research (natural experiments, mechanical filtration studies, randomized controlled trials, etc), a review of other studies, or “gray” literature (not new research but based on the research of others). This is what I found:

  • 25 of the links were original research of varying types
  • Four of the studies were “gray” literature and/or not applicable to the claim(the blog post, news article, opinion letter, and mathematical model).
  • Nine were reviews of research that attempted to combine multiple studies into a greater whole (systematic reviews and meta-analyses) or define the current state of knowledge on the subject of masks (scoping review). Many of 26 original research studies were referenced in these nine reviews.

The table below details the links by reference number and what type of study the links are.

Summary table for type of study (original research, review of other studies, gray literature) [Update 10/20/2012: Updated all “original research” to identify whether the study was a natural experiment, a mechanical filtration experiment, or a randomized controlled trial.] [Update 9/8/2021: corrected for #4 being a meta-analysis rather than original research. List above has been updated as well]

Claim Ref. # Study Type Claim Ref. # Study Type
1 Natural experiment
25 Synthesis review of other studies
2 Scoping review of other studies 26 Mathematical model
3 Meta-analysis of other studies 27 Systematic review of other studies
4 Meta-analysis of other studies 28 Systematic review of other studies
5 Systematic review of other studies 29 Meta-analysis of other studies
6 Mechanical filtration experiment 30 Mask contamination study
7 Mechanical filtration experiment 31 Randomized controlled trial
8 Mechanical filtration experiment 32 Duplicate of #5 & #41
9 Mechanical filtration experiment 33 Randomized controlled trial
10 Mechanical filtration experiment 34 Mechanical filtration experiment
11 Mechanical filtration experiment 35 Mechanical filtration experiment
12 Mechanical filtration experiment 36 Mechanical filtration experiment
13 Mechanical filtration experiment  37 News article
14 Mechanical filtration experiment  38 Duplicate of #28
15 Mechanical filtration experiment  39 Duplicate of #29
16 Mechanical filtration experiment  40 Duplicate of #30
17 Blog review of other studies 41 Duplicate of #5 & #32
18 Duplicate of #17 42 Randomized controlled trial
19 Systematic review of other studies 43 Duplicate of #33
20 Letter/opinion review of other studies 44 Duplicate of #35
21 Randomized controlled trial
45 Mechanical filtration experiment
22 Mechanical filtration experiment  46 Viral isolation description
23 Mechanical filtration experiment  47 Duplicate of #46
24 Mechanical filtration experiment 

Finally, I cross-referenced the two tables to see if there were any patterns between the groups and I didn’t identify any.

  • Of the four links that supported the claim that masks were ineffective/useless, two were meta-analyses of other studies and two were original research.
  • Of the 24 links that found masks were effective at reducing the spread of COVID at least some, six were literature reviews (scoping, systematic, and meta-) and the remaining 18 were original research.
  • Of the six where I couldn’t get the full paper, one was a systematic review of other studies and the other five were original research.
  • The remaining four were gray literature and/or not applicable to the claim at hand.

I will add my review details below this point as time allows. The details will include any statement from the LifeSiteNews author, a link to the publicly available full text if available, the study title, and my assessment of why the study does or does not support the false claim that masks are ineffective/useless.


Claim #1

The US Centre for Disease Control performed a study which showed that 85 percent of those who contracted Covid-19 during July 2020 were mask wearers. Just 3.9 percent of the study participants never wore a mask. (emphasis added)

Study title and link: Community and Close Contact Exposures Associated with COVID-19 Among Symptomatic Adults ≥18 Years in 11 Outpatient Health Care Facilities — United States, July 2020

Study type: Natural experiment

What the study actually found: First, the study was not designed to measure the effectiveness or lack thereof of masking. The goal of the study was to determine whether there were places that study participants visited that resulted in more or less likelihood of having COVID-19 symptoms. What the study found was that people visiting bars and restaurants – places where masking is difficult if not impossible – were the most likely to come down with COVID. In other words, the researchers found indirect evidence that masks are probably effective, which is opposite the false claim that masks are ineffective/useless. [Ed. note: I debunked this one previously at this link.]

Second, if you review the emphasized sections of the claim above, you’ll see that there’s a subtle difference in language between the two. The first is written in a way that suggests it applies to everyone in the US, while the second is strictly limited to the study participants. Lots of people have assumed that the 85% claim applies universally, which the study does not say. In fact, the study explicitly says that the results may not be representative of the larger US population. This makes the claim misleading.

Conclusion: The study does not support the claim that masks are ineffective/useless, and the specific claim as written is misleading.


Claim #2

This study used 5462 peer-reviewed articles and 41 grey literature records.

“Conclusion: The COVID-19 pandemic has led to critical shortages of medical-grade PPE. Alternative forms of facial protection offer inferior protection. More robust evidence is required on different types of medical-grade facial protection. As research on COVID-19 advances, investigators should continue to examine the impact on alternatives of medical-grade facial protection”

So how is your cloth and surgical mask working again if EVEN medical grade alternatives are failing ?

Study title and link: Facial protection for healthcare workers during pandemics: a scoping review

Study type: Scoping review

What the study actually found: First, the study did not use “5462 peer-reviewed articles and 41 grey literature records.” That was the starting point for the scoping review. It was filtered down, like all literature review are, to 48 peer-reviewed studies and 19 gray literature records, for a total of 67.

Second, the authors concluded that most of the studies they reviewed didn’t have enough data on improvised masks to draw any firm conclusions, but that 10 studies found that improvised masks were better than no mask at all.

Third, the authors wrote the following (emphasis added): “Evidence supports the superiority of medical-grade masks for protection against respiratory viruses. In cases when medical-grade masks are not available, the summarized research indicates that protection offered by certain improvised or reused masks is better than no protection at all.”

Conclusion: The study does not support the claim that masks are ineffective/useless, and the specific claim as written is misleading.


Claim #3

“There is moderate certainty evidence that wearing a mask probably makes little or no difference to the outcome of laboratory-confirmed influenza compared to not wearing a mask”

Study title and link: Physical interventions to interrupt or reduce the spread of respiratory viruses

Study type: meta-analysis

What the study actually found: This study reviewed multiple other original research studies and focused on three different infection types – laboratory confirmed influenza, unconfirmed influenza-like illnesses, and any respiratory viral illness. And the results they found varied depending on the type of infection.

For lab-confirmed influenza, the reduction in cases in mask wearers over non-mask wearers had a wide confidence interval that overlapped the risk ratio value of 1 (in this case, the risk ratio is the number of mask wearers who got sick divided by the number of non-mask wearers who got sick). The study’s authors concluded that this result had a decent chance of being accurate after they made several assessments of bias in the studies.

For influenza-like illnesses, the authors found with a confidence interval of 85% that masks reduced risk, but they also concluded that the referenced studies were less likely to be accurate due to their bias assessments

And for all viral infections, the confidence interval was the lowest of all, but the confidence in the results was also very low.

Conclusion: The study does support the claim that masks are are ineffective/useless.


Claim #4

“We included three trials, involving a total of 2106 participants. There was no statistically significant difference in infection rates between the masked and unmasked group in any of the trials”

Study title and link: Disposable surgical face masks for preventing surgical wound infection in clean surgery

Study type: meta-analysis (initially classified as original research in error – corrections made to summary information above)

What the study actually found: First, the criteria used by the study’s authors were very restrictive. While the study was an update to a study originally performed in 2002, this 2016 update found only three original research studies that met their criteria. This is a very small number for a meta-analysis and limits the applicability of their conclusions.

Second, the three studies included in the meta-analysis were from 1984, 1991, and 2010. While the authors don’t address this particular point, there are likely differences in mask construction and materials over the course of the 26 year span of the three studies, and there may have been changes in surgical protocols that have an impact on wound infection. Such details may, or may not, affect the results of the meta-analysis.

These concerns do alter the conclusion of the meta-analysis, however, which is that there is no evidence of masks being effective at reducing surgical wound infections.

Conclusion: The study does support the claim that masks are are ineffective/useless.


Claim #5

Two randomized controlled trials were included involving a total of 1453 patients. In a small trial there was a trend towards masks being associated with fewer infections, whereas in a large trial there was no difference in infection rates between the masked and unmasked group.

Study title and link: Disposable surgical face masks: a systematic review

Study type: Systematic review

What the study actually found: I was unable to locate a full copy of the paper. As an observation, one of the study’s authors is the same as one of the authors of the study referenced in Claim #4 above.

Conclusion: No way to assess whether the full study does or does not support the claim that masks are ineffective/useless.


Claim #6

“Our results suggest that cloth masks are only marginally beneficial in protecting individuals from particles < 2.5 μm”

Study link and title: Evaluating the efficacy of cloth facemasks in reducing particulate matter exposure

Study type: mechanical filtration experiment

What the study actually found: This is a study of multiple different types of cloth masks as well as surgical masks and N95-style masks.

First, the study found that there was significant variation between different types of cloth masks, and as you would expect, all the masks filtered better at the low flow rate (essentially breathing while sitting) vs at the high flow rate (essentially light to moderate activity).

Second, the study found that there was significant variation in effectiveness of masks at different particle sizes. Smaller particles made it through the mask more easily than larger particles. Particles of the size of a single SAR-CoV-2 virus (~ 100 nm) were filtered better by the N95 masks (80%+ filter efficiency at both flow rates) and surgical masks (60-70% filter efficiency at high flow and 90%+ at low flow) than by two of the three cloth masks tested (30-40% filter efficiency at high flow and 50-70% at low flow). But one of the cloth masks was as good at filtering virus-sized particles as the N95 mask was.

Third, effectiveness of cloth masks improved with larger particles, just as it did with the N95 and surgical masks.

The authors had this to say as well: “Overall, cloth facemasks lead to a measureable reduction of total particle counts. However, there was variability among the performance of cloth masks and also among different particle sizes.” In every case, masks were better than no masks.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #7

“The filtration efficiency of the filter materials was good, over 95%, for particles above 5 micron in diameter but great variation existed for smaller particles.”

Coronavirus is 0.125 microns. therefore these masks wouldn’t protect you from the virus

Study link and title: Face seal leakage of half masks and surgical masks

Study type: mechanical filtration experiment

What the study actually found: I was unable to locate a full copy of the paper to review.

Conclusion: No way to assess whether the full study does or does not support the claim that masks are ineffective/useless.


Claim #8

“The filter efficiencies against influenza virus particles were the lowest”

“We conclude that the filter efficiency test using the phi-X174 phage aerosol may overestimate the protective performance of nonwoven fabrics with filter structure compared to that against real pathogens such as the influenza virus”

Study link and title: Comparison of the Filter Efficiency of Medical Nonwoven Fabrics against Three Different Microbe Aerosols

Study type: mechanical filtration experiment

What the study actually found: The study tested two different surgical mask materials to determine if three different microbes would penetrate them – a virus that infects bacteria (a “phage”) that was 14 nm, deactivated influenza in the range of 40-200 nm (average of 68 nm), and a bacterium that averaged 358 nm but some of which were bigger than 1 μm. The deactivated influenza virus is the closest in size to the SARS-CoV-2 virus, but the bacterium could serve as a stand-in for larger droplets containing lots of viruses.

The study found that tests that both surgical mask materials were good at blocking the phage and the bacterium weren’t as good at blocking the deactivated influenza virus. One material was only 20% effective at blocking the flu, but the other was 80% effective.

The authors also point out that the influenza virus is small when compared to other viruses that infect people, so the results for influenza might or might not be applicable to other, larger viruses. And while one material’s filter efficiency was low, it wasn’t so low as to indicate that the mask was ineffective or useless given SARS-CoV-2 is not spread just by individual viruses.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #9

“Although surgical mask media may be adequate to remove bacteria exhaled or expelled by health care workers, they may not be sufficient to remove the submicrometer-size aerosols containing pathogens ”

Study title and link: Aerosol penetration through surgical masks

Study type: mechanical filtration experiment

What the study actually found: The purpose of this 1992 study was to assess how effective surgical masks were for particles in the range of 150 nm to 4 μm and tested over aiflow ranges from 5 to 100 L/min (from resting to heavy exertion). They studied surgical masks that have a filter layer (the standard today) as well as masks that did not.

The first finding was that there is greater penetration of the mask at higher flow rates, but that for particles larger than about 2 μm the mask with the filter blocked nearly everything regardless of the flow rate.

The second finding was that the surgical mask with the filter was effective at filtering particles all the way down to 150 nm. At 100 L/min (heavy exertion) the worst filtering was about 30% for 400 nm particles, but at 5 L/min (resting) the worst filtering was about 65% at 700 nm. For SARS-CoV-2 virus-sized particles, the worst filtering effectiveness was 35% at 100 L/min and went up to 80% at 5 L/min.

Something to keep in mind is that this study was intended to determine how well surgical masks protected the wearer, not the people around them. Regardless, the study showed that surgical masks were effective at particle sizes similar and larger than SARS-CoV-2 virues.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #10

“We found that 60 GSM face mask had particle capture efficiency of 94% for total particles greater than 0.3 microns”

How big is the virus again? 0.125 microns.

Study title and link: Particle removal from air by face masks made from Sterilization Wraps: Effectiveness and Reusability

Study type: mechanical filtration experiment

What the study actually found: The purpose of the study was to assess if the use of sterilization wraps (the material that is wrapped around sterile medical tools to keep them sterile until use) could be used in place of surgical masks or N95 respirators. This study only analyzed filtration efficiency down to 300 nm, which is larger than the average size of most viruses but smaller than the average size for most bacteria.

The study found that sterile wraps were better than surgical masks and nearly as good as N95 respirators at capturing particles sized 300 nm, 500 nm, and 1 μm. The study did not test particles smaller than 300 nm, and so no conclusions can be drawn from this study about the effectiveness (or lack thereof) of the tested masks at 125 nm.

Conclusion: The study does not support the claim that masks are ineffective/useless and the specific claim as written is misleading.


Claim #11

This study states that “alternatives” like silk and gauze etc could possibly be good options in the pandemic. It’s done on starch particles.

Does not state how big they are either, but they can still get through the material and my research points out that starch particles are “big”, much bigger than most viruses.

Study title and link: A New Method for Testing Filtration Efficiency of Mask Materials Under Sneeze-like Pressure

Study type: mechanical filtration experiment

What the study actually found: In addition to the starch particles which averaged 8.2 μm in size, the researchers also studied latex microspheres that averaged 750 nm in size. The study showed that 4 layers of silk blocked nearly 94% of the microspheres, 2 layers of cotton blocked 75%, fine gauze blocked 79%, and 4 layers of linen blocked 67%. Four other options were also studied, with the poorest filtration of 53% for 2 layers of linen.

At sneeze-like pressures, however, all of the cloth masks blocked both microspheres and starch particles better than the surgical masks did.

Note that the specific claim for this study says that study missed the microspheres and the fact that the sizes of the particles were described in the study, indicating that the claim’s author did not read the study closely enough.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #12

“The laser mask provided significantly less protection than the FFP2 respirator (P=0.02), and only marginally more protection than the surgical mask. The continued use of laser masks for respiratory protection is questionable. Taping masks to the face only provided a small improvement in protection”

Study title and link: Protecting staff against airborne viral particles: in vivo efficiency of laser masks

Study type: mechanical filtration experiment

What the study actually found: This is one of the studies that I was unable to review due to it being publicly unavailable, but I was able to review the abstract. In this case, the abstract alone was sufficient to find that the specific claim above is quote mined. What the abstract says prior to the quote above is that particle counts were reduced 3x in the worst case, with the best case being just over 100x reduction in particles.

Conclusion: Study unavailable – no conclusion drawn.


Claim #13

“Worn as designed, both commercial surgical masks and cloth masks had widely varying effectiveness (53 – 75 percent and 28 – 91 percent particle removal efficiency, respectively)”. Different brand, different results and only when they applied “nylon layers” did the “efficiency” improve. Synthetic fibres do not breathe, so this will inevitably effect your breathing.

Study title and link: Quantitative Method for Comparative Assessment of Particle Removal Efficiency of Fabric Masks as Alternatives to Standard Surgical Masks for PPE

Study type: mechanical filtration experiment

What the study actually found: This study measured the filter efficiency of multiple different types of materials (different cloths, multiple surgical masks, and several N95 respirators. When looking at strictly material filtration, the least efficient mask material was still 55% effective at filtering out particles. And in what is perhaps a completely unsurprising result, N95 respirators were the most effective.

When fit was taken into account, there was a great deal more variability in the filter efficiency of masks. Surgical masks varied from 53% to 78% efficient. Cloth masks varied even more widely, filtering between 28% and 91%, making the best quality masks nearly as good as N95 respirators. And adding a loose nylon mesh as a way to improve the fit of the mask to the wearer’s face always improved filter efficiency.

Construction also mattered. Flat masks similar in construction to disposable surgical masks were more variable than cone-shaped masks like N95 respirators and fitted cloth masks. Actual surgical masks performed better than surgical style cloth masks, but cone-shapped cloth masks worked better than surgical masks even when the nylon mesh was added to improve fit.

Unfortunately, I couldn’t find anywhere in the study where the sizes of the particles were characterized, so it’s likely that these results would vary if the particle sizes were accounted for as well.

The author of the LifeSiteNews article is significantly confused about what “breathability” of fibers means. It doesn’t mean that the fibers can’t be breathed through, which seems to be what the specific claim above is implying. Breathability the term used for fabrics that allow perspiration to escape from the body instead of being held next to the skin or being absorbed by the fabric. In addition, most breathable fabrics are, in fact, synthetic or synthetic/natural blends (with the exception of wool).

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #14

“Since the microspheres were not identified on the exterior of these face masks, they must have escaped around the mask edges and found their way into the wound”. Human albumin cells, aka aborted fetal tissue, is much larger than the virus and still escaped the mask.

Study title and link: The Efficacy of Standard Surgical Face Masks: An Investigation Using “Tracer Particles”

Study type: mechanical filtration experiment

What the study actually found: This study from 1980 looked at albumin spheres that are 10-35 μm in diameter, much larger than virus particles and significantly larger than many bacteria. It found human albumin particles in the wounds of patients, but when the cause was identified, it was due to air escaping from around the edges of the masks, rather than from penetration of the particles through the mask material itself. The researchers determined this by studying the exterior of the masks and finding no albumin particles on it.

When the researchers changed the order in which the doctors wearing the masks garbed up for surgery (putting the mask on first, underneath the head covering instead of over the top of the head covering), no albumin particles were found in the wounds of patients.

There is also a dishonest appeal to emotion logical fallacy in the specific claim above, specifically that human albumin is from “aborted fetal tissue.” Albumin is a protein, not a cell, and it’s extracted from adult blood plasma. Also, as someone who has worked in cleanrooms, this study’s finding that wearing a mask with the edges underneath the head covering (bonnet) is exactly what we do for cleanrooms specifically because wearing the mask that way significantly reduces the particles emitted while exhaling.

This study does point out the importance of fit to proper mask wearing, but also shows that, at least for larger particles, masks filter the particles very well. But this study is from 1980 and without a detailed literature review, it’ll be difficult to say how relevant this study is to modern masks and masking protocols.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #15

“Our findings suggest that a homemade mask should only be considered as a last resort to prevent droplet transmission from infected individuals” so why has the government suggested you make your own when they are not effective ?

Study link and title: Testing the Efficacy of Homemade Masks: Would They Protect in an Influenza Pandemic?

Study type: mechanical filtration experiment

What the study actually found: This study used harmless bacteriophages (small viruses that infect bacteria) that were about 23 nm in size (4-5x smaller than typical viruses that infect people) and a bacteria that varies from 0.95-1.25 μm as proxies for the influenza virus. Using these two pathogens, the researchers found that all of the homemade masks had some degree of effectivness at filtering even the smallest particles. The worst filter efficiency was a scarf material against the phages, which was only 49% efficient, while the other materials varied from 51% efficient to 90% efficient.

The authors of the study found that the best cloth masks they researched were 2.5-2.9x more protective than nothing, while surgical masks were between 4.1 and 5.3x more protective. And the researchers remarked that cloth masks seemed to get more protective as they were worn for some reason they didn’t understand while N95 masks degraded as they were worn.

Regardless, wearing a homemade cloth mask as a “last resort” when it’s 2.5x more protective than nothing still shows some effectiveness. And the actual quote from above is quote mined. The full quote is “Our findings suggest that a homemade mask should only be considered as a last resort to prevent droplet transmission from infected individuals, but it would be better than no protection (emphasis added).”

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #16

“Increasing the filtration level of a particle respirator does not increase the respirator’s ability to reduce a user’s exposure to contaminants”

Study title and link: Using half-facepiece respirators for H1N1

Study type: mechanical filtration experiment

What the study actually found: I was unable to find this study available to the public for free, so I could not assess it.

Conclusion: Study unavailable – no conclusion drawn.


Claim #17

The site is full of studies proving masks dont (sic) work for coronavirus or the flu.

Study title and link: Masks Don’t Work: A Review of Science Relevant to COVID-19 Social Policy

Study type: Blog post/literature review

What the blog actually found: First, this is not a study. It is a blog review of other studies by a physicist (non-expert) who claims that all his referenced studies show masks are ineffective. As such it not be included in an article that claims to point to 47 scientific studies. However, as the blog post does reference a number of other studies within it, I have chosen to review those studies briefly.

Study A: This is a duplicate of study #32 (Use of surgical face masks to reduce the incidence of the common cold among health care workers in Japan: A randomized controlled trial), which does not support the claim that masks are ineffective/useless.

Study B: This is a duplicate study of #19 (Face masks to prevent transmission of influenza virus: A systematic review), which does not support the claim that masks are ineffective/useless.

Study C: This is a new study not referenced in the original 47. It’s titled “The use of masks and respirators to prevent transmission of influenza: a systematic review of the scientific evidence” and it’s a systematic review of other studies. The review found three randomized controlled trials that showed masks reduced infections among healthcare workers and in households, five RCTs masks plus improed hygiene resulted in reduced infections but not necessarily a the classical 95% confidence interval, and nearly all observational studies found that mask wearing reduced infections among the groups observed. This does not support the claim that masks are ineffective/useless.

Study D: This is a new study not referenced in the original 47. It’s titled “Effectiveness of N95 respirators versus surgical masks in protecting health care workers from acute respiratory infection: a systematic review and meta-analysis” and it is a meta-analysis of other studies. The analysis assessed the effectiveness of N95 respirators as compared to surgical masks, not masks vs. no-masks as the blog’s author suggests. And while it did find that N95 respirators were more effective than surgical masks, the finding was not at the classical 95% confidence interval. As a result, this study is not applicable to the blog author’s claim and, therefore, does not support the claim that masks are ineffective/useless.

Study E: This is a new study not referenced in the original 47. It’s titled “Effectiveness of Masks and Respirators Against Respiratory Infections in Healthcare Workers: A Systematic Review and Meta-Analysis” and it’s another meta-analysis. The analysis found various studies that, combined, showed N95 respirators were much better than no protection and that respirators were better than surgical masks (although not at the 95% confidence interval for influenza). In fact, the figure from the paper that the blog’s author used was the one case in the entire paper that showed a low statistical confidence interval – all the others showed near or greater than 95% confidence, indicating that the blog’s author dishonestly cherry picked his data. While the analysis says nothing about whether medical or cloth masks are better than no masks, it finds conclusively that N95 respirators are better than nothing, so this analysis does not support the claim that masks are ineffective/useless.

Study F: This is a new study not referenced in the original 47. It’s titled “N95 Respirators vs Medical Masks for Preventing Influenza Among Health Care Personnel” and it’s a randomized clinical trial. The trial did not assess the effectiveness of masks vs no masks, only N95 masks vs. surgical masks. As such, it is not applicable to the question at hand and therefore does not support the claim that masks are ineffective/useless.

Study G: This is a duplicate of study #29 (Effectiveness of N95 respirators versus surgical masks against influenza: A systematic review and meta-analysis) which does not support the claim that masks are ineffective/useless.

Conclusion: None of the studies referenced in the blog support the claim that masks are ineffective/useless.


Claim #18

This is full of studies proving mask protection is negligible for coronavirus, flu etc.

This is a duplicate of Claim #17 above.


Claim #19

There is less data to support the use of face masks or respirators to prevent becoming infected.

Study title and link: Face masks to prevent transmission of influenza virus: a systematic review

Study type: Systematic review

What the study actually found: This systematic review started with a list of 279 studies that they then reduced to just 12 that met their standards for inclusion: one experimental study, six studies set in healthcare settings, and five studies in community settings.

The review summarized four of the six healthcare studies. One found no statistically significant (at the 95% confidence interval) difference between N95 respirators and surgical masks, another out of Japan had too little data to draw any conclusions, and another from Hong Kong found that failing to adhere to protections increased infections, but nothing more definitive. A study from 1918 found some evidence for masking plus hygiene reducing infections, but masks were not studied in isolation. And the last two studies were mentioned but not described in detail.

The review summarized four of the community studies as well. One study out of Hong Kong found that masks plus improved hygiene had a statistically significant (at the 95% confidence interval), but masks were not checked in isolation. Another found a significant decrease in SARS infections for masking plus hygiene plus distancing, but again the factors were isolated. A dormitory study found masks found that masks did reduce infections, but it wasn’t significant to the 95% confidence interval, and the review concluded that the study had issues with data quality. And the last study found a massive reduction in risk of infection for people wearing N95 or surgical masks that was significant to the standard 95% confidence interval.

All in all, the review concluded that the there’s not a lot of good data on the subject of masking vs. no masking because most studies weren’t designed to measure that, because the few studies that were designed for that had very small sample sizes, because the targeting of interventions (masks) was too narrow, and/or because the testing methods were inconsistent.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #20

No evidence to suggest that wearing a mask during exercise offers any benefit from the droplet transfer from the virus.

“Exercising with facemasks may reduce available Oxygen and increase air trapping preventing substantial carbon dioxide exchange. The hypercapnic hypoxia may potentially increase acidic environment, cardiac overload, anaerobic metabolism and renal overload, which may substantially aggravate the underlying pathology of established chronic diseases”

Study title and link: “Exercise with facemask; Are we handling a devil’s sword?” – A physiological hypothesis

Study type: Editorial letter

What the study actually found: The editorial talks about a lot of potential health concerns of forcing the body to work harder to breath through a mask during heavy exercise. And they refer to one study about the possible risk of touching one’s face more often increasing the risk of infection. But they also suggest that wearing a mask for low or moderate intensity exercise would be OK because the exerciser isn’t putting as much stress on their cardiovascular and/or pulmonary systems.

It’s important to note that the authors of this study are sports scientists, and thus are themselves not experts on virus transmission or masking, only on how the body responds to exercise.

Conclusion: The editorial is not applicable as it offers no data to support or refute the claim that masks are ineffective/useless.


Claim #21

Surgical site infection rates did not increase when non-scrubbed personnel did not wear face masks.

Study title and link: Use of face masks by non-scrubbed operating room staff: a randomized controlled trial

Study type: randomized controlled trial

What the study actually found: The study investigated the effectiveness of masking non-scrubbed staff in an operating room. The non-scrubbed personnel are those support staff who are not involved in the surgery itself, so not the surgeon, surgical assistant(s), or surgical technician(s). Non-scrubbed staff are required to stay a minimum of 12 inches away from the sterile field (the area defined by the scrubbed staff as sterile via sterile covers and wraps) at all times in order to minimize infection risk to the patient.

What the study found is that there did not appear to be an increase in the risk of infection if non-scrubbed staff were unmasked vs. masked. Instead, the study found that the duration of the surgery and the length of hospital stay were the two greatest factors contributing to increased infection risk.

Something to keep in mind about modern operating rooms is that they use HEPA filters to deliver sterile air into the area over the patient and blow out to the edges of the room where contaminated air is removed.

Note also that the scrubbed staff all wore masks.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #22

When the wearing of face masks by non-scrubbed staff working in an operating room with forced ventilation seems to be unnecessary.

Study title and link: Surgical face masks in modern operating rooms-a costly and unnecessary ritual?

Study type: mechanical filtration experiment

What the study actually found: The study placed settle plates on the surface of an operating table and had operating room personnel whisper or talk normally while wearing masks. The settle plates were then incubated and the number of resulting bacterial colonies were counted.

For scrubbed personnel (those who would be operating on the patient or directly helping the surgeon operate), there were 12 or 29 colonies when masks were not worn, but zero colonies when masks were worn. For non-scrubbed personnel standing at least 1 meter from the operating table there were zero colonies wearing a mask or not.

Note that the study was published in 1991

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #23

Wearing multi layer operating room masks for every visit had no effect on nose and throat carriage rates.

Study title and link: Masks: a ward investigation and review of the literature

Study type: mechanical filtration experiment

What the study actually found: Unknown – the study was not publicly available. Note that the study was published in 1986.

Conclusion: Study unavailable – no conclusion drawn.


Claim #24

The protection provided by surgical masks may be insufficient in environments containing potentially hazardous submirconometer-sized aerosols.

“Conclusion: We conclude that the protection provided by surgical masks may be insufficient in environments containing potentially hazardous submicrometer-sized aerosols”

Study title and link: Aerosol penetration and leakage characteristics of masks used in the health care industry

Study type: mechanical filtration experiment

What the study actually found: The study tested how well masks filtered out particles of corn oil ranging in size from 100 nm to 4 μm at a flow rate of 32 L/min (moderate exertion). The study found that different materials filtered differently, with some filtering very well and others very poorly – the best masks filtered 80% of the smallest particles (100 nm is about the size of a virus), while the worst filtered 0%. It also found that leaks due to poor fit reduced the filter efficiency as much as 20%. And it found that the thickness of the mask material had the greatest effect on filter efficiency – thicker masks filtered better.

Note that this study was published in 1993. Also note that “insufficient” protection is not the same as “no protection.”

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #25

Meta analysis review that says there is limited evidence to suggest that the use of masks may reduce the risk of spreading viral respiratory infections.

Study title and link: Masks for prevention of viral respiratory infections among health care workers and the public

Study type: synthesis review

What the study actually found: This synthesis review looked at a number of other studies, both randomized controlled trials and systemic reviews, and attempted to draw them all together into a single conclusion about the effectiveness of masks.

In general, the review found that previous randomized controlled trials had a high risk of bias in the 12 community studies that the review looked at, and that bias may be why the review’s authors were hesitant to claim that there was any significant effect of masking at the community level. Their conclusion was overly conservative for two reasons. First, as the review makes clear, the two highest quality community studies both found a statistically significant reduction in influenza-like illnesses. Second, the risk reduction in community settings was significant to 93%, instead of the 95% standard that the review’s authors used.

In healthcare settings, there were two studies that compared masks with no masks, but one was very small and thus there wasn’t enough data to draw firm conclusions and the other reported that most of the “no mask” control group reported wearing a mask at some point during the study period, essentially contaminating the data. The other four studies found surgical masks were somewhat better than cloth masks, and N95 respirators were somewhat better than surgical masks.

Ultimately, the synthesis review’s strongest conclusion is that assessments of the effectiveness of masks in randomized controlled studies are plagued with low power (number of infections were insufficient to differentiate between control and test), poor adherence to instructions (control group wearing masks or test group not wearing masks, or changing the type of mask worn), and small sample sizes.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #26

Evidence to suggest that transmission probability is strongly driven by indoor air quality, followed by patient effectiveness and the least by respiratory protection via mask use. So this could explain “second waves” and has nothing to do with hand shaking, or not wearing a mask.

Study title and link: Modeling of the Transmission of Coronaviruses, Measles Virus, Influenza Virus, Mycobacterium tuberculosis, and Legionella pneumophila in Dental Clinics

Study type: modeling review

What the study actually found: This is an assessment of a mathematical model for the transmissibility of multiple pathogens developed previously. It assumes that the previously developed model is correct and then assesses the effects of various terms of the model on transmission.

Put another way, if the model is wrong, then every conclusion drawn by this mathematical assessment is also wrong.

The model itself is based on a number of assumptions: that droplets and aerosol (breath) are both infectious, that infectious particles are the same concentration everywhere within a room (aka homogeneous distribution), and that all the particles remain infectious for the entire time the room is occupied. The authors made a number of additional assumptions, namely that surgical masks are only 24% effective (they can be significantly more effective depending on fit and how they’re made) and that FFP2 masks (similar to N95) are 85% effective for particles less than 2 μm in size.

The issue with this kind of study is that it’s entirely dependent on the accuracy of the model. First, if the model is wrong, then all the conclusions are wrong – garbage in, garbage out. Second, the conclusions are only applicable in situations where the model’s assumptions are accurate. In this case, the homogeneous distribution assumption is only applicable after a pathogen has been in the air for a period of time. And if only droplets are infectious, then the fact that they settle quickly will change the model significantly.

This review is not a review of science, it’s a review of mathematics, and as such it’s misleading to include it in a review of studies of masking.

Conclusion: This study is not applicable to the claim that masks are ineffective/useless and is thus misleading.


Claim #27

The use of face masks, either by infected or non infected peresons (sic), does not have a significant effect on influenza transmission.

SO MASKS DON’T PROTECT YOU FROM ME, AND VICE VERSA.

Study title and link: Nonpharmaceutical Measures for Pandemic Influenza in Nonhealthcare Settings—Personal Protective and Environmental Measures

Study type: systematic review

What the study actually found: The review assessed multiple other studies to determine whether they collectively found a statistically significant (to the usual 95% confidence interval). The conclusion is that they did not because in all three cases assessed, the confidence interval included the value 1.00 (which is equal risk – less than one is lower risk, greater than one is increased risk).

However, the data can be recalculated to determine what the confidence interval actually is and, from that value we can determine the percentage likelihood that masks reduce the risk of infection. When we do that, we find the following for the three conditions assessed:

Face masks alone: risk ratio of 0.78, 95% confidence interval of 0.51-1.20. Confidence interval for a reduced risk of infection from wearing a mask (<1.00) using a 1-tailed Student’s T-test: 87%.

Face masks plus hand hygiene: risk ratio of 0.91, 95% confidence interval of 0.73-1.13. Confidence interval for a reduced risk of infection from wearing a mask (<1.00) using a 1-tailed Student’s T-test: 80%.

Face masks with or without hand hygiene: risk ratio of 0.92, 95% confidence interval of 0.75-1.12. Confidence interval for a reduced risk of infection from wearing a mask (<1.00) using a 1-tailed Student’s T-test: 80%.

Put another way, there’s at least an 80% chance that masks are effective at reducing the risk of infection.

As a final obeservation, this review found that masking alone protected better than masking plus hand hygiene. This is a counter-intuitive result – washing pathogens off your hands should not increase the chance of infection. The review’s authors found that most of the studies were underpowered due to small sample sizes and had issues with mask compliance, which means that data issues are likely confusing the results.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #28

Meta analyses suggest that regular hand hygiene provided a significant protective effect over face masks and their insignificant protection.

Study title and link: Effectiveness of personal protective measures in reducing pandemic influenza transmission: A systematic review and meta-analysi

Study type: systematic review

What the study actually found: This review assessed many of the same randomized controlled trials that other systematic reviews have and as such it has many of the same issues – small datasets, low power, et al. The authors described in detail what their concerns were with each of the studies that met the criteria to be included in the review: poor definition of what was a “case”, poor representativeness of “cases,” difficulty in ascertaining who was and was not exposed, problems with the method of ascertaining exposure, and poor response rate. All of these “biases” reduce the ability of each study, and of this review, to determine how effective masks and hygiene are at reducing infections.

All that said, the review concluded that hand hygiene was significantly effective at reducing infections and that face mask use showed “mixed results” with a randomized controlled trial suggesting effectiveness.

Specifically, the review found that face masks showed a mean risk ratio of 0.53 with a 95% confidence interval of 0.16 to 1.71. The confidence interval for a reduced risk of infection from wearing a mask (<1.00) using a 1-tailed Student’s T-test is 86%.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #29

Use of n95 respirators compared to surgical masks is not associated with a lower risk of laboratory confirmed influenza.

Study title and link: Effectiveness of N95 respirators versus surgical masks against influenza: A systematic review and meta‐analysis

Study type: meta-analysis

What the study actually found: This analysis assessed six other studies (which have been assessed in other systemic reviews in this list) and concluded that N95 respirators were not better than surgical masks against lab-confirmed influenza at the 95% confidence interval (CI). With respect to lab-confirmed influenza, the study found that surgical masks were better than N95 respirators. It also found no significant differences (at the 95% CI) for other respiratory viruses or overall infections.

However, the study did find that there was an 85% confidence interval that N95 respirators were protective against non-influenza respiratory viruses and a 94% confidence interval that N95 respirators reduced infections overall. And N95 respirators showed a huge improvement over surgical masks for bacterial infections.

The analysis did not review any studies of masks vs. no masks.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Claim #30

Face mask surfaces can become contamination sources. People are storing them in their pockets, bags, putting them on tables, people are reusing them etc. This is why this study is relevant: [Ed. note: no additional information]

Study title and link: Adolescents’ face mask usage and contact transmission in novel Coronavirus

Study type: mask contamination study

What the study actually found: This study assessed the amount of contamination that can transfer between a mask and surfaces if the outside of the mask is touched. It did not assess the effectiveness of masks in any way.

Conclusion: This study is not applicable to the claim that masks are ineffective/useless and is thus misleading.


Claim #31

Loosely folded face masks and “bandana style” face coverings provide minimum stopping capability for the smallest aerosolized droplets.

This applies to anyone who folds or shoves a mask into their pockets or bag. It also applies to cloth and homemade cloth masks:

Study title and link: Visualizing the effectiveness of face masks in obstructing respiratory jets

Study type: Mechanical filtration study

What the study actually found: The study looked at emulated coughs in order to determine what materials (none, bandana, folded handkercheif, stitched mask from quilting cotton, or commercial cone face mask) stopped a cough the most effectively. It found that all masks had leakage around the mask, but all masks significantly reduced the distance a cough would travel, from 8 feet without a mask to a minimum of 2.5 inches for quilting cotton.

The study found that bandanas weren’t that good at stopping coughs compared to other materials, that a handkerchief was OK, and that the commercial cone mask was only about 8 inches.

The comment in the specific claim above doesn’t actually apply to the study, but it takes reading the study to understand that.

Conclusion: The study does not support the claim that masks are ineffective/useless.


Description of updates

[Updated 11/28/2021: Added claims 29-31.]

[Updated 11/26/2021: Added claims 25-28.]

[Updated 11/25/2021: Added claims 19-24.]

[Updated 10/21/2021: Added claims 16-18.]

[Updated 10/20/2021: A deeper review of claim #33 shows that it does not support the claim either, reducing the number of supporting studies to three. Updated study descriptions. Added claims 10-15]

[Updated 9/11/2021: Claims 1-9 are described below the summaries.]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Website Powered by WordPress.com.

Up ↑