Skip to main content

The Danish Mask Study

By November 18, 2020Commentary

This is the fuller post.   The headline result–wearing a mask makes no difference in cases or in level of transmission in the community.   It is obvious to all of us now, that this is true, as we can see the case growth in areas of the country with extremely high mask-wearing rates.  Here is the cite.  (Danish Mask Study)  Coupled with the withdrawal of the county study which purported to show a benefit based on limited time period study, this should put an end to any Governor claiming that mask mandates are backed by science and data, but it won’t.  As I noted in my flash report, people are already coming out with all kinds of BS criticisms of the trial. It is amusing in a sense, that people who just accepted all the mask modeling “research” without any quibbles or skepticism are finding all kinds of criticism directed at an actual randomized trial attempting to answer a real-life question–what is the effect of wearing a mask in the community.

Before describing the study and its findings, let me make a couple of preliminary observations.  I have been reading medical research for 40 years and science papers for even longer.  When you read this study and you know the history of the attempt to get it published, you are immediately struck by what the authors must have been forced to do to get any reputable journal to publish it.  They were clearly forced to constantly refer to limitations and caveats about the study and there is a discussion about the confidence intervals around the results that you absolutely never see in the published reports of research.  What kind of science is it that decides in advance what the results have to be and tries to suppress any research that challenges those pre-ordained results?  When I complain that science has become completely politicized, what happened to this study is Exhibit A.

Unlike all the garbage modeling studies about mask mandates and mask wearing reducing cases, this was an actual randomized trial on the critical question–does wearing a mask in the community affect the level of transmission, the number of cases, in that community.  It is real-life, we aren’t wearing masks in a model simulation or in some experimental room.  We wear them in the community, in real life, properly, improperly, touching them, not touching them, with gaps, without gaps, changing them frequently or not, washing them or not.  That is the reality of mask-wearing.  That is what this study replicated.

6000 people participated in the study.  3030 were randomly assigned to the mask group and 2994 to the control arm.  Out of the entire group 4862 completed the study period.  These were adults who spent more than 3 hours a day outside their homes and didn’t wear masks at an occupation.  Both the mask wearing arm and the non-mask wearing arm were told to follow social distancing measures.  The mask-wearing group was encouraged to wear a mask outside their home and were given 50 surgical masks for this purpose.  So better masks than cloth ones.  The primary outcome was number of infections after one-month.    42 people got infected in the mask wearing group and 53 in the control group.  Because more dropouts occurred in the mask group, there was no statistical difference in infection rate.  One interesting little nugget in the characteristics of the groups is that more people in the control arm had what might be considered higher-risk occupations than in the mask wearing arm.  Yet still no difference in infection rate.

Some of the criticisms are that the data on mask usage were self-reported, this only tested the effect on the mask wearer, not whether they might be less likely to infect others, and other niggling concerns.  The real criticism is that it doesn’t fall in line with the orthodoxy on all of us needing a mask glued to our face 24 by 7.  Self-reported adherence was good, with over 90% of the mask-wearers saying they largely complied with the recommendation and almost half saying they always did.  Taking out the 7% who reported poor adherence did not change the primary finding.  And very importantly, looking only at those persons who reported always complying with the mask recommendation also did not change the outcome.  It appeared there was very low in-home transmission that caused the infection in either arm of the study, so most transmission was occurring in the community.  Another interesting finding was no difference in infection with other respiratory viruses.

This was a well-designed study. If masks made a difference in the community, it would have been seen in this large a group. And that distinction between protecting the wearer versus others makes no logical sense. Think about it. If wearing a mask doesn’t do a better job of keeping you from getting infected, i.e. keeping the virus from coming in, why would it do a better job of keeping it from going out? In fact, you might expect the opposite, there is less pressure in the inhale than the exhale.

And if the problem is that people who are infected aren’t wearing a mask then wouldn’t we expect to see even higher levels of community spread among the non-mask wearing population–if masks worked, at least on a relative basis the masked people would be better protected from infection than the unmasked ones.  Think about it logically for a minute, if lots of people aren’t wearing masks, and they weren’t at this time in Denmark, according to mask theory there is a tremendous risk of exposure in public.  So there should be an even greater protective effect of a mask protecting the wearer from getting infected, because there is more opportunity for exposure to the virus.  But according to this study, a mask did no better job of protecting the wearer in this hazardous environment than did not wearing a mask.  If anything, this aspect of the trial strengthens the finding.

So wear them if you want to or have to, but don’t for one second think they are offering you or others any significant degree of protection.  And that is my concern about the mask mandate fetish, not only are people being lied to about the science, but they are being misled into thinking they will have a level of protection they won’t.

Join the discussion 83 Comments

  • Doug Young says:

    This makes a Denmark COVID hat-trick over the last week. They saved the 15 million minks (not that they’re long for this world), fought off a draconian forced-vaccination bill, and now, finally released this study that shows what all the other similar pre-COVID RCT masks studies did. Are they accepting new immigrants?

  • Rob says:

    I guarantee Fauci will disavow this entire study. And then go right on about how PCR testing is the greatest thing despite there being less science behind it (and thus its “emergency use only” designation – apparently we have 350 million emergencies a day).

  • M. McRae says:

    Yep. The backlash from the covid propagandists will be swift and severe. They will suddenly rediscover the ability to scrutinize data and methodology — which is conveniently abandoned when reporting on scientific findings supportive of their ideological/political goals. Besides, what some of us have cynically come to understand about scientific studies is that they are only as good what someone says about them, rather than what they actually say.

  • Dennis Boone says:

    As the folks at Powerline have been saying for months, the failure of masks to live up to their safety billing is the “out” that the political class needs to keep telling us it’s our fault for not stopping the virus dead in its tracks. We wear masks, but the virus spreads, so obviously it’s our bad behavior causing all this. We must therefore be spanked with lockdowns. “Thank you, sir, may I have another?”

  • Mike Timmer says:

    Well analyzed. Over at Powerline there persists a sect of belligerents who fail to see the logical catch-22 they paint for themselves. They simply can’t extrapolate from their concession that mask wearers are not protecting themselves from the virus but are protecting everyone else to, if everyone is masked, no one is protected. Only one-way masks could possibly make a difference.

  • Colonel Travis says:

    A study like this is most interesting, but just a cursory look around will tell you masks haven’t done squat. I go into the grocery store, for example, and the PA system blasts out Orwellian “Do your part, please wear and mask so we can stop the spread.” Sorry, where has this happened anywhere in this country? Where I live, mask mandates have been in place since the early summer. Cases are skyrocketing. I see what I see, yet I’m being told what I see isn’t happening. No, sir. Masks work. Trust us and shut up. Don’t show your face. Live in fear.

    We’re involuntarily participating in the biggest con job in world history.

  • SteveD says:

    ‘The headline result–wearing a mask makes no difference in cases or in level of transmission in the community.’

    The strength of the study is that it is large and measures as close to a real-life situation as possible. (there is no way to run a blind mask study) In terms of a recommendation, the result is inconclusive because an absence of evidence does not equal evidence of absence. In other words, they failed to prove that masks make a difference, but they didn’t prove the contrary either, that masks make no difference. The study was set up to prove a difference. (The experimental design intended to prove two means are equivalent, is different and requires more power, than a design intended to prove that two means are different).

    Also, while the differences between the control and the masked group were statistically insignificant, note that the p value was 0.079. That indicates a strong trend toward lower infections in the masked group and would provide more than enough evidence to follow up with a higher-power trial if that were possible. That said, even if the difference between the two groups were later proved to be real, it is exceedingly small (2.1% vs 1.8%). Interestingly, this study matches several mask-influenza studies I’ve seen which all seem to indicate a non-statistically significant trend toward fewer infections upon wearing masks.

  • Christopher B says:

    My suspicion would be that the trend reflects mask-wearers being more likely to do things like hand-washing and especially avoiding situations where transmission is likely. Mask-wearing seems to be the ultimate in social signaling, both by governments looking to ‘do something’ that is really minimally invasive, and people looking for an outward signal that says “I’m concerned”. The side benefit, as Dennis points out, is that if infection rates go down, mask wearers are heroes, and if they go up, well that’s the fault of all those other people doing things wrong.

  • Alex says:

    Thus confirms 100 years of mask studies.

    Fraud and deceit had a place in Dante’s ‘Inferno’. Just saying.

  • SteveD says:

    ‘My suspicion would be that the trend reflects mask-wearers being more likely to do things like hand-washing and especially avoiding situations where transmission is likely.’

    I’m certain hand washing has zero effect on Covid19 transmission. Also, in an RCT, mask use should be randomly determined by the investigator.

  • Drew Haney says:

    You guys had to have a Danish Study to tell you that masks don’t do anything. Wow, that’s even more concerning than the mask issue.

  • Tom Walker says:

    Isn’t the main benefit of mask wearing the prevention of the wearer from infecting others? It seems this study only concludes little to no benefit to the wearer, but does not address the claim that it lessons transmission to others. The reason doctors wear masks in surgery is to prevent infecting the patient, not the other way around.

  • Penny says:

    Hi, in the early 2000’s I worked as a nurse in the Puget Sound as a nurse in some of the most contagious MRSA and VRE units. We took our precautions and kept ourselves safe. Now, one third of our USA populations has tested positive for COVID and we are caving to the “experts and the politicians” because they know much better how to protect us. We are being manipulated by big tech and the unjournalistic media. Both of whom have failed us over the past four years. THEY are responsible for the demise of terminination of our culture and YOUR loved ones.

  • rick8e114d36bf9 says:

    There is a great spiritual deception in America and around the world. Praying for the fraud, corruption and deception to be exposed. For great COVID info, visit –

  • Ron says:

    Mask-wearing seems to be the ultimate in social signaling….. by…….. people looking for an outward signal that says “I’m concerned”.

    Kind of like wearing a “rainbow ribbon” to virtue-signal you are “woke” on “gay rights”?

  • James Alt says:

    It isn’t specifically addressed, but can we deduce that since there is no benefit to the wearer, that there is also no benefit to anyone else? The mask shamers are always saying we wear masks for everyone else.

  • AnthonyD says:

    Excellent review, Mr. Roche; good points from all others.
    The authors gloss over an interesting incidental finding: Some study participants had (likely) COVID patients AT HOME, but few of the former got infected!! “A total of 52 participants in the mask group and 39 control participants reported COVID-19 in their household. Of these, 2 participants in the face mask group and 1 in the control group developed SARS-CoV-2 infection, suggesting that the source of most observed infections was outside the home.” This also suggests that factors other than mere exposure determine the development of disease. Yet Pennsylvania is making people wear masks inside their homes! Recall the Diamond Princess floating petri dish last February: Only 700/3,700 people (~20%) developed documented infection – and they did not have any masks!

  • dirtyjobsguy says:

    In my experience this is a household spread disease for the most part which seems to be be backed up by evidence. I’m pretty sure most if not nearly all people do not wear masks at home. Our brief encounters in most other locations are pretty low risk. Even 100% effective masks would have little impact.

  • Kevin Roche says:

    If you read the full post you will see that the notion that somehow you can separate the effect of wearing a mask in terms of being a transmitter versus being the transmittee doesn’t hold up

  • Kevin Roche says:

    I agree as the post explains, you can’t separate the supposed benefits in terms of the transmitter or the transmittee

  • Paul Ashley says:

    But, but, but government should force people, every single one, to wear and maintain their masks perfectly so we can prolong the life of this virus that inappropriate and inaccurate tests show is deadly to all … well to a narrow group. But no one shoukd ever, ever get sick and feel bad! Losing a job or business is nothing compared to feeling off for a while. /sarc

  • Burt b. says:

    I can see the logic in suggesting that wearing a mask restricts the distance that particles can travel from a sneeze by a mask wearer.
    But so many people don’t wear their mask right, that I wonder how effective even this aspect is.

  • Peter S. says:

    The biggest question is why masks seem to work in a health care setting, preventing medical workers from getting covid, but don’t work in the wild.

    The way I’ve seen them actually used, I can come up with some explanations, but the Danish study used high quality masks and wearers used on average 1.3-1.7 masks a day. That eliminates two of the “in the wild” explanations (poor quality masks and masks worn too long). That still leaves masks improperly worn, frequently adjusting them by touching without hand sanitizing, and pulling them up and down without hand sanitizing, etc. The biggest would be tied to one of the reasons health experts were reluctant to suggest them in the first place; once people are masked, they stop doing the other precautionary behaviors: don’t touch your face, avoid crowds, keep your distance, wash hands frequently, etc.

  • SteveD says:

    ‘The biggest question is why masks seem to work in a health care setting, preventing medical workers from getting covid, but don’t work in the wild.’

    This does not surprise me because health care workers may behave differently than the general populace. They may be more careful to not change their behavior because they are using masks. They may be better trained. People undergoing scientific trials might behave differently. It all has to do with human behavior. In fact, it would not surprise me if masks work better in some countries or cultures than others.

    BTW, from conversations with doctors and other health care workers, I was under the impression that masks ideally need to be changed several times per day.

  • Alex says:

    Peter S,

    ‘The biggest question is why masks seem to work in a health care setting, preventing medical workers from getting covid, but don’t work in the wild.”

    From what I’ve learned reading my fair share of mask studies (yes, I even read Dr. Kellogg’s 1920 study. It’s G-r-r-reat! Sorry, couldn’t resist), is the following:

    Hospital settings can’t be replicated in community settings. Two different universes. Health care professionals are trained on how to use masks and are used to the environment to which they exist. Proper handling of a mask right down to disposal is essential for effectiveness. You will NEVER get that in a community setting because, well, we’re not robots and nor should we be ‘trained’ into changing our social habits for it.

    In addition, there are also multiple studies that in fact show masks may not even be effective (outside N-95 especially) in hospital settings.

    Public officials are conflating the two to push this agenda.

    I’m sure Kevin can add to this better than me.

  • Carl Pham says:

    I think you’re missing the point. Many people have emphasized that the point of mask-wearing is not really to protect yourself — and this study does indeed suggest that it doesn’t work well that way — the point is to slow the spread in the community. And that was not tested in this study. The only way to test that would be to compare two very similar populations, one of which almost all wore masks, and the other of which did not. In the case of this study, most people did not wear masks, so it’s not possible to test the effect (if any) on community transmission rates.

    And arguing that “its obvious” or “just think about it” to say the protective effect should be symmetric is pure speculation. It’s not hard to construct arguments in the other direction which are just as “obvious” to people who want to come to a different conclusion.* The only way you would know is to do the actual experiment — which this study, unfortunately, did not do.
    * The simplest possible one: a mask cuts down the amount of passed virus by 25% in both directions, and the canonical sick person emits 50% more virus than is needed to infect someone else. Under this scenario, if most people don’t wear a mask, so on average there’s only one mask protecting a healthy person from a sick person, there’s no reduction in community transmission, since when a health person meets a sick person even if he’s wearing a mask he gets 0.75 * 1.5 = 125% the amount of virus needed to fall ill. But if almost everyone *is* wearing a mask, then on average when a healthy person meets a sick person he gets 0.75 * 0.75 * 1.5 = 84% of the amount of virus needed to fall ill, and community transmission drastically falls, since you need the much rarer event of two unmasked people meeting.

    This is a completely trivial model, and the point is not to construct a realistic one, but to observe that it’s easily possible to construct non-idiotic models in which mask wearing has a marginal benefit if hardly anyone does it, but a significant benefit if everyone does it. There are plenty of “critical mass” situations like that in percolation type problems, which is what community transmission is.

  • Kevin Roche says:

    Okay, I keep trying to make this point, no matter how much difference a mask might make in an individual encounter, no one has any evidence that it is 100% effective regardless of whether it is worn by the infector, infectee or both. In reality, the types, and as worn by most people, they aren’t very effective. But even if 90% effective, across large numbers of encounters, they aren’t going to make a difference. People just want to ignore what is right in front of their faces (maybe its blocked by a mask) and look at the population level data where there are mask mandates and/or very, very high levels of mask wearing and tell me how those case curves are any different than where there isn’t. And that inoculum, or viral dose theory that was making the rounds a few weeks ago has been debunked by virologists and infectious disease doctors. As far as I can tell from reading the research, an infectious dose is so small compared to typical viral exhalations that you can understand why any significant leakage undermines the value of a mask for an infector or infectee. But thank you for your comment, which at least is well thought out and rationally presented compared to most of what I have gotten in response to the post.

  • Papaw says:

    So, we negate scientific studies… with opinions?

    How does someones opinion cause a scientific study to become untrue or less valid or less reliable?
    Once again, masks are proven worthless and the political Sunday scientists of the world rush to negate it with their worthless words.

    Has anyone ever pointed out that a mask that is worn for 2 hours has a horrific viral load impregnated into it, that begins to eject viral particles at a constant rate as the viral load increases? Nobody? Huh, maybe that’s why people who wear masks have an infection rate the exact same as people that don’t.

  • Dave says:
    Perhaps consider this before jumping to conclusions

  • Kevin Roche says:

    i read and posted on that study at the time

  • Michael Dunn says:

    ‘The biggest question is why masks seem to work in a health care setting, preventing medical workers from getting covid, but don’t work in the wild.”

    900 cases in the last two weeks at Mayo Clinic. Presumably, they have the best PPE around. I guess they all forgot to wear their masks!

  • SteveD says:

    ‘The reason doctors wear masks in surgery is to prevent infecting the patient, not the other way around.’

    Yes, to prevent opportunistic bacteria from their digestive tracks from directly infecting an open wound. However Covid19 is not an opportunistic bacteria from a doctor’s digestive track which likes to infect open wounds.

  • Kris Hubby says:

    It seems you are right about criticism, but it’s reached much scarier censorship. facebook doesn’t want me to share this actual journal article or your commentary, it labels it “False Information”! What? The Annals of Internal Medicine journal is now “false information”?

  • Al Benoit says:

    It appears to me after having gotten a bit familiar with a number of studies on this that masking has a quite small if any benefit in terms of preventing viral transmission and thereby infections. How about a study on vitamin D, and graphing the incidence and severity/duration of infection vs the amount of vitamin D in the blood? Maybe more bang for buck there, than trying to get very high compliance with uncomfortable mask mandates.

  • Finally an unbiased review of the Danish study. I had been struggling to find one on the Internet as almost all of them (written by know-it-all journalists) have already jumped into the conclusion that it is inherently flawed without bothering to read it in the first place, citing all sorts of unfounded accussations. Indeed the study has been vastly politicized by journalists and politicians to prevent it from interfering with their agenda. Thank you for this article and for your open-mindedness.

  • Health Care Worker says:

    The authors of this study write the following, because they expected the study would be quoted incorrectly:

    ‘The findings, however, should not be used to conclude that a recommendation for everyone to wear masks in the community would not be effective in reducing SARS-CoV-2 infections, because the trial did not test the role of masks in source control of SARS-CoV-2 infection.’

    This is written in the first paragraph of the discussion, and it makes perfect sense because basically no one was masked except half of the study participants.

    So, we don’t know from this one way or the other. Speculating is not helpful. Do we then, for example, want to have everyone in hospitals stop wearing masks? Sounds like a bad idea to me.

  • Kevin Roche says:

    I don’t know how recapitulating the exact findings is quoting the study incorrectly. If I understand your point, and everyone has seized on this, somehow masks magically don’t stop the virus from coming in but do stop it from going out? That is the only logical explanation for the distinction you are trying to draw. And somehow, in places in the US where we have 90% plus mask compliance, case surges are occurring. Hmmmm. And as I have repeatedly said, i am not saying people shouldn’t wear a mask if they think it will protect them, they should just not be under any illusions that there is significant evidence that it will.

  • arch1 says:

    Kevin, you state
    1) “If masks made a difference in the community, it would have been seen in this large a group.”
    a. How can you conclude this without any quantitative analysis and by considering only the size of the study group? If the non-control group comprised (say) 1% of the potential transmission community, it seems implausible that any effect on overall community transmission would have been discernible, even if mask wearing made a significant difference in outbound infection rate. If it comprised a large fraction of the relevant transmission community (which is the scenario most relevant to real world policy questions), that is a different matter.
    2) “The headline result–wearing a mask makes no difference in cases or in level of transmission in the community”
    a. I think that you are basing this claim on claim #1, which (as pointed out above) you haven’t supported.
    b. Also, as a matter of integrity in reporting, you should have been upfront about the fact that this “headline result” conflicts with the study authors’ own statement as to the limitations of their study. The fact that you didn’t do this raises a red flag for me – not a conclusory one, but one which does increase my level of skepticism when reading this post..
    i. I imagine your position is that the authors were required to make such a statement in order to get their paper published, or that they were mistaken about (or deliberately misled readers about) the limitations of their own study. That doesn’t change the fact that your readers deserve to be clued in when your claims go well beyond those of the paper you are analyzing, or justify your failure to do so.

  • Kevin Roche says:

    this was a randomized trial of the effect of mask wearing on transmission in the community. What else would the headline finding be about? I keep asking people to explain how masks magically stop outflow but not inflow of the virus, when in fact exhalation pressure, and especially when talking, coughing, sneezing, is stronger than inhalation pressure. And yes the authors were pressured to stick to the usual recommendations to get this published.

  • Doug Young says:

    One of the authors of the Danish mask study, Dr. Henning Bundgaard, will be live on a podcast produced by a friend tomorrow at 9 am Pacific (11 am CST) here:

    Ask him some questions! I think this will be middle of the road show in general, so he will have some opposition in studio.

  • arch1 says:

    “this was a randomized trial of the effect of mask wearing on transmission in the community. What else would the headline finding be about?”
    Kevin, at a minimum your readers should be able to expect your headline finding to actually BE a finding, rather than a NON-finding, of the paper. In this case, your “headline finding” is not only NOT a finding, but rather it is a statement which, if true, would imply that universal mask usage would be ineffective in reducing COVID infection rate – a conclusion SPECIFICALLY DISAVOWED by the study authors. (If you disagree with this perspective, you may wish to contact the study authors to get their opinion on whether your headline finding is a fair summary of the main takeaway from their paper).

    Also, can you please respond to my first point, which I now realize is even stronger than I suggested above? If masks made a difference in the community, how exactly would this have been seen in the Danish Mask study? **Please be specific as to the different results you would have expected to see if that were the case**.
    (I believe the correct answer is that even if mask-wearing significantly reduced outbound transmission, AND the non-control group comprised a large fraction of the relevant transmission community, this study would STILL not have detected such an effect, simply because it wasn’t designed to do so. After all, the study only compared the infection rate of the control group vs the non-control group; how could any conceivable result of such a comparison establish that mask usage made a difference in overall community infection rate? The study simply wasn’t designed to detect such an effect.)

  • Kevin Roche says:

    I learned a long time ago the pointlessness of discussion with people who have their minds made up

Leave a comment