You may or may not have recently read the results of the November 2020 Danish face mask study. This was published November 18th, in the Annals of Internal Medicine, a much less prestigious journal than say the New England Journal of Medicine, Lancet, Nature or JAMA. You’d figure the only randomized study of its kind would be published with more pomp and circumstance? For goodness sake, we just had a political election largely swayed by COVID-19. Doesn’t everyone want to “follow the science”??? Except higher profile journals did not want to publish it… The Danish researches had their results accumulating dust since July, but couldn’t find an accepting journal until November. Well, let’s first look at what the study showed…
A bit of perspective, the Danes were noted to be extremely efficient in stopping and limiting the transmission of COVID-19. Denmark was the 2nd county in Europe to announce a lockdown, doing so before it even reported its first death. Their positivity rate was lower than many other countries. While Danish restrictions were never as stringent as most of Europe and surely not the US, their citizens seemed to adhere closely to recommendations.
From April to June 2020, wearing a face mask in public was not mandated in Denmark. Of course, persons with symptoms or diagnosed SARS-CoV-2 infection were quarantining. Cafes and restaurants were closed until May 18, 2020, midway through the study. 6,024 Danish citizens participated to see if they would develop antibodies against the SARS-CoV-2 virus while living-out their normal daily routines. Antibody testing, as opposed to symptoms, were chosen as the outcome since many COVID-19 positive people are asymptomatic. Half were assigned to act normally and not wear a face mask when in public. The other half needed to wear a surgical-grade mask, with a 98% filtration rate, compatible with the N-95. Participants in both groups reported having spent on average of 4 ½ hours per day outside the home, trying to maintain social distancing and other standard hygiene measures. Of the 4,862 people who completed the study, both non-mask wearers and mask wearers cited they were more cautious than normal due to their participation in the study.
The glaring critique is that of the mask-wearers, 46% wore the mask exactly as they were supposed to, while 47% “predominately as recommended”. (What this exactly means is unclear.) 7% didn’t really wear the mask and they were excluded from the results. This is always a problem with any study. We can’t lock-up humans like rats, as occurred in the Stanford prison experiment of the 1970s. In every experiment, some of the volunteers don’t abide by the exact rules, or cross-over to the other arm of the study, and this is a known and accepted limitation of research. But of participants who complied, most used more than 1 mask a day, changing them out as recommended. This is better than how most mask wearers around the world perform.
In the end, 1.8% of the mask group and 2.1% of the non-mask group developed a SARS-CoV-2 infection over the month they were studied. There was no significant difference in group interactions. In just looking at mask group wearers who wore them, “exactly as instructed”, 2% developed an infection, again compared to 2.1% of people not wearing a mask. In dissecting the data, the researchers could not find a constellation of characteristics where face masks wearing was more effective in reducing infection. The rate of COVID-19 positive antibodies in the general Danish population, outside of the study, was similar at about 1.9% . In summary, this community based, prospective randomized controlled trial, suggests mask wearing did not reduce the incidence of SARS-CoV-2 infection.
Now this study cannot be extrapolated exactly to the US, because these Danes were mingling with other people not wearing masks. In the US, we are all supposed to be wearing masks, so there may be less baseline SARS-CoV-2 particles frolicking in the US air than in Denmark’s. But the Danes are noted to be much more disciplined than Americans, with better social distancing and hand hygiene. Of note, SARS-CoV-2 could also be contracted across the mucous membranes of the eyes, and none of the participants in the study were wearing specialized eye covering, per se. Though, there appeared to be no difference in normal eye glass wearing either. But again, no country in the world is mandating wearing face shields.
Of great interest, over the course of the study period, Denmark was on full lockdown until half way through, around May 18, 2020. But the rate of new infection acquisition was the same, whether during the period of lockdown or after. Lock-down measures did not seem to affect the transmission. The authors of the study concluded that mask wearing did not diminish the rate of disease transmission. Whether you wore a mask in the study, didn’t wear a mask in the study, or weren’t a part of the study, you had about a 2% chance of contracting SARS-CoV-2 in Denmark at that time. As mentioned in prior RMBSI blogs, this likely was because SARS-Cov-2 may be small aerosolized particles that penetrate masks.
So why should we be worried about this study? Well, wearing masks and quarantining may not affect the rate of COVID-19 transmission. That’s a huge problem. But we’re already seeing this in America, where despite masks, lockdowns, social distancing, mass media warnings… our rates are still increasing. But purely from the academic sense, we should be really worried for different reasons.
This is the only prospective randomized study assessing the effectiveness of masks for SARS-CoV-2. Prospective in that the study design was formulated ahead of time, and was carried out without knowing how the chips would fall. Retrospective studies, in contrast, look at data already collected, and are at risk of extreme bias in analysis. Randomized trials are conducted by dividing patient populations into two groups, where one group receives the intervention to be studied while the other does not. Examining the differences between groups in these types of trials has ushered in an era of evidence-based medicine that continues to guide clinical practice on a daily basis. The only component this study lacks was “blinding”. It’s best to have the studied participants not know (be blinded) to which group, experiment or control, they belong. “Blinding” usually prevents participants from providing biased outcomes… if people know they’re receiving the treatment or know they are receiving placebo, their outcomes may be influenced one way or another. However it’s simply not feasible to “blind” the mask wearers from knowing they’re wearing a mask. On the other hand, the researchers could have been “blinded” when analyzing the results, but that likely would not have changed anything.
This study involved over 6,000 people, which is a large sample size for most medical research. For comparison, the 1998 case series published in Lancet which linked the MMR vaccine to autism studied 12… yes just TWELVE… children, and has led to millions of people world-wide refusing vaccines despite the paper’s later retraction and public humiliation. In comparison to influential neurosurgery research, the 2005 Stupp paper formulating the current standard treatment protocol for glioblastoma multiforme, studied just 573 patients. One of the most influential brain cancer papers the world has, studied 1/10th the number of participants in this Danish study. So this is a pretty well designed study. That part can’t be scrutinized too much.
Most people’s problem is that it did not offer the results for which they may have hoped. Masks must be beneficial… right? That Fauci guy said so. Nerds with glasses don’t lie… But those people probably should have read RMBSI’s October blog on masks? Most detractors so far have summarized that this study was “inconclusive”, and not “negative”. This is incorrect and tries to be confusing with semantics to lay people. A “negative” study means that a statistically significant conclusion was not found in either direction. The Danish mask study intended to prove that mask wearing would reduce SARS-CoV-2 transmission, and it is “negative” because wearing the mask did not demonstrate reduced infection outcome. There is nothing inconclusive about this result.
Other skeptics cite that this study did not assess whether wearing masks prevented spread to others. This is partially true. The study specifically looked a disease contraction to the study participants, and not disease transmission to people outside the study. Criticizing this outcome measure insinuates: if you wear a mask, you really still have no hope to avoid SARS-CoV-2 yourself, but it’s more important that everyone else around may be at less risk? That argument has less pragmatic sense, other than being a nice altruistic gesture. As a population, we expect mask wearing primarily protects us as its main function. As an analogy, if I take an ibuprofen because I have a headache, the benefit of ibuprofen is that it stops my headache or prevents it from worsening. I don’t wish anyone ill-will, but I’m less concerned if my taking ibuprofen doesn’t stop my neighbor’s headache. But this Danish study indirectly did address masks and disease transmission, because everyone in Denmark still had about a 2% risk of contracting the virus. Even though it wasn’t a primary outcome measure, those in the face mask arm didn’t seem to diminish disease transmission either. Of course, directly studying disease transmission is also operationally untenable. It would require the researchers testing every person the study participants came in contact with twice a week for long periods of time to see if the contacts became positive. This is unrealistic and… “spoiler alert”… the vaccine trials didn’t study whether vaccines stop disease transmission either.
Consequently, the most prestigious journals in the world refused to publish the data. We expect, out of anyone, that they always “follow the science” wherever it leads (assuming it’s a worthy research). That’s the whole point of prospective randomized controlled trials, and why we weigh them so heavily. Of course, if you think NEJM only publishes the most prestigious work of mankind, they included an article in 2011 about wild armadillos as natural carriers of leprosy in the southern US… how would the world have survived without that gem of an article? When the results nearly pulselessly scavenged their way to the Annals of Internal Medicine, social media platforms censored them. Social media purports freedom of speech yet sometimes acts selectively?
The Danish mask study reminds us a bit of the SPORT trials of the early 2000s. A group of well published researchers had spent a majority of their careers trying to prove that surgery for lumbar spine disease was futile. A series of Spine Patient Outcome Research Trial (SPORT) papers were organized in a prospective randomized fashion to look at disc herniations, spondylolisthesis, and stenosis. 528 eligible participants were studied for disc herniation, 607 for spondylolisthesis, and 654 for stenosis. Spine surgeons globally were apprehensive about the results, fearful that spinal surgery may not be any better than conservative treatment. Instead, surgical treatment was found to be twice as effective as nonsurgical care. And just like the Danish mask study, the participants did not adhere exactly to the treatment protocols as designed, making the data a little cloudy. Detractors claimed the study just showed “no difference” in outcomes between conservative and surgerized groups. Yet when assessing outcomes based on whether the patients received surgery or not, there was no debate. If you haven’t gotten better in 3 months of conservative care, the odds of improvement from further nonoperative treatment may be low and these patients tend to do better with surgery. However the glaring difference with the SPORTs trials were that they were published in JAMA and the New England Journal of Medicine. The results were not hidden, but published because science isn’t supposed to have an agenda.
The answer to solving COVID-19 is not suppressing data like that acquired in the Danish mask study. If some scientists don’t believe the results, they should form another prospective randomized study and see what happens. Use even more patients from many different countries and blind the researchers. Also study the close contacts of the participants to better assess transmission over contraction. That’s what occurs following any other important trial. Replicate or contradict. But trying to bury the research under sand, is not becoming. Continuing to urge public mask wearing as the primary way to defeat COVID-19, may be the greatest public health Ponzi scheme of our generation. If masks don’t work, we need to change course and start figuring out what does. Some would argue what’s the harm in wearing masks? There probably isn’t much, other than the false sense of safety wearing them incurs, which may lead to more risky behaviors that promote transmission. This is what happened with the advent of seatbelts in cars… we all just drove faster.
Regardless of the outcomes or validity of this trial, the new direction of medical research censorship should worry all of us.
Bundgaard, H et al. Effectiveness of adding a mask recommendation to other public health measures to prevent SARS-CoV-2 infection in Danish Mask Wearers. Annals of Internal Medicine. 2020.
Weinstein J et al. Surgical versus nonoperative treatment for lumbar disk herniation. JAMA. 2006
Weinstein J et al. Surgical versus nonsurgical treatment for lumbar degenerative spondylolisthesis. NEJM. 2007.
Weinstein J et al. Surgical versus nonsurgical treatment for lumbar spinal stenosis. NEJM. 2008.