There’s a “study” that’s gone viral in the COVID-19-denying/minimizing, antimask, anti-public health interventions, antivaccine crankosphere by someone named Baruch Vainshelboim entitled Facemasks in the COVID-19 era: A health hypothesis. For whatever reason, even though it was published in January, only now does it seem to be going viral, leading me to seeing it incessantly popping up on my Twitter timeline, with cranks emailing me a hearty “Ha! I told you so!” while readers send me the link to the article asking me to discuss it. Before I do just that, let me just note that that last word in the title ought to give you a hint that, yes, this not-study was published in Medical Hypotheses. For those of you not familiar with this highly dubious journal, I present a brief trip down memory lane before I delve into Vainshelboim’s article.
Medical Hypotheses: A brief history
Does anyone remember Medical Hypotheses? I certainly do. It’s a fringe journal published by Elsevier—yes, that Elsevier!—that has a long history of publishing, if you’ll excuse the term, complete and utter bullshit under the banner of “speculative” science. Examples are many, and I’ve written about some of them, for example, the bogus claim that aluminum in antiperspirants cause breast cancer appears to have originated in an article in Medical Hypotheses.
That’s nowhere near all, though. Back in my early days of blogging, Medical Hypotheses published antivaccine articles, for example an article by Len Horowitz published in 2001 “hypothesizing” that genetic recombinations with oncogene activation by the hepatitis B vaccine involving simian viruses (yes, SV40, the simian virus that antivaxxers blame for an “epidemic of cancer” due to the polio vaccine) that likely infected polio vaccinated blood donors to the initial hepatitis B vaccine trials conducted on gay men in New York City and Ugandan Blacks in the early to mid-1970s had resulted in the AIDS epidemic. Another notorious example was an article by antivaccine activists Mark Blaxill, Lyn Redwood, and Sallie Bernard in 2004 laying out the “medical hypothesis” that mercury in vaccines causes autism. Perhaps the most (in)famous antivaccine article published by Medical Hypotheses was written by the père et fils duo of antivaccine cranks Mark and David Geier in which they proposed using Lupron (a drug that shuts down sex hormone production) to treat autism, the worst autism quackery I had yet seen at the time. They based their “Lupron protocol” on a claim that testosterone formed “sheets” that prevented the chelation of mercury and thus led to higher levels of mercury (from vaccines, of course) in the brain. Where did they get this idea? It was from an observation that testosterone could bind mercury when precipitated out of a high temperature benzene solvent. I kid you not.
On its website, Medical Hypotheses describes itself thusly:
Medical Hypotheses is a forum for ideas in medicine and related biomedical sciences. It will publish interesting and important theoretical papers that foster the diversity and debate upon which the scientific process thrives. The Aims and Scope of Medical Hypotheses are no different now from what was proposed by the founder of the journal, the late Dr David Horrobin. In his introduction to the first issue of the Journal, he asks ‘what sorts of papers will be published in Medical Hypotheses? and goes on to answer ‘Medical Hypotheses will publish papers which describe theories, ideas which have a great deal of observational support and some hypotheses where experimental support is yet fragmentary’. (Horrobin DF, 1975 Ideas in Biomedical Science: Reasons for the foundation of Medical Hypotheses. Medical Hypotheses Volume 1, Issue 1, January-February 1975, Pages 1-2.). Medical Hypotheses was therefore launched, and still exists today, to give novel, radical new ideas and speculations in medicine open-minded consideration, opening the field to radical hypotheses which would be rejected by most conventional journals. Papers in Medical Hypotheses take a standard scientific form in terms of style, structure and referencing. The journal therefore constitutes a bridge between cutting-edge theory and the mainstream of medical and scientific communication, which ideas must eventually enter if they are to be critiqued and tested against observations.
Sadly, it didn’t turn out as Horroborin had intended. Instead, Medical Hypotheses became a repository for crank scientific “hypotheses,” such as the aforementioned antivaccine and cancer pseudoscience. I could go through even more examples of such nonsense published by Medical Hypotheses, including “hypotheses” claiming that masturbation is a treatment for nasal congestion or that high-heeled shoes are linked to schizophrenia, and even pseudoscience by cancer quack Stanislaw Burzynski, but why belabor the point? The last time I paid enough attention to Medical Hypotheses to write an actual post about it was a decade ago when, after the journal had published an article by the granddaddy of HIV/AIDS denialism, Peter Duesberg, an article so bad and irresponsible that Elsevier actually retracted it, making it the only article I’m aware of that was too much pseudoscience even for Medical Hypotheses. Elsevier told then-editor Bruce Charlton, in essence, “Enough’s enough,” firing him. Oddly enough, in his misguided defense of Medical Hypotheses and its reliance on “editorial review” for its articles rather than peer review (and also of the “censorship” of Andrew Wakefield), Dr. Michael Fitzpatrick, normally a sound voice about medicine and science, inadvertently revealed just why the journal was so bad and its continued indexing by the National Library of Medicine on PubMed was dangerous:
The notion that organisational methods of censorship and repression are the appropriate response to influential currents of pseudoscience has unfortunately become widely established. This is well illustrated by the recent controversy surrounding the journal Medical Hypotheses – published, like Vaccine and the Lancet, by Elsevier. Under its founding editor David Horrobin, and his successor Bruce Charlton, Medical Hypotheses has rejected the procedures of peer review now standard among academic journals in favour of a policy of selection by the editor, according to what he considers interesting, provocative, entertaining. The result is an eclectic mixture of science and pseudoscience, sense and nonsense.
For example, Medical Hypotheses has published articles suggesting that autism may be caused by iron, mercury, Vitamin D deficiency, electromagnetic fields, Lyme disease, the antibiotic Augmentin and improved childhood hygiene. I strongly suspect that these theories are as speculative and as insubstantial as the MMR-autism thesis, but this is no grounds for denying their authors a platform, particularly when it is clear that they have not been given the kite mark of peer review.
Medical Hypotheses and its editorial policy were safe – indeed the journal has flourished under Charlton’s editorship – until he published an article by Peter Duesberg, the notorious retrovirologist who rejects the theory that HIV is the cause of AIDS. This brought Charlton into conflict with one of the most powerful scientific advocacy lobbies, the AIDS establishment, which ranks second only to the climate-change crusade when it comes to trying to suppress its critics, who are stigmatised as ‘denialists’ of doctrinal orthodoxy.
This was almost 11 years ago. Doesn’t it sound an awful lot like COVID-19 contrarians when social media companies try to deprioritize and deplatform their dangerous pseudoscience, conspiracy theories, and disinformation? In any event, as I said at the time, I suppose you could describe the journal that way, as an “eclectic mixture of science and pseudoscience, sense and nonsense,” although the pseudoscience and nonsense have long seemed much more prominent. Such a description, however, was not what I considered a ringing endorsement of a journal that had represented itself as being every bit as serious as a peer-reviewed scientific journal, and whose “speculations” and “hypotheses” can in the age of COVID-19 be downright dangerous and deadly.
Which brings us back to Vainshelboim’s article.
A dangerous “medical hypothesis” in the age of COVID-19
When I first saw Vainshelboim’s antimask article spreading around the COVID-19 crankosphere in social media, my first temptation was very similar to Mark Hoofnagle’s sentiment:
Like Mark, I was tempted just to point to Medical Hypotheses and say that, “peer review” or no “peer review” enforced on the journal by Elsevier 10 years ago after Bruce Charlton was fired, this article is very much of a piece with the long, undistinguished history of the journal of publishing ridiculous and sometimes dangerous pseudoscience. Unfortunately, that’s not enough. Why? Just look at Twitter:
It’s not a study!
Look at this Tweet by rising star in the antimask, antivaccine, COVID-19 minimizing/denying conspiracy movement, Naomi Wolf:
As you can see, being an ostensibly peer-reviewed and PubMed-indexed journal published by Elsevier, a very large and established publisher of peer-reviewed biomedical journals, makes the bad science published by Medical Hypotheses seem credible to laypeople (and even some scientists) who don’t know the journal’s long and sordid history. So we have to look at what Vainshelboim wrote. It doesn’t matter if, like this article, it’s nothing more than an opinion piece masquerading as a review article. Cranks interpret it as a “study,” when it is not, and use it to bolster their views.
It helps that there is a lot in the article that is just plain wrong, and risibly so. Our friend the Health Nerd points out a couple of examples of this sort of thing:
I also can’t help but note that these quotes from Anthony Fauci come from a New England Journal of Medicine article published March 26, 2020. This might as well be ancient history in the context of the pandemic, as over a year ago there was considerably more uncertainty about case fatality rates (CFRs, the mortality in people diagnosed with COVID-19) and infection fatality rates (IFRs, the mortality in people infected with SARS-CoV-2, the coronavirus that causes COVID-19, including asymptomatic infections). It’s intellectually dishonest to use a quote from so long ago in this fashion, when there is much more up-to-date information now to be had.
Vainshelboim also invokes a favorite “blame the victim” conspiracy theory about COVID-19 that states that only the old and infirm need to worry about dying from the disease:
In addition, data from hospitalized patients with COVID-19 and general public indicate that the majority of deaths were among older and chronically ill individuals, supporting the possibility that the virus may exacerbates existing conditions but rarely causes death by itself , .
Neither of the articles cited actually supports this claim. The first is an early article in JAMA from last year that did show that comorbidities and age greatly increase the chance of serious disease or death, and the second is by John Ioannidis, who has a demonstrated record of using bad epidemiology to downplay the lethality of COVID-19. In any event, this is nothing more than the COVID-19 denial trope that I like to call the “6% gambit,” a conspiracy theory that claims that “only 6%” of the deaths from COVID-19 are actually from the disease because 94% of the death certificates list comorbid conditions.
Now let’s look at Vainshelboim’s “hypothesis” (“hypotheses,” actually, as there are four):
Of note, hyperoxia or oxygen supplementation (breathing air with high partial O2pressures that above the sea levels) has been well established as therapeutic and curative practice for variety acute and chronic conditions including respiratory complications , . It fact, the current standard of care practice for treating hospitalized patients with COVID-19 is breathing 100% oxygen , , . Although several countries mandated wearing facemask in health care settings and public areas, scientific evidences are lacking supporting their efficacy for reducing morbidity or mortality associated with infectious or viral diseases , , . Therefore, it has been hypothesized: 1) the practice of wearing facemasks has compromised safety and efficacy profile, 2) Both medical and non-medical facemasks are ineffective to reduce human-to-human transmission and infectivity of SARS-CoV-2 and COVID-19, 3) Wearing facemasks has adverse physiological and psychological effects, 4) Long-term consequences of wearing facemasks on health are detrimental.
This is a non sequitur. It does not follow from the observation that oxygen supplementation is used to treat acute and chronic respiratory diseases and conditions that masks are harmful. In addition, he cherry picks his references. The first two references is the World Health Organization’s (WHO) statements from April and June on the use of facemasks, reminding me how I, for one, have long bemoaned the epic screwups in messaging over masks early in the pandemic. For example, the World Health Organization didn’t change its recommendations that downplayed the importance of wearing masks until last June. This was particularly infuriating, given that, as was noted in March 2020, mask use had always been advised as part of the standard response to being around infected people, especially for people who may be vulnerable, WHO officials were wearing masks during their news briefings, and since the SARS experience in 2003 that health officials in many high-risk Asian countries had advised wearing masks. She also pointed out that the messaging at the time also recommended that people who were sick should wear masks to protect others, further noting that, given the increasing evidence of asymptomatic transmission of SARS-COV-2, the coronavirus that causes COVID-19, such advice would imply that everyone should wear masks when around others.
The third reference that “masks don’t work” is a paper from 2015 beloved of antimaskers the world over. It was a randomized trial of medical masks, cloth masks or a control group (usual practice, which included mask wearing) among healthcare workers in Vietnam, the main outcome being influenza or respiratory viral disease that found that cloth masks were associated with the highest rates of respiratory disease, which Vainshelboim misrepresents in the same way antimaskers always misrepresent this study:
With respect to cloth facemask, a RCT using four weeks follow up compared the effect of cloth facemask to medical masks and to no masks on the incidence of clinical respiratory illness, influenza-like illness and laboratory-confirmed respiratory virus infections among 1607 participants from 14 hospitals . The results showed that there were no difference between wearing cloth masks, medical masks and no masks for incidence of clinical respiratory illness and laboratory-confirmed respiratory virus infections. However, a large harmful effect with more than 13 times higher risk [Relative Risk = 13.25 95% CI (1.74 to 100.97) was observed for influenza-like illness among those who were wearing cloth masks . The study concluded that cloth masks have significant health and safety issues including moisture retention, reuse, poor filtration and increased risk for infection, providing recommendation against the use of cloth masks .
It’s always been questionable how applicable this study is to the wearing of masks outside of a healthcare institution, but here’s the misrepresentation. One huge problem with this study is that the “usual practice” arm involved frequent wearing of medical masks, not “no masks,” making the interpretation of this study as anything other than showing that cloth masks don’t work as well as medical masks to slow the spread of respiratory viruses difficult. (I also note that no one recommends the use of cloth masks over medical masks in a medical setting, except in conditions of extreme shortages of masks.) In other words, there was no true “no mask” control group, contrary to Vainshelboim’s spin on the study. His description of the study is intellectually dishonest in the extreme.
Elsewhere, Vainshelboim makes some very incorrect assumptions, for example:
If you want to get an idea of how desperate Vainshelboim is to make his point, laugh at this paragraph:
The adverse physiological effects were confirmed in a study of 53 surgeons where surgical facemask were used during a major operation. After 60 min of facemask wearing the oxygen saturation dropped by more than 1% and heart rate increased by approximately five beats/min . Another study among 158 health-care workers using protective personal equipment primarily N95 facemasks reported that 81% (128 workers) developed new headaches during their work shifts as these become mandatory due to COVID-19 outbreak. For those who used the N95 facemask greater than 4 h per day, the likelihood for developing a headache during the work shift was approximately four times higher [Odds ratio = 3.91, 95% CI (1.35–11.31) p = 0.012], while 82.2% of the N95 wearers developed the headache already within ≤10 to 50 min .
A 1% decrease in oxygen saturation? I marvel that the authors could reliably pick up that tiny a drop in saturation with a pulse oximeter! Oxygen saturation fluctuates by more than that just from regular activity, and Vainshelboim neglects to note that even the authors of the study concluded that “early change in SpO2 may be either due to the facial mask or the operational stress.” I also laughed at the authors writing, “Since a very small decrease in saturation at this level, reflects a large decrease in PaO2, our findings may have a clinical value for the health workers and the surgeons.” As far as oxygen delivery goes, it is the saturation of the hemoglobin molecules in your red blood cells that matters far more than anything else, as the amount of oxygen dissolved in your blood is very small by comparison. This is basic physiology, and if your saturation is between 95-100%, a change of 1% is basically meaningless, statistical noise. Yes, if you have a 99% oxygen saturation on room air and start breathing 100% oxygen by a non-rebreather, you will bump your saturation up to 100% (a mere 1% increase) while vastly increasing the partial pressure of oxygen in your blood, but it will make very little difference in the actual oxygen delivery to your cells.
As for headaches, I could be sarcastic and mention that I recently wore an N95 for over 10 hours in the operating room, and the only headache I suffered was a metaphorical one—briefly—when one of the cases turned out to be more difficult than I had anticipated. In reality, I’ve never suffered a headache from wearing an N95, except when the straps holding it on my head were too tight or the splash-proof safety glasses were pressing against the wrong part of my face. My sarcastic recounting of personal anecdotes aside, Vainshelboim again does not tell the whole story. This study was about the wearing of all forms of personal protective equipment (PPE), not just N95 masks. That includes eye protection, such as goggles or face shields, gowns, etc. The study actually goes along with my personal anecdotal experience, though, as this quote shows:
Of the 158 respondents, 128 (81.0%) reported de novo PPE‐associated headaches when they wore either the N95 face mask, with or without the protective eyewear. All respondents described the headaches as bilateral in location. Figure 3 illustrates the summated anatomical localization of headaches (marked by the study participants) and the corresponding frequencies of occurrence according to PPE subtype. Interestingly, the location of the discomfort experienced by the participants corresponded to the areas of contact from the face mask or goggles and their corresponding head straps. The majority [112/128 (87.5%)] reported a sensation of pressure or heaviness at the affected sites, with some [15/128 (11.7%)] characterizing it as a throbbing or pulling pain [1/128 (0.8%)].
Unlike the impression Vainshelboim tries to relate, PPE-associated headaches reported have nothing to do with lack of oxygen due to N95 masks and everything to do with pressure from the straps of the goggles or face shield. Presumably, most cloth masks don’t produce such pressure unless they’re really, really tight (in which case the mask wearer should adjust the mask or get a different one), and even N95 masks don’t produce quite that pressure either (although sometimes I do feel sinus pressure from the mask digging into my nose and cheeks). It’s also an intellectually dishonest trick that he plays, cherry picking data from N95 masks in terms of oxygenation and headaches and generalizing it to all masks.
I could go on and on and on. Doing what Medical Hypotheses authors do, Vainshelboim has cherry picked an impressive array of studies suggesting that masks don’t slow the spread of COVID-19 and cause all sorts of horrific problems. Neither contention is true. There is a considerable and growing body of evidence supporting the efficacy of mask wearing for slowing the spread of COVID-19.
Personally, I like to refer to a much better review article published in January in PNAS on the efficacy of masks. While acknowledging the lack of certainty, it reviews a wide variety of evidence regarding masks and COVID-19 and other respiratory viruses, including epidemiological, ecological, and modeling studies, as well as the limited number of randomized clinical trials, and concludes:
Our review of the literature offers evidence in favor of widespread mask use as source control to reduce community transmission: Nonmedical masks use materials that obstruct particles of the necessary size; people are most infectious in the initial period postinfection, where it is common to have few or no symptoms (45, 46, 141); nonmedical masks have been effective in reducing transmission of respiratory viruses; and places and time periods where mask usage is required or widespread have shown substantially lower community transmission.
The available evidence suggests that near-universal adoption of nonmedical masks when out in public, in combination with complementary public health measures, could successfully reduce Re to below 1, thereby reducing community spread if such measures are sustained. Economic analysis suggests that mask wearing mandates could add 1 trillion dollars to the US GDP (32, 34).
Models suggest that public mask wearing is most effective at reducing spread of the virus when compliance is high (39). We recommend that mask use requirements are implemented by governments, or, when governments do not, by organizations that provide public-facing services. Such mandates must be accompanied by measures to ensure access to masks, possibly including distribution and rationing mechanisms so that they do not become discriminatory. Given the value of the source control principle, especially for presymptomatic people, it is not sufficient for only employees to wear masks; customers must wear masks as well.
Basically, in his Medical Hypotheses article, Vainshelboim ignores all the positive evidence in his eagerness to cherry pick negative evidence and in doing so provides a perfect piece of deceptive antimask propaganda in the form of a seemingly “peer-reviewed study” from Stanford University.
Who is this Baruch Vainshelboim character, anyway?
When I first encountered his Medical Hypotheses article, I had never heard of Baruch Vainshelboim. In the article, his affiliation is listed as “Cardiology Division, Veterans Affairs Palo Alto Health Care System/Stanford University, Palo Alto, CA.” A quick Google search brings up mostly references to his Medical Hypothesis article, such as an article by the American Institute for Economic Research, a right wing propaganda group disguised as a “think tank” behind the Great Barrington Declaration (GBD). The GBD, unsurprisingly, argues that we should just let COVID-19 rip through the population of the young and healthy, who are supposedly at very low risk of dying, while using “focused protection” (which is never really defined) to protect the elderly and the vulnerable. It’s basically a crank document advocating eugenics lite. Meanwhile AIER likes to liken anti-“lockdown” protesters to slavery abolitionists. Also unsurprisingly, AIER loves Vainshelboim’s article, falsely citing it as slam-dunk evidence that masks don’t work.
A quick PubMed search showed 55 publications, which is respectable, and most of his papers (such as this one) do list Stanford and the Palo Alto VA as his affiliations. However, this 2021 paper lists his affiliation as “Rabin Medical Center, Pulmonary Institute (affiliated to Sackler Faculty of Medicine, Tel Aviv University, Ramat Aviv), Petah Tikva, Israel.” His Research Gate Profile also lists this affiliation, as does this source, which lists him as a Professor at the Pulmonary Institute at the Rabin Medical Center. My search also brought up a LinkedIn profile that doesn’t seem to show him with Stanford or the Palo Alto VA, as well as an entry in the Pulmonary Vascular Research Institute, which appears to be some sort of UK charity. Oddly enough, I didn’t see any reference to his being employed at Stanford University or the Palo Alto VA Hospital. I found all this puzzling, given that the email address he listed with his Medical Hypotheses paper was a Gmail account, not a university or VA account.
And then there’s this AP News story:
The study’s author, Baruch Vainshelboim, is listed in the study as being affiliated with the cardiology division at the Veterans Affairs Palo Alto Health Care System/Stanford University.
However, a representative for the VA Palo Alto Health Care System told the AP in an email that Vainshelboim does not work there.
“I can confirm this person is not one of our physicians,” wrote Michael Hill-Jackson, a public affairs specialist with the system. “I do not see him in our system and our Cardiology team has never heard of him.”
Vainshelboim also does not work for Stanford, according to Julie Greicius, senior director of external communications for the university’s medical school.
“Stanford University has never employed Baruch Vainshelboim,” Greicius wrote in an email to the AP. “Several years ago (2015), he was a visiting scholar at Stanford for a year, on matters unrelated to this paper.”
Vainshelboim, who lists himself on LinkedIn as a clinical exercise physiologist and does not list any current employment, did not respond to a request for comment.
Curiouser and curiouser. Before I saw this article, my first guess as to how to explain the discrepancies in his employment history was that Vainshelboim had worked for the Palo Alto VA and had some sort of Stanford clinical faculty or adjunct faculty appointment, but had recently moved to Israel to work at the Rabin Medical Center. That could still be true, I suppose, although even unpaid clinical faculty are usually given a university email account, to allow them access to the online library and various teaching tools, which makes Stanford’s denial of his association with the University very hard to explain. Even more difficult to reconcile with Vainshelboim’s citing affiliations with Stanford and the Palo Alto VA is the statement that no one in the cardiology team at the Palo Alto VA had heard of him makes me wonder what’s really going on. Could it be…grift?
In the end, it doesn’t really matter. What matters is that Vainshelboim took advantage of the tendencies of Medical Hypotheses to publish pseudoscience and nonsense in order to get his “hypothesis” that masks don’t work and are harmful. Medical Hypotheses has always been a dangerous—and not in a good way!—journal in the way it provides ammunition to antiscience cranks. In the age of the COVID-19 pandemic, its embrace of nonsense can have even more deadly consequences.
ADDENDUM: It appears that Baruch Vainshelboim lied about his affiliation with Stanford University and the Palo Alto VA Hospital.