One of the biggest mysteries early on in the COVID-19 pandemic that continues to some extent even today is how SARS-CoV-2, the coronavirus that causes COVID-19, spreads. While it’s always been known to spread primarily through respiratory droplets (and that is still thought to be the primary method of spread), there are still so many questions that remain. For instance, you might remember that early on in the pandemic, fomites (objects on which virus-laden respiratory droplets could land and thereby harbor the virus) were thought to be a major source of infection, as people would touch such objects and then touch their face, and no doubt many of you remember all the breathless stories in March about studies showing that SARS-CoV-2 could survive up to three days on plastic and metal surfaces and up to 24 hours on cardboard. Then, a month ago, the CDC published a news release and updated its website to say that indirect contact from a surface contaminated with coronavirus is a potential way to contract COVID-19 but not the most prominent way that the virus infects people, emphasizing that the “primary and most important mode of transmission for COVID-19 is through close contact from person-to-person” and stating that fomite transmission is not “thought to be the main way the virus spreads.” None of this makes it any less of a good idea or any less imperative that you wash your hands frequently and avoid touching your face. The virus can still spread that way; it’s just not the primary driver of infection.
The real debate though bubbled up in the news over the 4th of July weekend in the form of a story in The New York Times entitled “239 Experts With 1 Big Claim: The Coronavirus Is Airborne“:
The World Health Organization has long held that the coronavirus is spread primarily by large respiratory droplets that, once expelled by infected people in coughs and sneezes, fall quickly to the floor.
But in an open letter to the W.H.O., 239 scientists in 32 countries have outlined the evidence showing that smaller particles can infect people, and are calling for the agency to revise its recommendations. The researchers plan to publish their letter in a scientific journal next week.
Even in its latest update on the coronavirus, released June 29, the W.H.O. said airborne transmission of the virus is possible only after medical procedures that produce aerosols, or droplets smaller than 5 microns. (A micron is equal to one millionth of a meter.)
Proper ventilation and N95 masks are of concern only in those circumstances, according to the W.H.O. Instead, its infection control guidance, before and during this pandemic, has heavily promoted the importance of handwashing as a primary prevention strategy, even though there is limited evidence for transmission of the virus from surfaces. (The Centers for Disease Control and Prevention now says surfaces are likely to play only a minor role.)
I’ll try to remember to update this once the scientists’ actual letter has been published. In the meantime, I thought of two things. First, the reaction to this story reminds me of the reaction to stories that have emerged regarding how presymptomatic and asymptomatic COVID-19 patients can spread coronavirus to others. The second—and more relevant—thing was that this debate reminded me of a very similar debate that I wrote about nearly six years ago. At the time, the nation was gripped in fear that the deadly Ebola virus would make its way here from Africa in order to cause outbreaks, and there was a similar question: Could Ebola be spread by air? Because I’ve written about this before, it allows me to expand a bit on the same basic concepts that I discussed in the context of Ebola and discuss them in the context of what we know about COVID-19. I’ll be very straightforward in admitting that I don’t know yet what to believe on this question, but I’ll look at the basic concepts and the evidence.
Airborne transmission: Droplet size matters
Before I discuss evidence, it’s necessary to know what the experts are talking about when they debate whether COVID-19 spread is airborne. Basically, it all boils down to the size of respiratory droplets. To explain the difference, I’m going to go back to an article by an infectious disease expert named Heather Lander that I cited 6 years ago (remember, in this article she was discussing droplets in the context of claims that Ebola could undergo airborne spread, hence the mention of a bloody wash cloth):
Bodily secretions that make it into the air from various orifices (e.g., nose, mouth) are called droplets and are classified based on size and distance traveled. The smaller the droplet, the longer it stays suspended in the air, the farther it travels and the deeper into the respiratory tract it can go upon inhalation by the person sitting down the aisle from you on the airplane. Teeny-tiny droplets (less than 5 microns) are generally referred to as “aerosols” and can be generated by a cough, a sneeze, exhaling, talking, vomiting, diarrhea, passing gas etc. Aerosols can also be generated mechanically by things like flushing a toilet, mopping, or rinsing out a bloody wash cloth. When aerosols are infectious, they transmit disease when they are inhaled by an organism and its [sic] called “aerosol transmission”. When droplets are larger than 10 microns they are called “large-droplets” and if infectious, they transmit disease by inhalation if the organism being infected is close enough to inhale the particles before they settle out of the air. They can also transmit virus if someone gets showered with droplets from, for example, a sneeze, or touching a droplet that is on the surface of an object (fomite) or someone’s skin and it’s called “large-droplet transmission”.
By way of comparison, SARS-CoV-2 is a bit more than 100 nm in diameter, or 0.1 micron.
It’s more complicated than even that, though:
Not all viruses can form infectious aerosols. It depends on where the virus goes in your body and what happens when it gets there. Aerosol infectivity of a virus is determined by how long the virus remains infectious in the air, how deep into the lungs it can travel, and how many virus particles are actually in each droplet compared to how many are required to actually establish an infection. If a viral infection generates aerosols containing 10 virus particles per droplet, but it takes 1000 virus particles per human cell to establish an infection, then those aerosols are not infectious, even though they contain virus. In addition, while airborne, aerosols begin to lose water content by evaporation and virus particles, especially enveloped particles like Ebola, can be affected by other environmental conditions such as humidity, air currents, and sunlight. These particles are also subject to the laws of physics and mechanical forces. A good example of a virus for which these characteristics have been better defined is influenza and this is an excellent article [PDF] that really explains the different kinds of aerosols and how they are transmitted.
If you keep this in mind, then those studies that you’ve no doubt seen in the press trumpeting how researchers have “isolated coronavirus” from respiratory droplets make more sense. Just detecting virus in droplets doesn’t mean that those droplets are infectious. In particular, remember that most such studies use polymerase chain reaction to detect viral nucleic acid sequences, whose presence alone doesn’t necessarily equate to infectivity. Thus, such droplets might be infectious, but it depends upon how many virus particles are in them and how many virus particles are required to get an infection rolling.
Another important thing to remember is that we are dealing with a continuum. The 5-micron cutoff between an “aerosol” droplet that can hang in the air for a long time and a larger droplet that falls out of the air rapidly, to land on nearby objects, is arbitrary. Humans beings, when coughing, sneezing, or speaking, produce a range of droplet sizes, ranging from aerosols to larger droplets. In order to understand the range of possibilities, this article from the CDC is useful [PDF, linked in the above quote]. Basically, by definition, aerosols are suspensions in the air small enough that they remain airborne for prolonged periods of time because of their low settling density. The article reports that for spherical particles of unit density, settling times for a 3-meter fall are 10 seconds for 100 microns, 4 minutes for 20 microns, 17 minutes for 10 microns, and 62 minutes for 5 microns. Particles with a diameter less than 3 microns in essence do not settle. So for a disease to be truly airborne, it has to produce aerosols that hang in the air for a long time, such that prolonged contact (or even any direct contact at all) with the infected individual isn’t necessary for disease transmission. Also affecting this equation is how far these tiny particles can get into the lung. Particles greater than 6 microns tend to be trapped in the upper respiratory tract, while essentially no deposition of particles into the lower respiratory tract occurs for particles greater than 20 microns. To sum it up, a good rule of thumb is that particles in the micron or submicron range are referred to as, and will behave like aerosols; particles greater than 10 to 20 microns are referred to as large droplets, will settle rapidly, and won’t be deposited in the lower respiratory tract.
It’s even more complicated than that:
Whether propelled by sneezing, coughing, talking, splashing, flushing or some other process, aerosols (an over-arching term) include a range of particle sizes. Those droplets larger than 5-10 millionths of a meter (a micron [µm]; about 1/10 the width of a human hair), fall to the ground within seconds or impact on another surface, without evaporating (see Figure). The smaller droplets that remain suspended in the air evaporate very quickly (< 1/10 sec in dry air), leaving behind particles consisting of proteins, salts and other things left after the water is removed, including suspended viruses and bacteria. These leftovers, which may be more like a gel, depending on the humidity, are called droplet nuclei. They can remain airborne for hours and, if unimpeded, travel wherever the wind blows them. Coughs, sneezes and toilet flushes generate both droplets and droplet nuclei. Droplets smaller than 5-10µm almost always dry fast enough to form droplet nuclei without falling to the ground, and it is usual for scientists to refer to these as being in the airborne size range. It is only the droplet nuclei that are capable of riding the air currents through a hospital, shopping centre or office building.
When infectious disease experts say that a virus is “airborne”, they have a very specific meaning. What they are saying is that the virus is capable of aerosol transmission via inhalation, even when the person inhaling the virus is not in close proximity to the source of the aerosol. In other words, if someone with measles (a very highly infectious virus that can be transmitted through the air), coughs up droplets in a room and then leaves the room and then you enter the room, you can breathe in the measles aerosol and will be very likely to contract measles (that is, if you haven’t been vaccinated against measles or had it before). In other words, droplet transmission is not the same thing as airborne transmission. Airborne transmission can occur in places where the infected patient has been, even if it were hours ago, while droplet transmission requires being close to the infected patient.
That is what this debate is about.
Is COVID-19 transmitted by airborne route?
A few weeks ago, I discussed the evidence addressing whether facemasks work to slow the spread of COVID-19. The specific meta-analysis that I discussed also examined the effect of social distancing on the transmission of coronavirus. The meta-analysis found that a physical distance of more than 1 meter was associated with an 82% decrease in the risk of virus transmission, more precisely that the absolute risk of infection from an exposed individual was 12.8% at 1 m and 2.6% at 2 m, while each additional meter of increased distance resulted in a doubling in the change in relative risk. Taken at face value, this evidence tends to suggest that COVID-19 is transmitted through larger droplets that settle rapidly after short distances. However, it doesn’t exclude the possibility of aerosol transmission.
Most of the evidence for aerosol transmission is anecdotal thus far, albeit nonetheless worrying:
Dr. Morawska and others pointed to several incidents that indicate airborne transmission of the virus, particularly in poorly ventilated and crowded indoor spaces. They said the W.H.O. was making an artificial distinction between tiny aerosols and larger droplets, even though infected people produce both.
“We’ve known since 1946 that coughing and talking generate aerosols,” said Linsey Marr, an expert in airborne transmission of viruses at Virginia Tech.
Scientists have not been able to grow the coronavirus from aerosols in the lab. But that doesn’t mean aerosols are not infective, Dr. Marr said: Most of the samples in those experiments have come from hospital rooms with good air flow that would dilute viral levels.
The incidents cited above included the spread of coronavirus in a choir at choir practice in March:
The full choir consists of 122 singers, but only 61 made it that night, including one who had been fighting cold-like symptoms for a few days.
That person later tested positive for the coronavirus, and within two days of the practice, six more members of the choir had developed a fever. Ultimately, 53 members of the choir became ill with Covid-19, the disease caused by the virus, and two of them died.
The event, which was reported in March by various news organizations, demonstrated how contagious and dangerous the coronavirus is, especially among older populations. The median age for those attending the practice that night was 69.
On the other hand, the story notes that the choir seats were “were packed together, six to 10 inches apart, far closer than the minimum six-foot recommendation by the CDC during the pandemic.” This incident, to me at least, is not compelling anecdotal evidence of airborne spread beyond large respiratory droplets over short distances landing on fomites.
More suggestive of airborne spread is an incident that occurred in a restaurant in Guangzhou, China in January early in the course of the pandemic, in which one diner infected with coronavirus but not yet feeling ill appeared to spread the disease to nine other people in the restaurant. The hypothesis was that the restaurant’s ventilation system blew virus particles around the dining room. However, 73 other diners in the restaurant did not become sick, nor did the eight employees working the floor at the time. Moreover, all of the people who became sick at the restaurant were either at the same table as the infected person or at one of two neighboring tables. Thus, this particular anecdote isn’t very compelling evidence, either, at least not to me.
Still more suggestive anecdotal evidence of airborne spread comes from a German meatpacking plant, where coronavirus infected more than 1,500 workers:
An outbreak of coronavirus that infected more than 1,500 people at a German slaughterhouse may have been spread by “circulating air”.
Experts fear COVID-19 can be spread inside facilities like the Toennies meat plant in Guetersloh because of the systems they use to pump out cool, moist air inside enclosed rooms.
Martin Exner, director of the Institute for Hygiene and Public Health at the University of Bonn, said: “What has not been known so far is that under such conditions circulating air can keep an aerosol moving.
This outbreak is more consistent with airborne spread of coronavirus.
Basically, most of the evidence for airborne spread is either anecdotal or inferred. For example, one recently published review article notes:
While evidence for airborne transmission of COVID-19 is currently incomplete, several hospital-based studies have performed air-sampling for SARS-COV-2, including one published paper (Ong et al. 2020), one early-release paper (Guo et al., 2020) and 5 papers still in pre-print at the time of writing (Chia et al., 2020, Ding et al., 2020, Jiang et al., 2019, Liu et al., 2020, Santarpia et al., 2020). Four of these studies found several positive samples for SARS-CoV-2 genome (RNA) in air using polymerase chain reaction (PCR) testing (Chia et al., 2020, Jiang et al., 2019, Liu et al., 2020, Santarpia et al., 2020), two found very small numbers of positive samples (Ding et al., 2020), and only one (Ong et al., 2020) found no positive air samples. This evidence at least demonstrates a potential risk for airborne transmission of SARS-CoV-2.
A recent mechanistic modelling study showed that short-range airborne transmission dominates exposure during close contact (Chen et al., 2020). Other studies investigating the transport of human-expired microdroplets and airflow patterns between people also provide substantive support for this transmission route (Ai et al., 2019, Li et al., 2007, Liu et al., 2017). Therefore, in light of this body of evidence for these other respiratory viruses; we believe that SARS-CoV-2 should not be treated any differently – with at least the potential for airborne transmission indoors.
The modeling study by Chen et al is interesting in that it found that, even in short range transmission, smaller exhaled droplets are important, with the large droplet route only predominating when the subjects are within 0.2 m while talking or 0.5 m while coughing and that the large droplet route contributes less than 10% of exposure when the droplets are smaller than 50 microns at 0.3 m apart. Of course, this study is a mathematical modeling study and doesn’t have empirical support for its conclusions, but its finding is consistent with both small aerosol droplets being an important mode of transmission and the observation that transmission efficiency drops off rapidly after 1-2 meters.
The authors further note:
Whilst this evidence may be deemed to be incomplete at present, more will arise as the COVID-19 pandemic continues. In contrast, the end-stage pathway to infection of the droplet and contact transmission routes has always been assumed to be via self-inoculation into mucous membranes (of the eyes, nose and mouth). Surprisingly, no direct confirmatory evidence of this phenomenon has been reported, e.g. where there have been: (i) follow-up of fomite or droplet-contaminated fingers of a host, self-inoculated to the mucous membranes to cause infection, through the related disease incubation period, to the development of disease, and (ii) followed by diagnostic sampling, detection, sequencing and phylogenetic analysis of that pathogen genome to then match the sample pathogen sequence back to that in the original fomite or droplet. It is scientifically incongruous that the level of evidence required to demonstrate airborne transmission is so much higher than for these other transmission modes (Morawska et al., 2020).
They do have a point. It’s been accepted that fomites and self-inoculation are a means of spreading coronavirus, but the evidence has been mostly circumstantial and not that rigorous. The one recent study in the Proceedings of the National Academy of Science (PNAS) that found masks are very effective in slowing the spread of COVID-19 has been cited as indicating that the predominant mode of spread of COVID-19 is airborne was riddled with methodological shortcomings, so much so that dozens of scientists petitioned the journal to retract it, saying the study has “egregious errors” and contains numerous “verifiably false” statements, so much so that a scientist named James Heathers wrote an article for Retraction Watch about it (and one other paper) entitled I agree with your conclusions completely, and your paper is still terrible. (That about sums it up.) Part of the problem is a good thing for you to know, namely that PNAS has a special track, called the “contributed” track, by which members of the National Academies of Science (NAS) can solicit their own peer reviews and submit them with the manuscript. That’s why PNAS has long been known as a dumping ground for NAS members to publish their cast-offs or to publish in areas outside their area of expertise. It’s how, for instance, Linus Pauling got some of his terrible papers claiming that vitamin C is an effective treatment for cancer published. More recently, PNAS has switched to a more standard peer-reviewed model of publishing, but the contributed track still remains for NAS members.
The bottom line
The question of whether COVID-19 can spread by aerosol remains without a definitive answer; however, it could be that the debate has been framed in a way that isn’t helpful. Again, respiratory droplets exist on a continuum, from tiny droplets and droplet nuclei that can travel far and hang in the air a long time to larger droplets that float to earth within seconds to minutes. They can also contain varying numbers of infectious virus particles. As one scientist put it:
People generally “think and talk about airborne transmission profoundly stupidly,” said Bill Hanage, an epidemiologist at the Harvard T.H. Chan School of Public Health.
“We have this notion that airborne transmission means droplets hanging in the air capable of infecting you many hours later, drifting down streets, through letter boxes and finding their way into homes everywhere,” Dr. Hanage said.
Experts all agree that the coronavirus does not behave that way. Dr. Marr and others said the coronavirus seemed to be most infectious when people were in prolonged contact at close range, especially indoors, and even more so in superspreader events — exactly what scientists would expect from aerosol transmission.
Dan Diekema, an infectious disease specialist in Iowa, concurs, referring to the controversy as a “tiresome spat“:
As we’ve outlined here and here, a major problem plaguing this discussion is the false dichotomy between “droplet” and “airborne” transmission that we use in healthcare settings (for simplicity of messaging, and because it has served us well for several decades—for reasons I’ll get back to later). This dichotomy divides application of transmission-based precautions between those pathogens spread via respiratory droplets, all of which must absolutely fall to the ground within 6 feet of the source, and those pathogens which become airborne, meaning they travel long distances on air currents, remain in the air for very long periods of time, and most importantly, can cause infection after their airborne sojourns if they find the right mucosal surface.
But we know (and WHO experts know) that there is no such dichotomy—it’s more of a continuum. At the very least there is a middle category, let’s call it Small Particle Aerosol Transmission (or SPAT). Many respiratory viruses (not just SARS-CoV-2) can remain suspended in aerosols and travel distances > 6 feet. As Jorge outlined, it’s probable that transmission events occur when these aerosols are concentrated in closed, poorly ventilated spaces or in very large amounts (e.g. a 2+ hour choir practice, a 3 hour indoor birthday party, a crowded bar). This may explain the superspreading events that drive a lot of SARS-CoV-2 transmission.
The Center for Evidence-based Medicine (CEBM) published a recent review on the evidence base supporting the two-meter rule of social distancing to reduce COVID-19 transmission. In it the authors also discussed the evidence that COVID-19 is transmitted by air and concluded, among other things, that the “longstanding dichotomy of large droplet versus small airborne droplet transmission is outdated and SARS-CoV-2 may be present and stable in a range of droplet sizes, which will travel across a range of distances, including some beyond 2 metres” and that “single thresholds for social distancing, such as the current 2-metre rule, over-simplify what is a complex transmission risk that is multifactorial,” adding that “social distancing is not a magic bullet to eliminate risk,” while further recommending that a “graded approach to physical distancing that reflects the individual setting, the indoor space and air condition, and other protective factors may be the best approach to reduce risk.”
What’s really needed is more science. The amount of science, both excellent, terrible, and every level of quality in between, that’s been produced since the pandemic began is truly amazing, but it’s only been less than seven months since the pandemic started in China. As Kimberly Prather and co-authors argue, what’s needed is this:
Aerosol transmission of viruses must be acknowledged as a key factor leading to the spread of infectious respiratory diseases. Evidence suggests that SARS-CoV-2 is silently spreading in aerosols exhaled by highly contagious infected individuals with no symptoms. Owing to their smaller size, aerosols may lead to higher severity of COVID-19 because virus-containing aerosols penetrate more deeply into the lungs (10). It is essential that control measures be introduced to reduce aerosol transmission. A multidisciplinary approach is needed to address a wide range of factors that lead to the production and airborne transmission of respiratory viruses, including the minimum virus titer required to cause COVID-19; viral load emitted as a function of droplet size before, during, and after infection; viability of the virus indoors and outdoors; mechanisms of transmission; airborne concentrations; and spatial patterns. More studies of the filtering efficiency of different types of masks are also needed. COVID-19 has inspired research that is already leading to a better understanding of the importance of airborne transmission of respiratory disease.
The information gathered on these characteristics of SARS-CoV-2, how it spreads, and how it causes disease will then be of use in the study of other respiratory viruses and future pandemics. In the meantime, I’m coming to the conclusion that we should assume that respiratory aerosol is a major mode of spread of COVID-19 and act accordingly. In many ways, we are already doing that, as social distancing and masks are primary means of preventing the transmission of viruses transmitted this way. However, more can be done, particularly in buildings. Additional strategies to mitigate the spread of COVID-19 in buildings could include refreshing stale indoor air, passing recirculated air through a high-efficiency filter to prevent infecting people in adjacent rooms, and other means of keeping airborne virus confined to limited areas. Then, as the science comes in, recommendations can be fine-tuned based on what we learn. In the meantime, there is no reason to be any more alarmed or even, in most cases, to change what we’re doing to protect ourselves and others.