37

This doctor on YouTube claims that the SARS-CoV-2 virus can float in the air for up to 3 hours.

If it lands in the air, the virus can land on a particle and stay floating there for up to 3 hours.

Is that true?

D.W.
  • 107
  • 1
  • 1
  • 5
Franck Dernoncourt
  • 7,224
  • 4
  • 33
  • 76

2 Answers2

37

Author's note, 11 July 2020:

This answer was originally written in late March of 2020, at which point the coronavirus pandemic was, in many countries, far from its peak. In the intervening time, there have been various studies continuing to investigate the possible spread of coronavirus via aerosols and respiratory droplets. I haven't taken a detailed look at recent studies, so I'm not ready to update this, but the takeaway is that our understanding of the virus has gotten better, and it's possible that the study discussed below can now be corroborated or refuted. The underlying point - that the video is based on one set of data points which arguably don't mimic real-world conditions well - is of course still valid.

It's possible, but more research is needed.

Tl;DR: The claim in the video appears to be based on a single recent study of the median half-life of SARS-CoV-2 (more on that below). That study was based on the measurements of virus concentrations over only a short period of time and under conditions unlikely to be found in the real world. While the study's results indicate the virus can stay around for hours in the air, that doesn't necessarily jibe with what we've been seeing around the world.

A later, abbreviated version of the preprint Laurel mentions, published in the New England Journal of Medicine as Aerosol and Surface Stability of SARS-CoV-2 as Compared with SARS-CoV-1, measured the half-lives of SARS-CoV-1 and SARS-CoV-2 in a variety of environments, including aerosols, cardboard, and plastic. This version lists the median viable half-lives of SARS-CoV-2 and SARS-CoV-1 as 1.09 hours and 1.39 hours, respectively, with 97.5% upper confidence limits of 2.64 hours and 2.43 hours. I'm not sure how to account for the discrepancy between the two versions of the paper.

(Note that this is a half-life, so even if we take those median half-lives as the actual half-lives of the viruses, some samples stayed in the air for much longer than that - the half-life is not an upper limit. The paper does not provide an upper limit; the virus was only studied for 3 hours.)

Now, the aerosol measurements were conducted at 65% relative humidity and 21-23°C. This is important to note, because previous research on SARS-CoV-1 indicates that higher humidity and higher temperature can severely decrease the virus half-life (and vice versa for lower humidity and lower temperatures).

Another thing to bear in mind is that coronaviruses are not expected to spread primarily through aerosols, but rather in droplet form. To quote a microbiologist not associated with the study:

The NIAID study "is measuring virus under ideal conditions and with a lot of virus," said microbiologist Benjamin tenOever of the Icahn School of Medicine at Mount Sinai. "So their results are all likely to be overestimates. That said, I think those values should at least be used to let people know that things like subway poles can harbor virus for more time than I would have considered possible," because an aerosol that encounters a solid object can stick to it. "Washing hands is more important than ever."

And, as noted earlier in that article:

"If it could easily exist as an aerosol, we would be seeing much greater levels of transmission," said epidemiologist Michael LeVasseur of Drexel University. "And we would be seeing a different pattern in who’s getting infected. With droplet spread, it’s mostly to close contacts. But if a virus easily exists as an aerosol, you could get it from people you share an elevator with."

According to the Centers for Disease Control and Prevention, that is not happening.

In other words, let's take the study's half-life results - regardless of the version of the paper - with a grain of salt. They were derived under artificial conditions, not real-world environments, and it's not clear that aerosol transmission is significant at all, given actual data. More study is definitely needed.

HDE 226868
  • 11,143
  • 5
  • 61
  • 65
  • It seems obvious to me that aerosol transmission is pretty much the only way the disease moves from person to person. – Daniel R Hicks Mar 26 '20 at 22:21
  • 18
    @DanielRHicks That's not what doctors and epidemiologists are saying, as I've noted in this answer. I'm curious as to why you disagree. – HDE 226868 Mar 26 '20 at 22:24
  • 1
    But the virus needs to get into your lungs or eyes to take hold. Aside from rubbing your eyes you have to inhale it. What you quote is stuff about the virus existing on surfaces. How it then gets into your body is another matter. – Daniel R Hicks Mar 26 '20 at 22:27
  • 6
    @DanielRHicks I would recommend doing some reading; [current information](https://www.cdc.gov/coronavirus/2019-ncov/prepare/transmission.html) suggests that respiratory droplets and contact with surfaces. If you touch a surface with the virus and then touch your face, eyes, mouth, etc., it's quite easy for it to enter your respiratory system - hence why washing hands is exceedingly important. – HDE 226868 Mar 26 '20 at 22:29
  • How is it "not clear that aerosol transmission is significant at all, given actual data"? – Daniel R Hicks Mar 26 '20 at 22:37
  • @DanielRHicks Because, as I noted above, the study does not represent the majority of real-world conditions, and likely yielded overestimates. See the first quote by Benjamin tenOever. – HDE 226868 Mar 26 '20 at 22:43
  • 10
    @DanielRHicks - You appear to be conflating aerosol transmission and droplet transmission. They are different things. – David Hammen Mar 27 '20 at 10:18
  • 1
    @DavidHammen - I think that the Statnews writer does too. – Daniel R Hicks Mar 27 '20 at 11:54
  • @DavidHammen - "An aerosol is defined as a suspension system of solid or liquid particles in a gas. An aerosol includes both the particles and the suspending gas, which is usually air. Frederick G. Donnan presumably first used the term aerosol during World War I to describe an aero-solution, clouds of microscopic particles in air. This term developed analogously to the term hydrosol, a colloid system with water as the dispersed medium." [Wikipedia](https://en.wikipedia.org/wiki/Aerosol#Definitions) – Daniel R Hicks Mar 27 '20 at 12:20
  • @HDE226868 The website you link to essentially supports Daniel R Hicks sentiment that aerosol transmission is the main way of infecting. To quote "The virus is thought to spread mainly from person-to-person... Through respiratory droplets" versus "It may be possible that a person can get COVID-19 by touching a surface". Of course, that's not quite the same as saying the former is the ONLY way (as Daniel R Hicks claimed), however, clearly the CDC will never be able to categorically claim that it is impossible to get infected through touching surfaces with the virus. – user2705196 Mar 27 '20 at 21:41
  • 3
    @user2705196 Respiratory droplets and aerosols are two completely different things - see [one of the papers I linked to](http://eknygos.lsmuni.lt/springer/106/201-212.pdf) that makes the distinction. The two terms are not the same, and the CDC does not mention aerosols. – HDE 226868 Mar 27 '20 at 21:46
  • 2
    The CDC [has further made the distinction in the context of past epidemics](https://www.reliasmedia.com/articles/118763-cdc-adds-respirators-clarifies-confusing-issue-of-droplet-aerosol-and-airborne-transmission). – HDE 226868 Mar 27 '20 at 21:51
12

This is probably referencing Aerosol and surface stability of HCoV-19 (SARS-CoV-2) compared to SARS-CoV-1. (Note that CoV-1 is the 2003 Coronavirus.) They used a “Collison nebulizer” to get the virus into the air and took samples periodically, with the last one being 3 hours later.

According to the study:

We found that viable virus could be detected in aerosols up to 3 hours post aerosolization [...] HCoV-19 and SARS-CoV-1 exhibited similar half-lives in aerosols, with median estimates around 2.7 hours.

Laurel
  • 30,040
  • 9
  • 132
  • 118
  • Thanks for finding the study, interesting! This technically answers the question, but I wonder whether a Collison Nebulizer accurately simulates humans. – Franck Dernoncourt Mar 26 '20 at 13:39
  • 7
    I find the quote quite hard to understand. Could you summarise, and provide a more succinct answer to the OP's question, so that perhaps other readers like myself can learn more from this answer? – theonlygusti Mar 26 '20 at 15:07
  • They performed the aerosol experiment in a Goldberg drum. I don't know much about that, so it's unclear to me to which degree this would apply in a typical room or outdoors. Without more information, I would just take it as a way to compare it with other viruses rather than treating it as representative of how it long it would survive in some real-world conditions. They also mentioned that environmental factors such as temperature and relative humidity could also play a role. – NotThatGuy Mar 26 '20 at 15:28
  • 3
    @theonlygusti - I read the quote as saying that if there's absolutely no airflow and you wait for 3 hours, there will be 6 times less SARS-CoV-2 in the air than when you started. The study simply didn't run for longer than 3 hours, but they tried it thrice. In real life conditions, the aerosol will probably be gone faster. Where to? I don't want to know. – Jirka Hanika Mar 26 '20 at 15:29
  • @FranckDernoncourt - Collison Nebulizer was used with the intention to simulate "aerosol-generating medical procedures"; it probably generates way more aerosol than the average coughing human with a handkerchief. – Jirka Hanika Mar 26 '20 at 15:41
  • 3
    @theonlygusti: there was a recent [Epidemics podcast](https://www.stitcher.com/podcast/dr-celine-gounder/epidemic-with-dr-celine-gounder-and-ronald-klain/e/68181515) that covered the topic a week ago in human intelligible terms. – Denis de Bernardy Mar 26 '20 at 15:47
  • I've found [a later version of the study in NEJM](https://www.nejm.org/doi/10.1056/NEJMc2004973) by the same group that has different results - for instance, lower median half-lives. I'm trying to figure out the discrepancy - do you know what could be the issue? – HDE 226868 Mar 26 '20 at 17:57
  • 5
    Note that if the _half-life_ is 2.7 hours, then the virus will still be present for _much longer_ at decreasing concentration. – chrylis -cautiouslyoptimistic- Mar 26 '20 at 21:27
  • 2
    @HDE226868 - Lack of data. Without much more comprehensive measurements involving different concentrations and longer test durations, as well as assumptions regarding which specific concentrations are infectious to humans, you shouldn't even assume that "half-life" (i.e. exponential decrease) is the right model to use, and that any non-zero concentration is sufficient for infectivity. – Jirka Hanika Mar 27 '20 at 06:34
  • @JirkaHanika My concern is that it appears to be the same study - when I said "later version", I just meant that it appeared to be the same paper on the same experiment, but with some changes, including that data. Much of the actual writing is identical. – HDE 226868 Mar 27 '20 at 13:39
  • 1
    @HDE226868 - The two versions of the study seems to be milking the same (tiny) data set, and their respective numeric findings don't appear to me to be substantially different. Saying that median half life is between 1.1 hours and 1.2 hours, and saying that after 3 hours, the concentration drops 6 times, is about the same thing - especially if you use an a priori model of exponential decrease to justify the term "half-life" and not for example a linear snowfall model. Calculation: 2 to the power of (3 hours divided by 1.15 hours) equals 6. – Jirka Hanika Mar 27 '20 at 17:24