key: cord-0982899-d643r1v7
authors: Schijven, J. F.; Vermeulen, L. C.; Swart, A.; Meijer, A.; Duizer, E.; de Roda Husman, A. M.
title: Exposure assessment for airborne transmission of SARS-CoV-2 via breathing, speaking, coughing and sneezing
date: 2020-07-05
journal: nan
DOI: 10.1101/2020.07.02.20144832
sha: 4a38105234b08d510e45dbff22db12954932ddef
doc_id: 982899
cord_uid: d643r1v7

Background Evidence for indoor airborne transmission of SARS-CoV-2 is accumulating. If SARS-CoV-2 also spreads via aerosols, this has implications for measures taken to limit transmission. Objectives The aim of this study is to assess exposure to airborne SARS-CoV-2 particles from breathing, speaking, coughing and sneezing in an indoor environment. Methods An exposure assessment model was developed to estimate numbers of SARS-CoV-2 particles in aerosol droplets, expelled during breathing, speaking, coughing and sneezing by an infected person in an unventilated indoor environment, and subsequent inhalation by one or more persons. Scenarios encompass a range of virus concentrations, room sizes and exposure times. Results The calculated total volume of expelled aerosol droplets was highest for a sneeze, followed by a cough and speaking for 20 minutes, and lastly breathing for 20 minutes. A few to as much as tens of millions of virus particles were expelled. Exposure probability strongly depends on the viral concentration in mucus, as well as on the scenario. Exposure probabilities were generally below 1% at a virus concentration in mucus below 10^5 per mL for all scenarios, increasing steeply at different higher concentrations. According to nose / throat swab data collected from patients, 75%, 50% and 5% of infected individuals carry an estimated number of SARS-CoV-2 per mL mucus of at least 10^5, 10^6 and 10^8, respectively. Discussion Exposure to SARS-CoV-2 via aerosols generated during breathing, speaking, coughing and sneezing in an unventilated indoor environment is possible. This study forms a basis to estimate probabilities of exposure to SARS-Cov-2 by airborne transmission in indoor spaces. As long as it is uncertain what fraction of the airborne virus particles is infectious and as long as a dose response relation is lacking, it is recommended to be precautious.

The fate of droplets in the air is mostly determined by their size; larger droplets deposit quickly while 57 smaller droplets can stay airborne for longer periods, in so-called aerosols (a suspension of droplets 58 in air). Aerosols can arise from 'violent expiratory events' such as coughing and sneezing (Bourouiba <5 µm can penetrate more deeply into the lungs, while larger particles most likely impact the upper 65 airways (Gralton et al. 2011; Tellier et al. 2019 ). While it is true that the large majority of the volume 66 of fluids that is expelled during, for example, coughing and sneezing, is in droplets that deposit 67 quickly, this does not imply that airborne transmission is highly unlikely (Nicas et al. 2005) . 68 Furthermore, research suggests that the cut-off size of droplets (aerodynamic diameter) which 69 deposit quickly is higher than 5 µm, and not static but dependent on a number of factors, such as 70 relative humidity (Liu et al. 2017 ). In the absence of turbulence, droplets with an initial diameter 71 larger than 80 µm will be deposited on the floor from an initial height of 2 m at a distance away from 72 the mouth of around 1 m (Liu et al. 2017 ). The droplet with an initial diameter of 60 µm can reach 73 about 4 m, with a size of 0.32 times its initial diameter at a relative humidity (RH) of 0%, whereas it 74

can travel a distance of 1.85 m at a RH of 90% due to its larger droplet size of 0.43 its initial diameter 75 (Liu et al. 2017 ). In the case of turbulence, even initially larger particles could likely travel even 76

further. Therefore, airborne and droplet transmission occur on a continuum, and airborne 77 transmission can potentially occur in the size fraction of all particles less than about 60 µm (Gralton 78 et Similarly, van Doremalen et al. (2020) also found that SARS-CoV-2 remained viable for hours in 92 experimentally generated aerosols (reduction in infectious virus particles from 3100 to 500 per litre 93 air in 3 hours). 94

There is much discussion about the potential for airborne transmission of SARS-CoV-2. The World 95

Health Organization so far maintains that COVID-19 is not airborne (https://who.africa-96

newsroom.com/press/coronavirus-fact-check-covid19-is-not-airborne). Eissenberg et al. (2020) 97 discuss the evidence cited by WHO and conclude that it remains prudent to consider airborne 98 transmission of COVID-19 as an explanation for the rapid spread of the virus. If SARS-CoV-2 also 99 spreads via airborne transmission, this has implications for the measures that are being taken to 100 limit transmission, such as advice to keep a certain distance from other people. If SARS-CoV-2 is also 101 airborne, transmission would be plausible beyond the often advised 1.5 meters. 102

The aim of this study is to assess exposure to airborne SARS-CoV-2 particles from breathing, 103

speaking, coughing and sneezing in an indoor environment. Figure 1 shows a schematic overview of 104 the processes modelled in this study. The exposure assessment entailed estimating the numbers of 105 SARS-CoV-2 particles in aerosol droplets, expelled during breathing, speaking, coughing and sneezing 106 by an infected person in an unventilated indoor environment, and subsequent inhalation by one or 107 more persons in that environment. Exposure assessment is part of a Quantitative Microbial Risk 108

Assessment (QMRA), in which exposure assessment is followed by a risk characterisation. In risk 109 characterisation, estimates of risk of infection and/or illness are made using dose response relations 110 (Haas et al. 1999 (1) 167

where c is the virus concentration [numbers per mL mucus] with mean µ, standard deviation σ,

coefficients a0 and a1 and time t [days] from onset symptoms. Note that, in the Zou-data, a 170 distinction between the data from nasal and throat swabs was made, but this was not possible in the 171 RIVM-data where most samples were combined nose-throat swabs. SARS-CoV-2 concentrations are 172 determined as Ct values, which are inversely related to viral RNA copy number. Viral RNA copy 173 number in mucus was estimated from Ct values, as detailed in Supplementary material S1. 174

The total initial volume of aerosol droplets per cough, sneeze, 20 minutes speaking and 20 minutes 176 breathing was calculated from the number of aerosol droplets and their size distribution. Following 177 Liu et al. (2017) , in collecting literature data of the size distribution of expelled droplets by breathing, 178

speaking, coughing and sneezing, expelled droplets smaller than 60 µm were considered, when 179 measured directly in front of the mouth, assuming little evaporation had happened. Liu et al. (2017) 180 reported that the droplet nuclei size at a relative humidity of 90% (25°C) could be 30% larger than 181 the size of the same droplet at a relative humidity of less than 67.3% (25°C). In the case of a distance 182 of about 0.5 m or more, the size distribution of droplets of 20 µm or less was considered, and for 183 these droplets, their initial size distribution (at the point of leaving the mouth) was estimated by 184 multiplying their diameter by a factor of three, to correct for evaporation (Liu et al. 2017) . 185

For the total aerosol droplet volume by breathing, the data reported by the different size classes may exist that was not apparent from this chart (e.g. a subject that expels 193 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. . https://doi.org/10.1101/2020.07.02.20144832 doi: medRxiv preprint above average droplets in one size class is perhaps likely to also expel above average in other size 194 classes). The total volume of aerosol droplets per minute of exhaled air, vbr was calculated as 195 follows: 196 = 10 ( , ) 10 −12

where di is the diameter [µm] of aerosol droplets in the i-th class of six aerosol droplet diameters, of 198

which the logarithm of their concentration in air is normally distributed with mean µi and standard 199 deviation σi. The volumes of each aerosol droplet size were summed, then converted from µm 3 to 200 millilitres by a scaling factor of 10 -12 and multiplied by the tidal breathing rate that is normally 201 distributed on log-scale with mean µbr and standard deviation σbr (Table 2) .

The numbers of expelled aerosol droplets nsp,co,sn during speaking, coughing or sneezing, respectively, 203

were assumed to be lognormally distributed: 204

The ≈ sign denotes rounding to integer values. Values of parameters µsp and σsp are given in Table 2   206 and represent the range of numbers of aerosol droplets as reported by Duguid (1945) and is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. .

It was assumed that the expelled aerosol droplets were instantaneously evenly dispersed in the 233 room. The dose D was computed as follows: 234

where vr is the volume of the room [litres], and vinh is the inhaled volume of an exposed person 236 during 20 minutes using the same tidal breath rate as in equation (3), s is a sensitivity factor. 237

In equation (6), s*c can be interpreted as a change in c by factor s. Similarly, s*vbr,sp,co,sn, s/vr and 238 s*vinh can be interpreted as a change in vbr,sp,co,sn, 1/vr and vinh. In other words, any change in these 239 variable or combination of these variables by factor s has the same effect on the dose and 240

probability of exposure. Factor s can also be interpreted as the number of exposed persons, or the 241 number of persons expelling virus. Also note that a factor s change in vinh can be due to a change in 242 exposure time. 243

In the scenario with one exposed person in a bus for 20 minutes, s= 1. This is the reference scenario. 244

In the scenario with 30 exposed person in a bus for 20 minutes, s=30. In the larger room with 10 245

persons for 1 hour, s=1/7*10*3=4, and in that room for 4 hours, s=16. For the sensitivity analyses 246 two extra scenarios were added with factor s=0.1 and s=100. 247

Finally, the probability of exposure Pexp was computed as one minus the Poisson probability of 248 exposure to zero particles: 249 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. . range from 0.004 to 30 000 picolitres per 20 minutes), the speaking and coughing scenarios vary 273 about three orders of magnitude and the sneezing scenarios vary about two orders of magnitude. 274

The total aerosol volume from speaking loudly for 20 minutes is in between that of the low and high 275 scenarios from one cough. 276 277 3.3 Viral concentration data from swabs 278

Observed SARS-CoV-2 concentrations in mucus spanned a wide range, from 10 2 to 10 11 copies / mL 279 (corresponding to a range of Ct values from 40 to 10.5). Observed concentrations in mucus were on 280 average 2 orders of magnitude higher for the RIVM data (~10 6 copies / mL) than in the Zou data 281 (~10 4 copies / mL, Figure 3 bottom panels). SARS-CoV-2 virus concentrations from the nasal swabs of 282 Zou et al. (2020) were, on average, higher than those from throat wabs, and both decreased over 283 time (days since symptom onset) (Supplementary material S1). This decrease in concentration from 284 the onset of symptoms was also observed in the own data. The bottom panels of Figure 3 show probabilities, and these probabilities vary one to two orders of magnitude, like the number of 310 expelled virus particles. Generally, low probability of exposure is observed at a virus concentration 311 below 10 5 per mL for all scenarios, except the 100x reference scenario. Furthermore, of course 312 directly related to the volume of expelled aerosol droplets (Figure 2) , it is observed that the 313 probability starts to increase steeply at different concentrations for each of the scenarios. Going 314 from the high sneezing scenario to the low sneezing scenario and the high coughing scenario, then 315

to both speaking scenarios, followed by the low coughing scenario, and finally, the breathing 316 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. . https://doi.org/10.1101/2020.07.02.20144832 doi: medRxiv preprint scenario, for each these steps, about a ten times higher virus concentration is required for the same 317 exposure probability. Going to a scenario with more exposed persons, longer time for exposure, or 318 smaller room, the probability of exposure curves all shift to the left. 319

For example, at a virus concentration of 10 6 per mL mucus and in the high sneeze scenario, in the 320 bus with one exposed person, figure 3 indicates a probability of exposure of about 4%. In the same 321

scenario, but with 30 exposed persons, the probability of exposing at least one person amounts to 322 60%. In the larger room, with 10 exposed persons, the exposure probability is 14% and 41% for the 323 1-hour and 4-hours exposure times, respectively. In a full bus and nobody speaking, coughing or 324 sneezing, the probability of exposure is about 6% when an infected person expelled virus at a 325 concentration of 10 8 per mL. And if in that bus an infected person would sneeze heavily (the high-326 sneeze scenario), at a virus concentration of 10 8 per mL, probability of exposure of at least one 327 persons would be equal to one. At a concentration of 10 6 , this probability is 60%. The probability of exposure of at least one person to SARS-CoV-2 particles contained in aerosol 336 droplets that were expelled by an infected person was estimated in various scenarios wherein the 337 infected person was expelling virus by breathing, speaking, coughing or sneezing. An important 338 assumption was that the initial aerosol droplets have the same virus concentration as measured in 339 mucus from nasal and throat swab samples, and that the numbers of virus particles in mucus were 340 evenly distributed. Moreover, it was assumed that the expelled aerosol droplets were 341

instantaneously and evenly distributed in the air of the room. For the sake of simplicity for this first 342 study, it was assumed that an infected person leaves the room before one or more susceptible 343 persons enter. In further work, this could evolve to a dynamic model with rates at which virus is 344 expelled and inhaled, as was done by Buonanno et al. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. . https://doi.org/10.1101/2020.07.02.20144832 doi: medRxiv preprint concentrations, ranging from 10 -3 to 10 -1 RNA copies / L for patient areas, 10 -3 to 4 x 10 -2 for staff 361 areas, and 10 -3 to 10 -2 for public areas. Highest SARS-CoV-2 concentrations were detected in droplet 362 size ranges of 0.25 -1 μm and >2.5 μm. Guo The assumption that expelled aerosol droplets are instantaneously and evenly distributed in the air 369 of the room implies that there is an immediate dilution of the expelled virus concentration, which 370 lowers its concentration in the air, but also spreads the virus. Obviously, dilution will not really occur 371

instantaneously, it highly depends on the aerodynamics in the room. An exposed person directly in 372 front of the infected person, or in a flow path of the contaminated air, may inhale a much larger 373 dose than average. Clearly, air ventilation is very important. It may be surmised that in outdoor 374 spaces exposure probability will be much less, due to much more dispersion and dilution. 375

In the exposure assessment, virus inactivation was not included. According to van were also found to be highly expressed in nasal epithelium meaning that infection can likely 392 establish there as well (Sungnak et al. 2020 ). SARS-CoV-2 prevalence in the population of course 393

influences the probability that an infected person is present in a bus or room, this was not 394

accounted for in the model. Similarly, immunity in the population influences the probability that 395 susceptible persons are present. 396

The current study was entirely focused on estimating probabilities of exposure to SARS-CoV-2 in 397 aerosolized droplets that are small enough to be distributed in the air farther than 1.5 meter. From 398 this study, the relative importance of aerosol and droplet transmission cannot be determined. To 399 put this in perspective, Duguid et al (1946) , also captured droplets with an initial diameter of more 400 than 60 µm when speaking, coughing or sneezing, of which the total volume can be 3-5 orders of 401 magnitude higher than the total volume of the smaller droplets <60 µm. However, for droplet 402 transmission, different factors govern the probability of exposure, such as the probability of 403 expelling droplets directly onto the mucosa of another person, hand hygiene, hand-mouth contact 404 virus transfer rates, etcetera. It should be noted that, although transmission at short distances is 405 commonly thought to be droplet transmission, Chen et al. (2020) found that the short-range 406 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. . https://doi.org/10.1101/2020.07.02.20144832 doi: medRxiv preprint airborne route actually dominates at most distances during both talking and coughing. Chen et al. 407 studied airborne transmission in general, not specifically SARS-CoV-2 or another virus. 408

The highest numbers of viruses were expelled in aerosol droplets from a sneeze, followed by the 409 high coughing scenario. Persons that are sneezing and/or coughing are advised by governments to 410 stay at home. However, it is also reported that symptoms can start acutely and that the day before, 411

and first days with, symptoms, the highest virus loads are found in naso-pharyngeal swabs (Kimball 412 et on the quantity of viral RNA. Since in an efficient PCR, a Ct value that is 9-10 cycli lower represents a 440 virus RNA load that is 500-1000 higher, this information can be used to indicate a warning for those 441 that may possibly shed much more than the average person. In the own data set the average 442 detection is at Ct 26.1 cycli and about 2.5% shedders at Ct≤17 cycli were found, that can be indicated 443 as probable suppershedders. 444 A superspreading event probably has multiple components. A person producing high levels of virus 445

in the (upper) respiratory system can be thought of as a superreplicator. If a superreplicator is 446 average or above average in droplet and aerosol production he or she may be a supershedder: not 447

only producing high load of virus but also excreting them in transmittable form. If a supershedder 448

does not follow the mitigation protocols (stay at home when sick, proper coughing and sneezing 449 hygiene etc.) a supershedder can become a superspreader: someone causing many more infections 450 than anticpated based on the average R0. The exposure assessment of this study demonstrates that 451 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. . https://doi.org/10.1101/2020.07.02.20144832 doi: medRxiv preprint viral RNA copy concentrations in mucus above 10 8 per mL may easily give rise to very high exposure 452 probabilities, even during breathing and speaking. SARS-CoV-2 superspreaders might be pre-or 453 asymptomatic. Kimball et al. (2020) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 5, 2020. . . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 5, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 5, 2020. . https://doi.org/10.1101/2020.07.02.20144832 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 5, 2020. . https://doi.org/10.1101/2020.07.02.20144832 doi: medRxiv preprint CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 5, 2020. . https://doi.org/10.1101/2020.07.02.20144832 doi: medRxiv preprint

The size and the duration of air-carriage of respiratory droplets and droplet-nuclei

Treat covid-19 as though it is airborne: It may be

Estimating the overdispersion in covid-19 transmission 537 using outbreak sizes outside china

Origin of exhaled breath particles from 539 healthy and human rhinovirus-infected subjects

A field indoor air 541 measurement of sars-cov-2 in the patient rooms of the largest hospital in iran

Comparative 544 dynamic aerosol efficiencies of three emergent coronaviruses and the unusual persistence of sars-545 cov-2 in aerosol suspensions

Epidemiology: Dimensions of superspreading

Assessment of 548 experimental and natural viral aerosols

The role of particle size in aerosolised 550 pathogen transmission: A review

Aerosol and surface distribution of severe 552 acute respiratory syndrome coronavirus 2 in hospital wards

Quantitative microbial risk assessment

High sars-cov-2 attack rate 556 following exposure at a choir practice -skagit county, washington

Asymptomatic and presymptomatic sars-cov-2 559 infections in residents of a long-term care skilled nursing facility -king county, washington

Aerosol or droplet: Critical definitions in the covid-19 era

Evidence of 564 respiratory syncytial virus spread by aerosol. Time to revisit infection control strategies? American 565

Respiratory virus 567 shedding in exhaled breath and efficacy of face masks

Aerosol transmission of sars-cov-2 -569 evidence for probable aerosol transmission of sars-cov-2 in a poorly ventilated restaurant

Quantity and size 571 distribution of cough-generated aerosol particles produced by influenza patients during and after 572 illness

Evaporation and dispersion of respiratory droplets from coughing

Aerodynamic analysis of sars-cov-2 in two 576 wuhan hospitals

Superspreading and the effect of individual 578 variation on disease emergence

Singing and the dissemination of tuberculosis

Mycobacterium 582 tuberculosis miniepidemic in a church gospel choir

Influenza virus aerosols in 584 human exhaled breath: Particle size, culturability, and effect of surgical masks

Airborne transmission of sars-cov-2: The world should face the reality

Toward understanding the risk of secondary airborne 589 infection: Emission of respirable pathogens

Air, surface environmental, and 591 personal protective equipment contamination by severe acute respiratory syndrome coronavirus 2 592 (sars-cov-2) from a symptomatic patient

Viral load of sars-cov-2 in clinical samples. The 594

R: A language and environment for statistical computing

Sars-cov-2 is 598 transmitted via contact and via the air between ferrets

Transmission 600 potential of sars-cov-2 in viral shedding observed at the university of nebraska medical center

Evaluation of exposure 602 scenarios on intentional microbiological contamination in a drinking water distribution network

Airborne 605 transmission route of covid-19: Why 2 meters/6 feet of inter-personal distance could not be enough

Airborne transmission of covid-19: 608 Epidemiologic evidence from two outbreak investigations

Small droplet aerosols in poorly ventilated 610 spaces and sars-cov-2 transmission. The Lancet Respiratory Medicine

The airborne lifetime of small speech droplets and 612 their potential importance in sars-cov-2 transmission

Rstan: The r interface to stan

Q fever pneumonia: Virulence 616 of coxiella burnetii pathovars in a murine model of aerosol infection

Sars-cov-2 entry 619 factors are highly expressed in nasal epithelial cells together with innate immune genes

Recognition of aerosol transmission of infectious agents: A 621 commentary

A generalized dose-response relationship for adenovirus 623 infection and illness by exposure pathway

Aerosol 625 and surface stability of sars-cov-2 as compared with sars-cov-1. The New England Journal of 626 Medicine

628 Shedding of infectious virus in hospitalized patients with coronavirus disease-2019 (covid-19): 629 Duration and key determinants. medRxiv:2020

Covid-19 may transmit through aerosol

Clinical features of 69 cases with coronavirus disease 633 2019 in wuhan, china

Development of a dose-response 635 model for sars coronavirus

Inactivation of influenza a viruses in the environment and modes of 637 transmission: A critical review

Guidelines infection prevention and control of epidemic-and 641 pandemic-prone acute respiratory infections in health care

World Health Organization. 2020. Modes of transmission of virus causing covid-19: Implications for 643 ipc precaution recommendations, verison 29

Transmission routes of covid-19 virus in 645 the diamond princess cruise ship

Infectious virus in 647 exhaled breath of symptomatic seasonal influenza cases from a college community

On airborne transmission and control of sars-cov-2. Science of 650 The Total Environment

Evidence of airborne transmission of 652 the severe acute respiratory syndrome virus

Airborne spread and infection of a novel swine-654 origin influenza a (h1n1) virus

Sars-cov-2 viral load in upper 656 respiratory specimens of infected patients