Rapid review on SARS-CoV-2 aerosol transmission: update 2

Saidel-Odes (2021)^{Footnote 53}

Cluster Investigation

Israel

Sep – Oct 2020

new

Investigation of a COVID-19 outbreak linked to an index patient who was immunocompromised with multiple myeloma and underwent stem cell transplant. Patient was treated in an AIIR following COVID-19 diagnosis, and healthcare staff donned appropriate PPE (i.e., N95, face shield, gown and gloves) prior to entering the patient’s isolation room. All staff in the unit wore gowns, gloves and masks.

Seven healthcare staff in the transplant unit housing the index patient were diagnosed with the infection; attack rate of 19% (n=7/37). The infected staff included 2 doctors, 4 nurses, and 1 housekeeping worker. Two of the positive staff did not report any direct contact with the index patient. The outbreak investigation and univariate analysis found infected healthcare workers were more likely to have reported spending time in the unit’s corridor or nurses’ station ([RR], R = 7.2; 95% CI, 1.22–42.49; P = .018), but not in the index patient’s room. The investigators suggest aerosol transmission to be the only plausible explanation for this outbreak.

Eichler (2021)^{Footnote 14}

Cluster Investigation

New Zealand

Sep 2020

new

Investigation of a COVID-19 outbreak that occurred during repatriation, at a mandatory isolation and quarantine facility (hotel) and among household contacts in the community. Cases who were returning other countries were required to complete a mandatory. All persons went to a Mandatory quarantine and isolation at a facility for 14 days after landing at their destination.

Based on the application of video surveillance data and viral genomic analysis of cluster cases, the investigators suggest a multiple transmission chains. The suspected sites of transmission were international and domestic flights, the quarantine facility, and households. While aerosol transmission may have been the mode in each transmission event, but an in-depth investigation concluded aerosol transmission was the primary mode of infection transmission at the hotel quarantine facility.

The case exposures/transmissions in the outbreak are assumed to be as follows:

2 index cases infected around the time of repatriation travel from India to Christchurch, NZ. 1 exposed on the flight to Christchurch, 2 exposed in hallway at the quarantine facility (by aerosols see below), 1 exposed on the flight from Christchurch to Auckland (transmission from the 2 aerosol exposed cases who tested negative on day 12 of quarantine), with additional cases among household contacts of the last three cases.

Based on computer footage the investigators concluded aerosol transmission to have taken place at the doorway of two different hotel rooms that housed individuals in quarantine by suspended aerosols. Transmission was assumed to have occurred during a 50-second window between closing the door to the room of one case and opening the door to the room of other cases.

Note: A review of the hotel’s ventilation system found the rooms in question had a net positive pressure compared with the corridor so air flow into the corridor when the door was open was likely

de Man (2020)^{Footnote 15}

LTE

Cluster Investigation

Netherlands

Jun – Jul 2020

new

Investigation of a COVID-19 outbreak in a single ward of a 7 ward Dutch nursing home.

A total of 17/21 residents in the ward and 13/34 healthcare workers from the ward, and an additional 4 laboratory workers were confirmed cases linked to the cluster; attack rate among residents was 81% and among healthcare workers was 50%. All 106 healthcare workers and 95 residents of other wards remained negative for COVID-19. Healthcare workers wore facemasks during patient care and worked in designated wards to limit contacts. The outbreak was limited to a single ward (out of 7 wards in the facility) with a ventilation system that recirculated unfiltered air. The ventilation system was new and monitored CO2 concentrations to determine when to refresh the air with outside air. Viral RNA was detected in the ventilation system (air conditioners and ventilation cabinet dust filters).

The investigators state transmission was likely the result of aerosol transmission in an inadequately ventilated setting/ward given the simultaneous detection of a large number of cases limited to one ward, and occurred during a time of low community transmission.

Luo (2020)^{Footnote 7}

Cluster Investigation

China

Jan 2020

new

Investigation of a COVID-19 outbreak linked to multiple bus trips by the index case, in Hunan China.

A total of 12 cases were identified following the investigation of 243 individuals who were epidemiologically linked to multiple bus trips taken by the index case, attack rate of 7.0%.

The seated distance of the infected individuals during bus rides ranged from 1-4.5 meters from the index case. The 2.5 hour coach ride resulted in 8 cases and not all cases were clustered close to each other during the trip. Infection was also identified in an individual who sat (approximately) on the same seat as the index case, after the index case had existed the bus. 2 cases were identified from the 1 hour minibus trip. 2 cases were tertiary cases to one of those infected on the bus.

None of the identified cases reported wearing face masks during bus rides.

The investigators suggest aerosols traveling beyond 2 meters and air flow to have influenced infection transmission in the crowded and closed environment of these transportation buses.

Shen (2020)^{Footnote 54}

Cluster Investigation

China

Jan 2020

A COVID-19 outbreak among 128 people driven to a worship event in Eastern China on two separate buses. Round trip was 100 minutes on the bus.

Attack rates were measured for Bus 1 vs. Bus 2 that had the index case. Air conditioning systems of both buses were on recirculation mode. Spatial analysis of passenger seating was estimated.

None of the passengers on Bus 1 were infected, 24 of the 68 passengers on Bus 2 developed COVID-19.

Passengers riding Bus 2 with the index case had an attack rate of 34.3% (95% CI, 24.1%-46.3%), compared to passengers on bus 1.

Although sitting near bus windows and doors appeared to have had a protective effect on infection transmission, the authors conclude, the lack of a significant increase in infection risk between individuals sitting in high risk zones (i.e. closer to the index case) and low risk zones, and elevated attack rates among bus passengers riding with the index case, to be partially explained by aerosol transmission of infection.

Guenther (2020)^{Footnote 55}

Preprint

Cluster Investigation

Germany

Spring 2020 (est)

Investigation of a super-spreader event among meat processing plant workers that included: possible routes of transmission, spatial relationship between workers, climate/ventilation conditions, sharing of living quarters and transportation, and genetic typing of oropharyngeal swab samples.

The analysis of index cases (flatmates) and 18 co-worker cases suggest working the early morning shift (140 early shift workers) to be the common source of infection.

Statistically significant infection rates were observed for workers working within an 8-meter radius of the suspect index case.

Authors conclude indoor confined settings, demanding physical work, and the facility’s environmental conditions (i.e. air being constantly re-circulated and cooled to 10°C, with low air exchange rates) all created conditions for aerosol transmission.

Note: quantitative risk estimates were not provided.

Kwon (2020)^{Footnote 8}

Cluster Investigation

Korea

Jun 2020

new

Investigation of a cluster among 3 patrons at a restaurant. The investigation considered epidemiological data, television images, air flow patterns and cell phone location information.

The investigators concluded infection transmission occurred from an infected customer to two individuals at the restaurant venue, attack rate of 15.4% (2/13). Source and time of exposure was assumed to be one day prior to symptom onset in the index case. The assumed mode of transmission was aerosols that travelled greater than 6.5 meters via air conditioner air flow current inside the restaurant venue. The exposure time between the infector and the infected is reported to be approximately 5 minutes.

Mask use is reported to have been improper among staff and patrons.

Lu (2020)^{Footnote 5}

Cluster Investigation

China

Jan-Feb 2020

Investigation of a COVID-19 cluster among restaurant patrons. The investigation included a spatial analysis of restaurant table arrangement and where cases were seated.

An outbreak among 91 individuals at a restaurant, 83 had dined at 15 tables, and the remaining 8 individuals were staff. A single asymptomatic case led to 9 COVID-19 infections among diners from three families. None of the families had met previously and did not have any close contact during lunch. No additional cases were identified during the 14 days quarantine of the remaining diners.

Spatial analysis of case tables during lunch (i.e., exposure event reveal) found the affected tables had been arranged in line with airflow from an air conditioning unit. Authors suggest infection transmission could not be explained by droplets alone, and aerosols travelling with air flow may have contributed to infection transmission.

Li (2020)^{Footnote 57}

In silico study

China

Feb 2020

Note: Same outbreak described by Lu (2020).

An investigation and analysis of a COVID-19 cluster among 3 families who ate at the same restaurant. The analysis included: epidemiological data, spatial analysis of restaurant table arrangement, video surveillance data, and computer fluid dynamic and tracer gas simulations of event’s fine droplet spread.

10 people from three different families seated at different tables were found to have been infected with SARS-CoV-2 following a Chinese New Year’s Eve (January 24, 2020) lunch. None of the waiters or patrons at the remaining tables became infected. Ventilation rate was estimated to be 0.9L/s (0.75-1.04 L/s) per person.

No close contact or fomite contact was observed among cases, aside from back-to-back sitting by some patrons.

Using computer simulations the authors demonstrate infection distribution to be consistent with the spread pattern of exhaled virus aerosols. These results highlight transmission occurring in a crowded and poorly ventilated indoor situation.

Charlotte (2020)^{Footnote 16}

Cluster Investigation

France

Mar 2020

new

The investigation of a COVID-19 outbreak linked to a choir practice, in Whir au Val, France.

Twenty-seven participants (25 singers, 1 conductor and 1 accompanist) attended the indoor choir practice that took place in a non ventilated space of 45 m². No attendees reported being symptomatic in the 14 days prior to the practice date.

19/27 attendees developed COVID-19 infections at 1 to 12 days post practice date (median of 5.1 days). The secondary attack rate was 70% among all attendees.

The investigators assume infection transmission occurred by aerosols from presymptomatic or asymptomatic individual(s) who attended the choir practice as the index case was not definitively identified.

Hamner (2020)^{Footnote 58}

Cluster Investigation

US

Mar 2020

Note: Same outbreak described by Miller (2020).

The investigation of a COVID-19 outbreak linked to a choir practice, in Skagit County, Washington. The practice lasted for 2.5h. During practice people were singing and seated 6-10 inches apart, socializing with communal snacks, and stacking chairs. None of the attendees reported physical contact.

Among the 61 choir members attending the practice, at least one singer was known to be a symptomatic COVID-19 case. The epidemiological investigation reported 53 cases (33 confirmed, 20 probable cases). Secondary attack rates were 53.3% among confirmed cases and 86.7% among all cases.

The odds of infection were 125.7 (95% CI: 31.7-498.9) times greater among members who attended the March 10 practice (assumed exposure event).

The investigators introduce the potential for aerosol emission and COVID-19 transmission during singing in the COVID-19 literature.

Miller (2020)^{Footnote 59}

In silico study

US

Mar 2020

Note: Same outbreak described by Hamner (2020).

Monte Carlo simulations and mathematical modeling were used to estimate aerosol emission rates in the outbreak linked to a choir practice, in Skagit County. The applied model assumes infection transmission during the outbreak was dominated by inhalation of respiratory aerosols in a well mixed indoor environment (i.e. the aerosols were evenly distributed in the air).

The viral load emitted was expressed as quanta emission rate (quanta per hour⁻¹) where a quantum was defined as the dose of aerosol droplet nuclei required to cause infection in 63% of susceptible persons.

In silico analysis supported aerosol transmission from respiratory aerosols based on assumption that high emission rates occurred given the high attack rate (53-87%), which was higher than would be expected if the transmission was due to fomites or large respiratory droplets.

The model estimates the mean respiratory aerosol emission rate for a single infected case at the exposure event to be 970 [IQR 680-1190] quanta per hour⁻¹.

Buonanno (2020)^{Footnote 60}

In silico study

China and US (sites of applied outbreaks)

Feb-Mar 2020

Note: A different analysis of restaurant and choir practice outbreaks described above

This is an emission and exposure model that used a step-wise approach to quantify individual infection risk among susceptible subjects exposed to an asymptomatic/ mildly symptomatic case in choir practice and dine-in restaurant.

Also used Monte Carlo method; individual infection risks were calculated as a function of quanta emission characteristics.

The model illustrated individual infection risk increased based on ventilation rates, activities and amount of virus exhaled. For instance, sedentary activities for 1 hour may have an infection risk of 2.1%, which can increase to 27% with higher emission rates.

Based on risk assessment approach and available data, quanta emission rates were estimated to be 61 quanta/hour for the restaurant and 341 quanta/hour for the Skagit Valley choir practice. In both of the examples, varying the ventilation would not have achieved an individual risk <0.1.

The authors concluded aerosol transmission represents the main route of transmission for both outbreaks.

Kriegel (2020)^{Footnote 61}

In silico study

Germany, China, US, (sites of applied outbreaks)

Feb-Mar 2020

Note: Included the following clusters: Meat Processing plant- Guenther (2020), Choir Practice- Hamner (2020), Bus Passengers – Shen (2020), and Restaurant – Lu (2020)

An extension of the Wells-Riley equation was used to estimate predicted infection risk via aerosols in twelve published and unpublished COVID-19 outbreaks. Predicted infection risks were compared to observed attack rates in each event. To estimate a “credible interval” for model predicted infection risks, the quanta emission rate, the respiratory rate as well as the air volume flows were varied. The analysis assumes long range aerosol transmission in an ideally mixed environment.

In nine out of the twelve outbreaks the observed attack rates were in range with the predicted infection risk via aerosols and the corresponding ranges (with the variation of the boundary conditions).

Predicted Infection Risk via Aerosols (PIRA)/attack rate (AR)

Meat processing plant: 25% (17-35)/ 26%

Choir: 97% (88-99)/ 87%

Restaurant: 40% (35-56)/ 45%

Bus tour: 35% (19-58)/ 34%

The attack rates from all these outbreaks are reported to be in-line with estimated infection risk via aerosols.

Azimi (2021)^{Footnote 62}

In silico study

Cruise ship

Jan-Feb 2020

Note: Same outbreak described by Almilaji (2020) and Xu (2020)

Analysis of case data from the Diamond Princess cruise ship outbreak. Applied a framework based on stochastic Markov chain and negative exponential dose-response modeling with empirical data, to inform a modified version of the Reed-Frost epidemic model, to predict case count rates. Assumed infected individuals could be infectious upto 1 day post incubation period, effective incubation periods were estimated to be 5-11/13 days, and considered different modes of transmission.  

Note: Case data from early 2020 were included in the analysis.

712 COVID-19 cases were identified in 3711 passengers and crew members, yielding an attack rate of 19%.

Key estimates derived from the model included

• short-range droplets and aerosols (35%), long-range aerosols (35%), and fomite (30%) as modes of infection transmission
• large respiratory droplets (41%) and small respiratory aerosols (59%) as source of infectious virus
• Case transmission proportions prior to and after the passenger quarantine was 58% (±5% SD) and 42% (±5% SD), respectively. T
• Effective reproduction number before and after the quarantine period was 3.8 (±0.9 SD) and 0.1 (±0.2 SD).

Based on the modeled estimates the authors concludes smaller respiratory aerosols contributed to a greater proportion to infection transmission aboard the cruise ship, on average, across all time periods (i.e., both before and after passenger quarantine). It was estimated that aerosol transmission was the dominant mode of transmission (>70% of cases) despite the high ventilation rates (9-12 air changes per hour) with no air recirculation in the cruise ship.

Almilaji (2020)^{Footnote 63}

Cluster Investigation

Cruise ship

Jan-Feb 2020

Note: Same outbreak described by Azimi (2020) and Xu (2020)

Analysis of clinical and case count data from Diamond Princess cruise ship outbreak. Post quarantine symptomatic infection onset rates (SIRR) among lab confirmed cases were examined and the design of the cruise ship’s air conditioning system was considered.

Note: Case data up to February 20, 2020 were included in the analysis, and a median 5 day incubation period was assumed.

Infection rates among passengers in cabins without previously confirmed cases was 1.2%, which was higher than rates among passengers in cabins with previously confirmed cases 0.8%.

Based on this difference, the authors suggest airborne transmission of SARS-CoV-2 through the cruise ship’s ventilation system may have contributed to the outbreak, and explain the higher infection rates in cabins without previously confirmed cases.

Xu (2020)^{Footnote 64}

Preprint

Cluster Investigation

Cruise ship

Jan-Feb 2020

Note: Same outbreaks described by Azimi (2020) and Almilaji (2020).

Analysis of COVID-19 case data from the Diamond Princess cruise ship outbreak was analyzed based on individual risk factors, stateroom occupancy and the air conditioning (i.e. HVAC) system of the ship to explore the most plausible modes of transmission.

Case data from January 20 to February 18, 2020 were included in this analysis.

Daily infection rates for passenger cases (n=146) were predicted based on close contact vs. non-close contact status, and pre- and post-quarantine data (February 5 was the start of quarantine).

The investigators concluded most passenger cases were likely exposed before the passengers were quarantined and the cruise ship’s air conditioning system did not play a role in long-range aerosol transmission of COVID-19.

Groves (2021)^{Footnote 9}

Cluster Investigation

US

Jun 2020

new

Investigation of a COVID-19 outbreak associated with multiple exposure events in Hawaii between a stationary cycling class instructor, a kick boxing/personal training instructor and their clients.

Note: The average community transmission rate at the time of the outbreak was 2–3 cases per 100,000 persons per day, which suggests alternative exposures were unlikely.

20 cases were linked to two pre-symptomatic fitness instructors, at different fitness classes. The kick boxing/personal training instructor was a secondary case of the stationary cycling class instructor.

The index instructor wore a face mask during the stationary bike class but the participants did not. The instructor was > 6 feet away and facing participants during the class. The room windows and doors were closed and floor fans (for cooling) directed air at participants.

The transmission events with the second instructor are assumed to have occurred during small group kick boxing and personal training sessions when the instructor did not wear a mask and mask use was minimal among participants.

Both instructors taught classes prior to symptom onset and the following aggregate attack rates for were calculated:

• <1 day pre symptom onset (attack rate 95%)
• 1 to <2 days pre symptom onset (attack rate 13%)
• ≥2 days pre symptom onset (attack rate 0%)

The outbreak investigators suggest infection transmission to have been facilitated by the lack of face mask use, extended close contact, poor room ventilation, and aerosol emission during physical activity and loud speech.

Lendacki (2021)^{Footnote 10}

Cluster Investigation

US

Aug-Sep 2020

new

Investigation of a COVID-19 outbreak associated with indoor fitness classes that were conducted at ≤25% capacity (i.e., 10–15 persons) with participants ≥6 ft apart.

Note: the building was not originally designed for exercise classes and the buildings ventilation system was not assessed.

55 COVID-19 cases (49 confirmed and 6 probable cases) were identified among 81 individuals who participated in indoor high-intensity exercise classes, attack rate of 68%. Mask use during the exercise sessions was reported to have been infrequent and approximately 75%.

Outbreak was attributed to 2 index cases who attended multiple exercise classes when symptomatic and potentially infectious. Both attendees reported mask use ≤60% of the time in class (infrequent mask use).

The odds of infrequent mask use compared to consistent mask use during classes was more common among attendees with COVID-19 than attendees not diagnosed with COVID-19 (odds ratio [OR] = 3.5; 95% CI = 0.9–15.1). Attendees diagnosed with COVID-19 also reported attending more exercise classes (median of 5 vs. 3 classes)

The authors suggest infrequent mask use, increased respiratory exertion during exercise, aerosol transmission, and sub-optimal ventilation may have contributed to infection transmission in this outbreak.

Jang (2020)^{Footnote 65}

Cluster Investigation

South Korea

Feb-Mar 2020

Investigation of a COVID-19 outbreak associated with Zumba classes at 12 different fitness sports facility locations following an instructor workshop in Cheonan, South Korea.

The initial transmission event is assumed to have occurred among instructors at a 4-hour workshop where 8 of the 27 attendees tested positive for SARS-CoV-2. In the following weeks case counts associated with infected instructors grew to 112 cases across multiple fitness facilities.

The workshop attack rate was 26.3% (95% CI 20.9%–32.5%) and the secondary attack rate from 8 instructors was 4.10% (95% CI 2.95%–5.67%, 830 close contacts).

The investigators state approximately half of identified cases (50.9%) were due to transmission from instructors to fitness class participants; 38 cases (33.9%) were in-family transmission from instructors and students; and 17 cases (15.2%) were from transmission during meetings with coworkers or acquaintances.

No secondary cases were observed among Pilates and yoga class students, led by an infected instructor.

Authors state intense physical activity, large number of participants in a fitness class (i.e. crowded space), and the moist warm atmosphere of the sports facility may have contributed to high rates of infection in the outbreak.

Brlek (2020)^{Footnote 66}

Cluster Investigation

Slovenia

Feb-Mar 2020

Investigation of a COVID-19 cluster linked to a squash court in a fitness facility.

The cluster involved 6 cases assumed to be linked through indirect transmission of infection.

Epidemiological investigation indicated the index case developed symptoms during the game of squash, and four confirmed and one suspect case were linked to the same squash hall and potentially the same change rooms. None of the cases shared sports equipment or had contact with the facility staff. No additional cases were identified.

Authors suggest the infection transmission within the cluster likely occurred due to aerosolization of virus in the indoor setting including small confined space, inadequate ventilation and strenuous physical activity.

Hwang (2021)^{Footnote 11}

Cluster Investigation

South Korea

Aug 2020

new

Investigated the role of aerosol transmission in a COVID-19 outbreak associated with an apartment building in Seoul.

10 COVID-19 cases in 7 households were identified along two vertical lines of apartments in a building, each line of apartments were connected through a single air duct in the bathroom for natural ventilation. Attack rate of 2% among the apartment building residents (n=10/437).

None of the tested surfaces, including household ventilation grills and drains were positive for viral RNA.

Transmission through droplets when occupying the same common space (e.g., an elevator) was assumed to have been unlikely due to the spatial arrangement of case apartments.

The outbreak investigators assume vertical air flow through a single air duct in the bathroom (consistent with air flow in a vertical shaft) to have spread the infection to upstairs and downstairs apartment residents in the building.

Lin (2021)^{Footnote 12}

Cluster Investigation

China

Jan – Feb 2020

new

Note: A different analysis of the cluster described by Kang (2020).

Investigation of a COVID-19 cluster linked to families living in three vertically aligned units of the same apartment building. The index family reported possible travel related exposure in Wuhan, but the two other families with subsequent cases did not. Tracer gas was used to simulate air flow among units.

10 COVID-19 cases were identified in 3 households, phylogenetic analysis confirmed all cases were infected by the same strain.

Video surveillance footage did not identify exposures between the index household members and other cases.

Investigators conclude infection transmission to have likely occurred through drain pipes connected to the toilet which were connected to sewer pipes and flood drains, a system that was shared by the vertically aligned apartments.

Kang 2020^{Footnote 67}

Cluster Investigation

China

Jan – Feb 2020

new

Note: A different analysis of the cluster described by Lin (2021).

Investigate infection transmission among three families living in the same apartment building. The index family reported possible travel related exposure in Wuhan, but the two other families with subsequent cases did not. Investigators used Ethane tracer gas as a surrogate for gas in the buildings drainage system and computational fluid dynamics to investigate possible sources of infection and transmission among families.

10 COVID-19 cases among three families who lived in vertically aligned apartments connected by drainage pipes in the master bathrooms. Attack rate of 4% among apartment building residents and staff (n=10/217)

No exposure from the building’s elevators were identified, and viral RNA was not detected on elevator buttons or air vent surfaces. The surfaces most frequently contaminated with viral RNA were in the master bathrooms, suggesting these areas to be probable infection transmission sites.

Based on the epidemiological and in-silico analyses the investigators conclude infection transmission from the index family to the other two families likely occurred through fecal aerosols traveling within vertical drainage stacks of the apartment building.

Note: none of the collected air samples from this setting were positive for viral RNA.

Jiang (2020)^{Footnote 13}

Cluster Investigation

China

Jan – Feb 2020

new

Investigation of 43 SARS-CoV-2 cases linked to a cluster at a department store in Baodi, China. Epidemiological data, video surveillance footage, store layout and ventilation conditions were considered in the investigation.

43 COVID-19 cases linked to the outbreak at a department store: 6 salespersons, 18 customers, and 19 of their close contacts. The close contact cases were determined to have been secondary cases without exposure to the department store.

The investigators conclude aerosols to have been a significant mode of transmission among 11 cases of the outbreak. The index case was not identified but it was assumed to have been one of the infected salespersons.

Cai 2020^{Footnote 68}

Cluster Investigation

China

Jan 2020

Investigation of a SARS-CoV-2 cluster linked to a shopping mall. Clinical, epidemiological and laboratory (RT-PCR) data of cases was analyzed to assess possible modes of infection transmission.

Two shopping mall co-workers were the index cases: this was associated with 7 infections among co-workers on the same floor, 7 mall staff from other floors, 10 mall shoppers, and 2 close case contacts outside of the mall. Shoppers and co-workers from other floors denied close contact with the index cases.

Based on the available data the authors suggest infection spread could have resulted from spread via fomites or virus aerosolization in a confined public space (e.g. restrooms or elevators). 

(est) = Estimated based on author affiliations and publication date.
LTE= letter to the editor

Edwards (2020)^{Footnote 27}

In Vivo Study

US

Oct 2020 (est)

Eight non-human primates Macaca mulatta (rhesus macaque) and Chlorocebus aethiops (African green monkey)) were infected with aerosols (≈2 µm) containing SARS-CoV-2 (~2.5x103 TCID50) using a laboratory inhalation system

Mucosal sampling by nasal swabs showed viral RNA detected as early as +1 day post infectious aerosol exposure with viral titers reaching peak levels at +7 days post infection, and clear decline of RNA by day +14 days post infection, and undetectable by +28 days post infection. These findings confirm SARS-CoV-2 infection by infectious aerosols can occur.

Mucosal sampling by nasal swabs showed viral RNA detected as early as +1 day post infectious aerosol exposure.

Exhaled breath particle production started 3 days post infection rose to day 7 and decreased to baseline by day 14 in primates. Exhaled breath particle production was temporally consistent with viral replication in nasal swab samples.

There was a significant association between exhaled breath particles and viral load in most primates and correlated with viral kinetics.

Viral RNA was undetectable in nasal swab samples of infected primates by day 28 post-infection.

Port (2020)^{Footnote 17}

Preprint

In Vivo Study

US

Dec 2020 (est)

new

Three groups of female Mesocricetus auratus (Syrian hamsters) (n=36) were infected with SARS-Cov-2 virus aerosols and other types of exposures (i.e., fomites and intranasal). Infected animals were compared to unexposed controls (n=12). Exposure by SARS-CoV-2 aerosols (1.5x10³ TCID50) was by a 3 jet collision nebular; particle size ranged from 1-5 um. Viral shedding and replication patterns measured in respiratory tissue and by fecal and oropharyngeal swabs.

During separate airborne transmission experiments two transmission pairs (n=4) were co-housed in cages separated by a perforated plastic divider that prevented direct contact. The susceptible animals were placed in the direction or against the air flow from infected animals; four transmission pairs (n=8). Infection detected based on serum seroconversion.

The authors found mode of SARS-CoV-2 infection transmission played a factor in disease severity, viral load, and virus shedding in the animal model.

Early virus replication (1 day post infection), peak respiratory shedding of virus at 2 days from exposure, and higher early viral loads in lung and tracheal tissue (p = <0.0001), was observed among aerosol exposed animals when compared to other exposure types. Viral titers in lung tissue showed a positive relationship with upper and lower respiratory tract pathology and weight loss, which lead the authors to suggest early respiratory shedding (as observed among animals infected by aerosols) may predict acute disease manifestation.

In the indirect contact experiments assessing airflow based transmission no symptoms linked to infection were observed in the susceptible animals but 25% (n=1/4) seroconverted. This seroconversion (i.e., virus exposure) was linked to directional airflow from the infected to host animals.

Of note, it was observed fomite transmission was linked to delayed disease manifestation with longer duration of time between exposure and viral replication in respiratory tissue, which led to reduced disease severity.

Zhang (2021)^{Footnote 18}

Preprint

In Vivo Study

US

Jan 2021 (est)

new

Investigated the risk of SARS-CoV-2 infection transmission from aerosol exposure, and risk of re-infection at re-challenge between naïve infected and previously infected transmission pairs of Mesocricetus auratus (Syrian hamsters).
In the aerosol transmission experiments infected and donor hamsters (n=6) were housed in transmission cages with wire mesh partitions that prevent indirect and direct contact between animals, but permitted airflow. Live virus levels in respiratory tissue over the post exposure period was measured to confirm re-infection.

Experiments found SARS-CoV-2 was efficiently transmitted from infected naïve hamsters to previously infected hamsters by airborne transmission, among all transmission pairs.

Based on viral RNA levels in respiratory tissue, the investigators conclude prior infection to have provided provide some good protective immunity but not complete immunity against re-infection at re-challenge; replicating live virus was present in the re-infected animals.

Sia (2020)^{Footnote 28}

In Vivo Study

Hong Kong (est)

May 2020 (est)

Experimental study that investigated SARS-CoV-2 infection transmission via aerosols in Mesocricetus auratus (Syrian hamsters). Infected and susceptible animals were housed in adjacent wire cages placed 1.8 cm away from one another (3 different pairs) were exposed to one another for 8 hours.  

Efficient indirect transmission of infection to susceptible hamsters occurred in all three pairs within experimental settings.

Peak viral load in aerosol exposed hamster was at 3 days post contact.

Kutter (2020)^{Footnote 29}

In Vivo Study

Netherlands (est)

Oct 2020 (est)

An experimental set-up where infected and susceptible pairs (i.e. transmission pairs) of individually housed Mustela putorius furo (ferret) (n=8) were connected through a hard 15 cm air duct opening with multiple 90° turns. Airflow was directed upwards from the donor to indirect recipient animals. Air travelled an average of 118 cm (approx. 3 ft.) through the tube systems.

Note: transmission of SARS-CoV-2 and SARS-CoV was investigated.

Indirect transmission of SARS-CoV-2 between two ferrets more than 1 meter away was confirmed in 50% separately housed transmission pairs.

Infection in susceptible animals were confirmed through the detection of viral RNA in throat and nose swabs.

Kim (2020)^{Footnote 30}

In Vivo Study

Hong Kong (est)

May 2020 (est)

Experimental study of SARS-CoV-2 transmission via aerosols in Mesocricetus auratus (Syrian hamsters). Infected and susceptible and infected ferrets.

Efficient indirect transmission to have occurred among indirect contact ferrets.

(est) = Estimated based on author affiliations and publication date.

Dabisch (2020)^{Footnote 19}

Simulation experiments

US (est)

Aug 2020 (est)

new

Measured the viability and persistence of SARS-CoV-2 in artificially generated aerosols within a drum aerosol chamber across a range of temperatures (10 °C, 20 °C, 30 °C and 40 °C), relative humidity (20%, 45% and 70%), and simulated sunlight levels of zero integrated UVB irradiance (i.e., indoor/night/darkness), 0.9 W/m2 UVB irradiance and 1.9 W/m2 integrated UVB irradiance (i.e., summer midday sunlight). Viral concentration was held at mean value of 2.3 ± 0.4 log₁₀ TCID₅₀/L-air during the experiment. Infectiousness of virus within aerosols was measure using micro titration assay and Vero cells. Experiment results informed a regression model predicting aerosolized SARS-CoV-2 decay under variable conditions.

The time needed for a 90% decrease in infectious virus at 40°C, 20% relative humidity increased from 4.8 min with sunlight representative of noon on a clear summer day outdoors, to more than 2 hours under conditions representative of indoors or at night.

Across other temperature and humidity levels the decay per minute in midday sun simulations ranged from. 38.1% ± 8.9% per minute at 40 °C and 20% relative humidity, to 18.9 ± 4.8% per minute at 10 °C and 20% relative humidity. For moderate sunlight representing spring and fall intensity at 40 °C, decay rates ranged from18from.0 ± 6.2% per minute at 30 °C and 45% relative humidity, to 11.1 ± 4.6% per minute at 10 °C and 20% relative humidity.

In the absence of sunlight <2% decay (mean) per minute rate was estimated, for most tested temperature and humidity levels.

Exceptions were in the higher temperature and/or humidity ranges,

• Decay rate at 30 °C with 70% relative humidity was 6.3 ± 2.6% per minute.
• Decay rate at 40 °C with 20% relative humidity was 3.9 ± 0.4%.

In the regression analysis sunlight had the largest influence on decay followed by temperature and a sunlight x temperature interaction. Humidity (relative or absolute) had the lowest influence on the decay constant. Thus, more intense sunlight and high temperatures >30°C had faster decay rates.

Fears (2020)^{Footnote 31}

Simulation experiments

US (est)

Sep 2020 (est)

The long-term persistence of artificially generated viral aerosol suspensions of SARS-CoV-2 was measured at different time intervals. Viral contents were quantified by RT-PCR, and infectiousness of virus was measured by plague assay. Samples were qualitatively assessed by electron microscopy.

Infectious SARS-CoV-2 was detected at 10 minutes, 30 minutes, 2, 4, and 16 hours during the aerosol suspension stability experiment.

A minimal reduction in viral genome copies in aerosol samples (as measured by RT-PCR) was noted for the measured time points.
A minor but constant fraction of the SARS-CoV-2 virus in aerosols maintained replication-competence at all measured time points, including at 16 hours.

Qualitative assessment of virion integrity revealed virions were either ovoid or spherical in shape, and maintained the expected morphologies up to 16 hours in aerosol suspension.

Van Doremalen (2020)^{Footnote 32}

LTE

Simulation experiment

US (est)

Spring 2020 (est)

In this experiment SARS-CoV-2 and SARS-CoV-1 virus titer stability and decay was measured from artificially generated aerosols, in Vero 6 cell culture. Analysis used a Bayesian regression model.

SARS-CoV-2 virus remained viable in experimentally generated aerosols up to 3 hours (duration of the experiment), with a reduction in infectious titer from 10^3.5 to 10^2.7 TCID₅₀ per liter of air

In aerosols the half life of SARS-CoV-2 virus was estimated to be 1.1-1.2 with a 95% credible interval of 0.64-2.64.

(est) = Estimated based on author affiliations and publication date.
LTE= letter to the editor

Ryan (2020)^{Footnote 34}

Exhaled breath condensate monitoring study

Ireland

Apr - May 2020

new

Exhaled breath condensate samples were collected from COVID-19 patients using RTUBE condensers. The sample included nasopharyngeal swab positive (n=16) and nasopharyngeal swab negative with clinical diagnosis of COVID (n=15) patients. Additional samples from pre-SARS-CoV-2 were included as controls (n=14). Virus in samples was detected by RT-PCR using different viral gene assays.

Note: Clinical diagnoses of COVID-19 were based on clinical expertise and imaging results.

93.5% (29/31) of the collected exhaled breath samples from SARS-CoV-2 patients (clinically confirmed and/or positive nasopharyngeal swab results) were positive by RT-PCR targeting all four genes (E, S, N, ORF1ab). All pre-pandemic control samples were negative. In this study exhaled breath was shown to be a sensitive and non-invasive sample type.

Positivity of samples varied by RT-PCR assay target sequence among nasopharyngeal swab negative patients (n=15):

• 66% (10/15) positivity for viral envelope (E)/ spike (S) proteins gene assays (used for the nasopharyngeal swab).
• 73% (11/15) positivity for viral nucleocapsid (N)/open reading frame (ORF1ab) gene assays.

The combined results equated to 14/15 clinically diagnosed cases positive on at least one assay.

Zhou (2021)^{Footnote 33}

Exhaled breath condensate monitoring study

China

Feb - Mar 2020

new

Note: Additional results on viral RNA in air samples are summarized in Table 5.

COVID-19 patients (n=10) about to be discharged from hospital (as per negative throat and nasal swabs) were recruited from multiple hospital sites. Exhaled breath condensate was sampled using using a BioScreen II device.

SARS-CoV-2 RNA in breath samples was quantified using RT-PCR.

22.2% of 9 COVID-19 patients about to be discharged from hospital had SARS-CoV-2 in their exhaled breath samples at a concentration of ~105 RNA copies/m3. Both patients were over the age of 70.

It was estimated that some COVID-19 patients in the sample were exhaling the virus at a rate of ~1400 RNA copies per minute into the air at discharge.

Ma (2020)^{Footnote 35}

Preprint

Exhaled breath condensate monitoring study

China

Spring 2020 (est)

Note: Additional results on viral RNA in environmental air samples are summarized in Table 5.

Exhaled breath condensate samples were collected from COVID-19 patients (n=30) using a BioScreen device.

The study confirms the emission of SARS-CoV-2 virus RNA into the air from exhaled breath condensate of infected individuals (16.7% n=5/30). The positive samples were detected either <3 days from symptom onset (n=3) or within 7-14 days from symptom onset (n=2).

SARS-CoV-2 levels in exhaled breath were estimated to reach 105-107 copies/m3 if an average breathing rate of 12 L/min is assumed and is highest during early stages of infection.

Di Carlo (2021)^{Footnote 20}

Exhaled air monitoring study

Italy

Apr - Jun 2020

Note: Additional results on viral RNA in environmental air samples are summarized in Table 5

new

Tests the release of SARS-CoV-2 RNA into the air during normal breathing – without coughing, sneezing or talking among infected cases (n=5). These cases were all in hospital patients admitted to hospital for non-COVID symptoms who later developed COVID symptoms during hospital stay.

Sampled individuals were housed in airborne infection isolation rooms (AIIR) and exhaled breath samples were collected using Sartorius AirPort sampler. Patient oropharyngeal, nasopharyngeal and salivary swabs samples were also collected at the time of breath samples.

The days since symptom onset to sample collection varied from 7- 56 days in the patient sample. Oropharyngeal and nasopharyngeal swabs of 4 patients were positive at the time of sampling.

Viral RNA was detected at 1 cm distance from patients’ mouths in two patients (40% n=2/5). Both had positive oropharyngeal, nasopharyngeal and salivary swabs. Salivary swabs were negative in the patients with negative exhaled air samples who were reported to be infected based on consecutive positive swab samples. In one of these patients, wearing a surgical mask effectively blocked viral RNA detection at 1 cm (masked samples were not collected from the other patient with positive exhaled air samples).

Air samples at 1 cm from RT-PCR negative patients were negative for viral RNA.

Feng (2020)^{Footnote 36}

Exhaled air & exhaled breath condensate monitoring study

China

Feb - Mar 2020

Note: Additional results on viral RNA in environmental air samples are summarized in Table 5.

Sampled exhaled breath and environmental air of COVID-19 patients using a NIOSH bio-aerosol sampler. Exhaled breath condensate was sampled using a sterile laboratory-made collection system. Air samples were segregated by aerosol size. Samples were collected from COVID-19 patients in the later stages of infection in hospital settings.

SARS-CoV-2 RNA was not detected in any of the patients’ expired breath samples (n=0/9). Viral RNA was isolated in exhaled breath condensate (25%; n=2/8), and bedside environmental air samples (8%; n=1/12).

The authors attributed minimal contamination of viral RNA in study samples to reduced respiratory viral shedding among patients in later stages of infection.

Ding (2020)^{Footnote 37}

Preprint

Exhaled air and exhaled breath condensate monitoring study

Hong Kong

Feb 2020

Note: Additional results on viral RNA in environmental air samples are summarized in Table 5.

Exhaled breath condensate samples (n=2) and expired air samples (n=2) were collected from COVID-19 patients housed in airborne infection isolation rooms (AIIR). Multiple devices were used for air sample collection (n=27), which was conducted on different days.

Note: sample collection distances from patient(s) are not reported.

All collected exhaled condensate samples and expired air samples were negative for SARS-CoV-2 RNA.

(est) = Estimated based on author affiliations and publication date.

Lednicky (2020)^{Footnote 40}

Biological monitoring study

US (est)

Nov 2020 (est)

Air samples were collected in triplicate from hospital rooms of COVID-19 patients in the absence of aerosol generating procedures. Air samples were collected using a VIVAS air sampler at 2 to 4.8 meters away from patients, with and without a HEPA filter on the air sampler inlet tube. Virus in isolated air samples was measured using RT-PCR, and infectiousness was measured based on cytopathic effects in cell culture (LLC-MK2 and Vero-E6). The genomes of isolated virus was sequenced.

Viable (infectious) SARS-CoV-2 was found to be present in aerosols sampled from hospital patient rooms by RT-PCR and cell culture (via cytopathic effects)

All air samples collected without a HEPA filter was positive for viral RNA.

A single nearly complete virus sequence was isolated from the air samplers that collected environmental air. This genetic sequence matched the virus strain isolated from nasopharyngeal sample of one of the two patients who occupied the room during sampling. The matched person was diagnosed with acute infection at the time of air sampling.

Santarpia (2020)^{Footnote 41}

Preprint

Biological monitoring study

US

Apr 2020

Patient generated aerosols in hospital air samples were collected using a NIOSH BC251 aerosol sampler at the foot of COVID-19 patient beds. Aerosol sizes and concentration was concurrently measured during sample collection using an Aerodynamic Particle Sizer Spectrometer. Aerosols were distinguished by the proportion of different sizes (>4.1 μm, 1-4 μm, and <1 μm) among samples.

Presence of the virus in isolated aerosols (<5μm) was measured using RT-PCR, western blot, and transmission electron microscopy and infectiousness of isolated viral particles was examined using cell culture (Vero-E6).

Note: The study does not specify if the patients were housed in regular wards or AIIR.

RNA was detected in all six patient rooms, and included all aerosol particle size fractions (defined as >4.1 μm, 1-4 μm, and <1 μm).

Replicating virus in cell culture was observed in most <1 μm aerosol samples, two of the 1-4 μm size aerosol samples and two of the >4.1μm samples.

Western blot and TEM analysis of these samples also showed evidence of viral proteins and intact virions.

The authors conclude the infectious nature of the aerosols collected in this study suggests that aerosol transmission of COVID-19 is possible.

Santarpia (2020)^{Footnote 42}

Biological monitoring study

US

Mar 2020

Air samples from negative pressure isolation spaces and wards housing COVID-19 cases were collected using a Sartorius Airport MD8 air sampler and tested for SARS-CoV-2 viral RNA by RT-PCR. A subset of positive samples were examined for viral propagation in Vero E6 cells. Several indicators were utilized to determine viral replication including cytopathic effect, immunofluorescent staining, time course PCR of cell culture supernatant, and electron microscopy.

63.2% of in-room air samples were positive by RT-PCR (mean concentration 2.42 copies/L of air).

Two samples placed at different proximity to a patient, including a sample from <2 meters away the patient, were positive. Viral concentration was higher in the air sample collected closer to the patient (4.07 vs. 2.48 copies/L of air).

58.3% of air samples collected from hallways were positive (mean concentration of 2.51 copies/L of air). Some viral replication was observed in a single positive sample collected from a hallway.

Munoz-Price (2021)^{Footnote 24}

Biological monitoring study

US

Spring 2021 (est)

new

Note: Additional results on viral RNA in environmental air from household settings are summarized in Table 6.

Air samples (n=16) collected from hospital ICU inpatient rooms occupied by confirmed COVID-19 cases. Air samples were collected using the Sartorius MD8 airscan sampler, positioned at 0.3 -1.8 meters from patients’ head. Viral RNA in samples was measured by RT-PCR.

12.5% (n=2/16) of hospital air samples were only positive for viral RNA. These samples where collected at 0.3 meters from confirmed cases. The patients had mild severity of illness and were not on supplemental oxygen at the time of samplesamples collection.

Razzini (2020)^{Footnote 69}

Biological monitoring study

Italy

Apr 2020 (est)

new

Air samples (n=5) collected from hospital areas with COVID-19 cases using a MD8 Airport
Portable Air Sampler with gelatin
membrane filters. Samples were collected from three zones, contaminated (corridor for patients and ICU), semi-contaminated (undressing room) and clean areas (medical staff dressing and locker areas).

Viral RNA in samples was measured by RT-PCR.

All the air samples collected from COVID-19 cases’ ICU and corridor were positive for viral RNA; samples from other areas were negative.
Viral concentration in (mean Ct) across ICU air samples was 22.6, and 31.1 for corridor air samples.

Ge (2020)^{Footnote 70}

Biological monitoring study

China

Jun 2020 (est)

new

Air samples (n=33) were collected using the NIOSH bioaerosol sampler BC251. Sampled settings were ICU, hemodialysis clinics, fever clinics, and respiratory wards.

Viral RNA in samples was measured by RT-PCR.

All air samples (n=3) from the ICU setting were positive for viral RNA. Viral concentrations in samples ranged from Ct 36.5 - 37.8.

Hu (2020)^{Footnote 71}

Biological monitoring study

Feb – Mar 2020

China

new

Air (aerosol) samples were collected from multiple hospital sites using a centrifugal aerosol-to-hydrosol sampler (n=123). Viral RNA in samples was measured by RT-PCR and viability by Vero-E6 cell culture.

Eight air samples (21%) from ICU environments and one same from (16%) from the CT (computerized tomography) room were positive for viral RNA. The range of virus concentrations in the positive air (aerosol) samples was 1.11 × 10³ to 1.12 × 10⁴ RNA copies m⁻³. The virus could not be cultured from positive samples.

Seyyed (2020)^{Footnote 72}

Biological monitoring study

Iran

May 2020 (est)

new

Air samples (n=10) were collected from COVID-19 patient ICU wards using a sampling pump with a porous midget impeller. Viral RNA in samples was measured by RT-PCR.

60% of air samples were positive for viral RNA. The highest RNA concentrations were measured in samples from between patient beds (3913 copies per ml). These were air samples collected at 1.5 to 2 meters from patient beds.

Moore (2020)^{Footnote 43}

Biological monitoring study

England

Mar –May 2020

new

Air samples (n=55) were collected from COVID-19 patients with and without respiratory symptoms, from eight different hospitals. Sampled settings included AIIR, general COVID-19 wards and non-COVID-19 wards. Samples were collected using a Coriolis μ air sampler and/or Sartorius MD8 air sampler.

Viral RNA in samples was measured by RT-PCR, and viability by Vero-E6 cell culture.

7% (n=4) of air samples were positive for viral RNA, at low concentrations from <10 to 460 genomic copies/m³ air. The virus could not be cultured from positive samples. All positive samples were collected at 1 meter distance from patients, and days since symptom onset among the patients ranged from 8-10 days.

Hernández López (2021)^{Footnote 44}

Biological monitoring study

Jan 2021 (est)

Mexico

new

Air samples (n=15) from emergency, internal medicine and COVID-19 patient rooms and multiple bed patient care rooms in two hospitals were collected using a Millipore sampler.

Viral RNA in samples was measured by RT-PCR.

Three air samples (20%) were positive for viral RNA. All positive samples were from the COVID-19 patient room and near the patient.

Tan (2020)^{Footnote 73}

Biological monitoring study

China

Mar 2020

new

Air samples (n=12) were collected from ICU and isolation wards of COVID-19 cases in hospital, and corridors. Air samples were collected at 1 meter from a patient’s head.

A single air sample from a patient care area, collected when the patient was undergoing intubation, was positive for viral RNA.

Gehrke (2021)^{Footnote 45}

Preprint

Biological monitoring study

Germany

Oct 2020 - Jan 2021

new

Note: Additional results on viral RNA in community setting air samples are summarized in Table 6.

Air samples were collected from patient rooms in two COVID-19 isolation units, and an outpatient endoscopy facility using non-powered cold traps. Viral RNA was quantified by RT-PCR.

Hospital setting samples:

No viral RNA was isolated from the first isolation unit ventilated permanently by two windows. Viral RNA was detected in samples collected from a non-ventilated corridor next to the isolation room.

In the second isolation unit samples, SARS-CoV-2 RNA levels reached concentrations of 10⁵/mL in non-ventilated rooms, but when windows were open to increase ventilation SARS-CoV-2 RNA concentrations in samples were reduced to 10⁴/mL or less.

In the endoscopy facility, 50% of cold traps were positive for viral RNA (n=6/12). 57% of cold traps (n=4/7) from endoscopy operation rooms were positive for SARS-CoV-2, but RNA concentrations (initially at 12 copies/ml) and the number of positive samples were reduced when the ventilation levels in the room was increased.

Zhou (2021)^{Footnote 33}

Biological monitoring study

China

Feb- Mar 2020

new

Note: Additional results on viral RNA in exhaled breath samples are summarized in Table 4.

Air samples (and exhaled breath samples) of hospitalized COVID-19 patients were collected from multiple hospital sites. Air samples were collected from hospital corridors, waste storage rooms, ICU rooms, toilets, medical preparation rooms, clinical observation rooms, and general wards. Air samples were collected using the Air-nCoV-Watch (ACW) system and viral RNA quantified by RT-PCR.

6.8% of the air samples (n =3/ 44) were positive for viral RNA at digital PCR concentration levels ranging from 9-219 viruses/m³.

Yarahmadi (2021)^{Footnote 47}

Biological monitoring study

Iran

Feb 2021 (est)

new

Air samples were collected (n=20) from an ICU ward at, 1) a COVID-19 patient breathing zone (i.e., side table next to patient’s head), general area (10 meters from the ICU unit), and breathing zone of health care personnel near the COVID-19 patient’s bed, and 1 meter from the patient’s bed. Samples were collected using NIOSH and ASHRAE bioaerosol samplers. Viral RNA was detected by RT-PCR.

50% (n=2/4) of samples from the patient breathing zone were positive for viral RNA. The positive samples were from a confirmed case, and negative samples were from a suspected case.
12.5% (n=1/8) of samples from health care personnel breathing zone were positive for viral RNA.

12.5% (n=1/8) of samples from the general area were positive for viral RNA.

Authors suggest bioaerosols maybe present due to the re-aerosolization of previously airborne SARS-CoV-2 particles from health care personnel walking between different wards and stations of the ICU.

Ong (2021)^{Footnote 46}

Biological monitoring study

Singapore

Jan 2021 (est)

Air samples were collected from AIIR in a hospital and community isolation facility (naturally ventilated) housing confirmed COVID-19 patients, at 1 meter away from the patient. Air samples were collected using a BioSpot-VIVAS BSS300-P bioaerosol sampler (collecting particles <4.34 µm in size). NIOSH aerosol samplers were used to validate the Biospot results. SARS-CoV-2 RNA was detected by RT-PCR, and cultured using Vero C1008 cell culture.

50% of BioSpot air samples from hospital AIIR (n=6/12) were positive for SARS-CoV-2 RNA (concentrations ranging from 178.9 to 2,738.4 copies/m3). Positive samples were collected from rooms with at least one symptomatic patient (<7 days post symptom onset). NIOSH aerosol samplers detected SARS-CoV-2 RNA in aerosols <1 µm, 1–4 µm, and >4 µm in diameter.

Only 1 of 9 samples from the community isolation facility was positive for SARS-CoV-2 RNA, with a concentration of 978.3 copies/m³.

Virus cultures of all positive samples were negative.

Dumont-Leblond (2020)^{Footnote 22}

Biological monitoring study

Canada

Mar-June 2020

new

Sampled air from AIIR rooms from a Quebec hospital housing non-severe COVID-19 patients (n=22). Air samples were collected (with 3 µm gelatine filters and 0.8 µm polycarbonate filters) using a SASS 3100 dry sampler, at 1.5m away from the patient bedside. Viral contamination was measured by RT-PCR and Vero E6 cell culture.

11% (n=11/100) of air samples collected at the bedside of 6 patients were positive for viral RNA. These patients had higher frequency of fever, dyspnoea, and cough compared to others in the sample. Most were males and had a slightly longer duration of hospitalization.

Average SARS-CoV-2 viral load in the positive sample rooms was estimated at 4.86E+4 virus genomes per hour.
No virus isolated from air samples could be cultured.

Binder (2020)^{Footnote 74}

Biological monitoring study

US

Apr-May 2020

new

Environmental swabs and aerosol samples were collected from single patient hospital rooms housing COVID-19 patients (n=20, 16 symptomatic and 4 asymptomatic) within a dedicated COVID-19 ward. Bioaerosols were collected using a NIOSH BC 251 sampler placed at different distanced from the patients’ heads (1, 1.4, 2.2, and 3.2 meters). Empty hospital rooms, hallways, staff break room and staff workstations were sampled as controls. SARS-CoV-2 viral contamination was measured by RT-PCR, and Vero EC6 cell culture.

Three samples collected from three of the COVID-19 patient rooms were positive for viral RNA; samples were collected at 1.4-2.2 meters from the patients. These patients were 4-10 days post symptom onset (runny nose, headache, fever, cough, difficulty breathing, fatigue, loss of smell, gastrointestinal symptoms).

Viable virus could not be identified in bio aerosols samples via cell culture.
Samples from other area were negative for SARS-CoV-2 RNA.

Passos (2021)^{Footnote 75}

Biological monitoring study

May-Aug 2020

new

Air samples were collected from two different hospital environments (n=52) using active and passive aerosol sampling methods.

Active sampling used multiple portable/hand-held air sampler models (AIRIDEAL 3P, MD-8 AirPort, HANDI-VOL) and filter types (cellulose, PTFE or quartz microfiber filters). The passive method used petri dishes for the collection of sedimentable particles. Both hospitals housed COVID-19 patients in dedicated ICU wards at the time of sample collection. Aerosol samples were also collected from multiple outdoor public spaces using high-volume air samplers (AGV, Energética).

SARS-CoV-2 RNA in samples was detected by RT-PCR.

Air samples from four areas sampled in one hospital (at 100% ICU occupancy at the time of sampling) were positive for viral RNA in suspended particles (filter pore size >0.2 μm and >0.3 μm), as well as 11% of sedimentable particles (n=4/36).

Positive suspended air particle samples were detected from inside the COVID-19 ICU ward (concentration of 0.33 genomic units m⁻³ @ distance >2 from patient), in a protective apparel removal room (0.14 genomic units m⁻³), storage room for patient mobile toilets and soiled patient linens (0.19 genomic units m⁻³). The storage room was open to natural ventilation.

SARS-CoV-2 RNA positive sedimentable particles were detected in samples collected from an external corridor adjacent to the COVID-19 dedicated ICU.

All outdoor air samples and samples collected at the other hospital (i.e., Hospital 1) l (at 33% occupancy at the time of sampling) were negative for SARS-CoV-2 RNA.

Lane (2021)^{Footnote 48}

Biological monitoring study

Jan 2021 (est)

new

Air samples were collected from multiple hospital units, AIIR, ICU nursing stations, family/visitor corridors outside of ICUs, medical unit, and patient room hallways. NIOSH 251 2-stage cyclone samplers collected air samples (limit of detection of 8 viral copies/m³ of air).

Two patient rooms housing COVID-19 patients were used as positive control samples. Viral RNA was detected by RT-PCR.

The two positive control samples collected from two confirmed COVID-19 patient rooms were positive for viral RNA.

None of the air samples (n=528) from other sampling area were positive for SARS-CoV-2 RNA.

Liu (2020)^{Footnote 76}

Biological monitoring study

China

Feb-Mar 2020

SARS-CoV-2 RNA concentration and aerosol size distributions in air samples (n=30) from multiple sites within or near a hospital and field hospital.

All aerosol samples (n=30) were collected on pre-sterilized gelatin filters (Sartorius). Three size-segregated aerosol samples were collected using a miniature cascade impactor (all sampled from staff areas). Viral RNA was detected by RT-PCR.

SARS-CoV-2 contamination in air samples was low to undetectable.

In the field hospital setting, the greatest suspended SARS-CoV-2 RNA in aerosols was identified in a temporary patient toilet room (1 m² area) with low ventilation, likely from the patient breathing or aerosolization of virus from feces and urine of infected patients.

Samples from the field hospital staff personal rooms demonstrated the greatest virus concentrations. Aerosols from 0.25 to > 2.5 μm were identified. The authors hypothesize this came from healthcare worker PPE surfaces and apparel. Low but detectable viral RNA concentrations were found at a department store entrance and an outdoor site near the hospital suggesting this may have occurred due to high traffic flow and crowding.

Note: The specific concentrations of airborne SARS-CoV-2 in each aerosol sample by site are provided in the publication.

Chia (2020)^{Footnote 77}

Biological monitoring study

Singapore

Spring 2020 (est)

Detection of air contamination by SARS-CoV-2 in airborne infection isolation rooms (AIIR) housing COVID-19 patients, in hospital settings. Air samples were collected, and aerosol sizes were measured by NIOSH BC 251 bio-aerosol samplers. Viral RNA was detected by PCR.

66% (n=2/3) of the air samples collected from AIIR environments were SARS-CoV-2 RNA positive. The smallest aerodynamic size fraction that contained detectable levels of SARS-CoV-2 RNA was 1–4 µm.

Total SARS-CoV-2 concentrations in air ranged from 1.84 × 10³ to 3.38 × 103 RNA copies per m³ air sampled.

The authors suggest the presence of SARS-CoV-2 in the sampled air is likely highest during the first week of illness, when respiratory viral load is high.

Jin (2021)^{Footnote 78}

Biological monitoring study

China

Feb-Mar 2020

new

Air samples were collected from an ICU housing a single ready-for-discharge COVID-19 patient, 2 days after the patient tested negative for SARS-CoV-2. Air samples were collected from the isolation room and staff PPE dressing room. High-volume air samples were collected using a WA 400 Portable viral aerosol sampler. Viral RNA in samples was detected using RT-PCR.

A single air sample from the ICU ward isolation room was positive (1/7, 14.29%) for viral RNA.

The authors state their findings suggest the virus may be shed via aerosol for days, even after a patient has tested negative.

Zhou (2020)^{Footnote 79}

Biological monitoring study

UK

Apr 2020

Three to five air samples were collected from multiple hospital environments using a Coriolis air sampler, viral RNA was measured by RT-PCR, Vero E6 and Caco2 cells cultures were used to culture virus.

38.7% (n=14/31) of the collected air samples were positive for viral RNA, but SARS-CoV-2 virus could not be cultured due to low recovered viral loads.

The odds of contamination in public areas was lower than areas immediately occupied by a COVID-19 patient (OR 0.5 95% CI 0.2-0.9).

Orenes-Piñero (2020)^{Footnote 80}

Biological monitoring study

Spain

Spring 2020 (est)

Investigators develop and apply “COVID traps” to measure the capacity of SARS-CoV-2 aerosol transmission in hospital patient care settings. “COVID traps” were placed 1 meter away from patients in ICU and ward settings. Viral RNA was detected by RT-PCR.

In the ICU, none of the “COVID traps” were positive for COVID-19; all COVID-19 patients were intubated. In the ward setting, two “COVID traps” were positive for SARS-CoV-2, both were near a patient requiring the use of respiratory assistance. The authors conclude it was unequivocally the result of virus transmission in air.

Feng (2020)^{Footnote 36}

Biological monitoring study

China

Feb-Mar 2020

Note: Additional results on viral RNA in exhaled breath samples are summarized in Table 4.

Environmental air from the rooms of recovering COVID-19 patients in isolation hospital wards and ICU were sampled using a NIOSH sampler. Air samples (n=12) were collected and aerosol size measured. Samplers were also placed on a tripod 1.2 m in height and 0.2 m away from the bed at the side of the patient’s head for 30 minutes.

SARS-CoV-2 RNA was detected in a single air sample from SARS-CoV-2 patients. The maximum viral RNA concentrations detected in the positive air sample by particle size was 1112 copies/m³ (<1 μm) and 745 copies/m³ (>4 μm).

The authors attribute minimal contamination of viral RNA in study samples to reduced respiratory viral shedding among patients in later stages of infection.

Ding (2020)^{Footnote 37}

Preprint

Biological monitoring study

Hong Kong

Feb 2020

Note: Additional results on viral RNA in exhaled breath samples are summarized in Table 4.

Air samples (n=46) were collected from airborne infection isolation rooms (AIIR) housing COVID-19 patients, nursing stations, corridor and air-conditioning units at a hospital treating COVID-19 cases. Multiple air samplers were used for sample collection, which was conducted on different days, and RNA was detected by RT-PCR.

A single air sample (n=1/46) from the corridor outside a storage room with a medical waste bin was weakly positive for SARS-CoV-2 RNA. All other tested air samples from patient rooms, washrooms, and air supply inlets were negative.

RNA copies for the weakly positive sample was not quantified.

Guo (2020)^{Footnote 81}

Biological monitoring study

China

Feb-Mar 2020

Air samples were collected from hospital ICU (n=40) and general wards housing (n=6) COVID-19 patients, at different distances from patients and the doctors office (n=8). Air samples were collected using a SASS 2300 Wetted Wall Cyclone Sampler.

SARS-CoV-2 virus particles were identified in 35% (n=14/40) of ICU air samples, 12.5% (n=2/16) of general ward air samples.

SARS-CoV-2 aerosol was detected in 35.7% (n=5/14) of samples near air outlets, 44.4% (n=8/18) of samples in patients’ rooms, and 12.5% (n=1/8) in the doctors’ office area. No SARS-CoV-2 virus were identified in patient corridor air samples.

Based on site(s) of positive air sample collection authors conclude virus-laden aerosols to concentrate near and downstream from patients, and the maximum transmission distance of virus laden aerosols to be 4 meters.

Nissen (2020)^{Footnote 82}

Biological monitoring study - Hospital setting

Sweden

Spring 2020 (est)

Open liquid containing petri dishes were placed at air entrances to ward rooms and near exhaust filters of a hospital’s ventilation system for 24 hrs to collect viable virus. Infectivity was assessed using Vero E6 cell culture.

SARS-CoV-2 RNA was detected in fluid samples placed in the ventilation system, and in 33% of samples (n=1/3) placed near air entrances of wards.

Viability of the isolated virus could not be established by cell culture.

Zhang (2020)^{Footnote 83}

Preprint

Biological monitoring study - Hospital setting

China

Mar-Apr 2020

The study sampled outdoor environment aerosols (n=16) at three hospitals receiving COVID-19 patients. Aerosol samples were collected using bioaerosol samplers WA-15. Viral RNA was quantified by RT-PCR.

SARS-CoV-2 virus was identified within sampled aerosols at 285-1,130 copies/m3 concentrations, similar to contamination levels observed in ICU units.

Viral RNA was identified up to 5 meters away from outpatient buildings (n=1/2), as well as in hospital waste water treatment areas (n=1/3).

Dumont-Leblond (2021)^{Footnote 22}

Biological monitoring study

Canada

Spring 2020

new

Air and no-touch surfaces of 31 rooms from multiple long-term care facilities (n=7) in Quebec were sampled. The air samples were collected 8 to 30 days after the residents’ symptom onset using an IOM Multidust sampler attached to a portable pump Gillian Air 5, placed ~2m from SARS-CoV-2 positive residents. The sampled rooms did not have significant aerosol mitigation measures in place. SARS-CoV-2 RNA was quantified by RT-PCR, and cultured by Vero E6 cells.

Note: the distance of surfaces from SARS-CoV-2 patients was not provided.

All air samples (n=31) were negative, but viral RNA was recovered from 32% (n= 20/62) no-touch surface samples of door frames and shelving units (concentrations range 13-36,612 genomes/surface).

Authors suggested the viral load recovered from the no touch surfaces without direct contact to be indicative of viruses spread through air.

Virus could not be cultured from positive samples.

Song (2020)^{Footnote 84}

Biological monitoring study

China

Feb 2020

new

Air samples were collected from Airborne infection isolation rooms (AIIR) housing COVID-19 cases. Air samples were collected using the automatic sampling system Derenda PNS 16T-3.1, at approximately 1 meter from patient beds. Viral RNA in samples was measured by RT-PCR.

All of the samples were negative for SARS-CoV-2 RNA.

Faridi (2020)^{Footnote 85}

Biological monitoring study

Iran

Mar 2020

new

Air samples (n=10) were collected from COVID-19 patient wards housing individuals with server and critical symptoms.

Air samples were collected into the sterile standard midget impingers using a vacuum pump. Air samplers placed 1.5 to 1.8 m above the floor and approximately 2 to 5 m away from patients' beds. Some patients coughed during the sample collection. Viral RNA in samples was measured by RT-PCR.

All of the samples were negative for SARS-CoV-2 RNA.

Declementi (2020)^{Footnote 86}

Biological monitoring study

Italy

Sept 2020 (est)

new

Air samples (n=8) were collected from a COVID-19 non-Intensive Care Unit, following cleaning and disinfection processes. Viral RNA in samples was measured by RT-PCR.

Note: The sampler model, the presence of COVID-19 cases during sampling nor the sampling distance from cases are provided.

All of the samples were negative for SARS-CoV-2 RNA.

Ahn (2020)^{Footnote 87}

Biological monitoring study

Korea

Mar 2020

new

Air samples were collected from the isolation rooms of severe COVID-19 patients requiring mechanical ventilation or high-flow oxygen therapy. Samples were collected using a SKC BioSampler and swab sampler.

Viral RNA was measured by RT-PCR.

All of the samples were negative for SARS-CoV-2 RNA.

Masoumbeigi (2020)^{Footnote 88}

Biological monitoring study

Iran

Sept 2020 (est)

new

Air sampling (n=31) was conducted in a hospital with COVID-19 patients; all of the patients had a severe form of cough and sneezing. Air samples were collected from Emergency wards, ICU, CT-SCAN, and laundry unit. Air sampling was performed by all-glass impinger (AGI) at ~0.5-4 m away from patients’ beds. Viral RNA was detected by RT-PCR.

All of the samples were negative for SARS-CoV-2 RNA.

Alsved (2020)^{Footnote 89}

Biological monitoring study

Sweden (est)

Spring 2020 (est)

SARS-CoV-2 RNA measured from COVID-19 cases (n=2) within 2 days of symptom onset. Air samples were collected 0.8 meters away from the case, as the individual was talking or singing. The measurements were carried out in an experimental airtight chamber with human volunteers.

Air samples collected within 0.8 meters of COVID-19 cases were negative for viral RNA.

Viral loads in subject airways at the time of the experiment could not be obtained. Authors state qPCR Ct values of 22–25 to have been reported in clinical reports for the subjects within 24hrs of the experiment.

Cheng (2020)^{Footnote 90}

Biological monitoring study

China

Jan-Apr 2020

Air samples were collected within 10 cm of asymptomatic and symptomatic COVID-19 patients (n=6) with and without surgical masks in AIIR were tested for SARS-CoV-2 contamination. Viral loads in respiratory patient fluid samples were also tested by having patients sneeze and spit into gelatin filters within air samplers. Viral loads were measured using assays (not specified) and RT-PCR.

No virus was detected in air samples from rooms with both surgical masked and non-masked patients.

Except for one patient who had a respiratory fluid viral load of 2.54 x 10⁴ copies/ml, all other patients’ samples from sneezing were negative for virus RNA.

Authors suggest aerosol transmission is not the predominate mode of infection transmission in the sampled settings. Appropriate PPE use, environmental disinfection, and single occupancy within AIIR are provided as reasons for observed results.

Kim (2020)^{Footnote 91}

Biological monitoring study

South Korea

Mar-Apr 2020

Air samples (n=52) were collected 2 meters away from COVID-19 patients (n=8), before admission, and on hospital days 3, 5, and 7 using a MD8 Airport Portable Air Sampler.

Some patients were housed in negative pressure rooms (e.g. AIIR).

RNA was measured by RT-PCR.

All collected air samples were negative for viral RNA.

Ong (2020)^{Footnote 92}

Biological monitoring study

Singapore

Jan-Feb 2020

Air samples were collected from COVID-19 patients (n=3) in AIIR at a dedicated SARS-CoV-2 outbreak center between day 4 and day 11 from symptom onset using SKC Universal pumps a Sartorius MD8 microbiological sampler. RNA was measured using RT-PCR.

No air samples were positive for SARS-CoV-2 viral RNA.

Ma (2020)^{Footnote 35}

Preprint

Biological monitoring study

China

Spring 2020 (est)

Note: Additional results on viral RNA in exhaled breath samples and community setting air samples are summarized in Table 4 and Table 6.

Air samples (n=26) were collected from hospital settings and unventilated quarantine hotel rooms of cases using a robot. RNA was detected by RT-PCR.

No air samples from hospital settings were positive for SARS-CoV-2 viral RNA.

(est) = Estimated based on author affiliations and publication date.

Lednicky (2021)^{Footnote 40}

Preprint

Biological monitoring study

US

Jan 2021 (est)

new

Air samples were collected from a car driven by a COVID-19 patient (mildly symptomatic, no cough) using a Sioutas personal cascade impactor sampler at multiple sampling stages of aerosol sizes (PCIS); the sampler was attached to the sun-visor on the passenger side of the car and approximately 3 feet from the subject’s face. The air conditioner in the car was on during the 15 minute long drive.

SARS-CoV-2 RNA was detected by RT-PCR, virus viability measured by Vero E6 cell culture, as well as genome sequencing.

SARS-CoV-2 viral RNA was detectable at all PCIS aerosol size sample stages, at concentrations ranging from 1.24E+03 to 3.14E+04 genome/ m³ of air.

These findings suggest virions are present in different sized respiratory secretions.

SARS-CoV-2 was cultured from the sampling stages collecting particles in the 0.25 to 0.50μm size range.

Munoz-Price (2021)^{Footnote 24}

Biological monitoring study

US

Spring 2021 (est)

new

Note: Additional results on viral RNA in environmental air from patient care settings are summarized in Table 5.

Air samples (n=9) collected from households occupied by confirmed COVID-19 cases. Air samples were collected using the Sartorius MD8 airscan sampler, positioned at 0.3 -1.8 meters from patients’ head. Viral RNA in samples was measured by RT-PCR.

55% (n=5/9) of air samples, collected from three different households, were positive for viral RNA. The odds of positive air samples in household settings, when compared to hospital settings, werewas estimated to be 8.75 [95% CI 1.21–63.43; P = .058]. The median number of days from the last positive SARS-CoV-2 test to the day of air sampling was 3 (range, 2.5–5).

Of the households with positive air samples, one had A/C running and all three had opened windows or doors immediately prior to air sampling. Anecdotally, most households felt warm and humid at the time of sample collection.

Gehrke (2021)^{Footnote 45}

Preprint

Biological monitoring study

Germany

Oct 2020 - Jan 2021

new

Note: Additional results on viral RNA in hospital settings air samples are summarized in Table 5.

Air samples were collected from a concert hall and a shopping mall for 6 hours during operation using non-powered cold traps. Viral RNA was quantified by RT-PCR in an area with high community transmission in an effort to identify potential hotspots for transmission.

Indoor COVID-19 hotspots were found in non-ventilated areas and in zones that are predisposed to a buoyancy (chimney) effect.

In the concert hall, 1/10 cold traps were positive for viral RNA.

In the shopping mall, 35% of traps were positive for viral RNA (n=5/14) located on the ground floor, 2^nd and 3^rd kiosk area as well as around the escalators. The 4^th floor fashion section that usually has 25% capacity was negative. The highest viral concentration (2.7-5.4 x10³ copies/mL) was found on the ground level between the escalators. However, follow-up measurements where cold traps were placed directly under 2 inflows and 2 outflows on the 3rd floor (to exclude any contamination of the central ventilation system) were negative.

Samples taken on Friday when a high volume of patrons were twice as high as those taken on Monday.

Moreno (2021)^{Footnote 52}

Biological monitoring study

Spain

May- Jul 2020

new

Air samples were collected from subway trains and buses during operation. Ambient air (PM2.5) was sampled using Teflon filters with PEM (Personal Environmental Monitor) equipment, A/C filters were also sampled. Viral RNA in samples was measured by RT-PCR.

2/6 subway train air samples were SARS-CoV-2 positive and none of the air conditioner filter samples were positive.

1/6 bus air samples and 4/9 air-conditioning filters were positive.

Cleaning and disinfection results indicated this process was successful at decontamination.

de Rooij (2021)^{Footnote 49}

Preprint

Biological monitoring study

Netherlands

Apr-Jun 2020

new

Air samples were collected from mink farms with ongoing COVID-19 outbreaks (3 farms at late stage of outbreak, 1 at early stage of outbreak). Samples were collected from indoor farm environments, outdoors at the farm premises, and nearby residential sites. Air samples were collected using teflon filters, and in parallel with particulate size fraction PM10 (<10μm) and inhalable dust (<100μm) collection. Personal air samples from the investigation field staff were also collected. Viral contamination in air samples was measured by RT-PCR.

Based on sample contamination the investigators conclude mink farms to be highly contaminated with viral RNA inside but the dispersion of virus outdoors and to nearby areas to be negligible.

Farm in acute stage of outbreak:

• SARS-CoV-2 RNA was detected in 62.5% (n=5/8) of the inhalable dust samples and 50% (n=4/8) of PM10 samples collected inside the farm. All (n=2) personal air samples were positive for viral RNA.
• SARS-CoV-2 RNA was also detected in outdoor samples at 1.5-10 meters from the farm’s open entrance, but samples collected at 20m from this entrance were negative.

Farms in late stage of outbreak:

• Viral RNA was detected in 9.8% (n=4/41) of inhalable dust samples, at ~4x103 copies/m3. SARS-CoV-2 RNA was not detected in the inhalable dust samples collected outside the mink houses (n=9) or from the PM10 samples collected inside the farm (n=9) or near the mink houses (n=9). One of the personal air samples was positive for SARS-CoV-2 RNA.
• All of air samples from residential sites were negative for SARS-CoV-2 RNA.

Ma (2020)^{Footnote 35}

Preprint

Biological monitoring study

China

Spring 2020 (est)

Note: Additional results on viral RNA in exhaled breath samples and hospital settings air samples are summarized in Table 4 and Table 5.

Air samples (n=26) were collected from hospital settings and unventilated quarantine hotel rooms of cases using a robot. RNA was detected by RT-PCR.

A single positive air sample (3.8%) was identified in an unventilated quarantine hotel toilet room.

Wong (2020)^{Footnote 50}

Preprint

Biological monitoring study

Singapore

Feb-Mar 2020

new

Air samples (n=6) were collected from accommodation rooms, corridors, toilets and elevators used by COVID-19 cases in a community settings (not specified). Viral RNA in samples was measured by RT-PCR.

All air samples were negative for viral RNA.

Döhal (2020)^{Footnote 51}

Preprint

Biological monitoring study

Germany

Mar 2020

new

Environmental air from 21 quarantine households (n=15) were collected by cyclone sampling using a Coriolis 154 Micro Air sampler. Viral RNA in samples was measured by RT-PCR and Vero-E6 cell culture.

Note: At least one person in each included household was positive for SARS-CoV-2 at the time of sample collection.

All air samples were negative for viral RNA.

Di Carlo (2020)^{Footnote 93}

Biological monitoring study

Italy

May 2020

new

Samples were collected on a city bus in a town which had a high number of COVID-19 cases, during the last week of lockdown and first week of gradual opening (approx. 1100 passengers used the bus during the study period). Air was sampled using microbiological gelatin membrane sample filters of 80 mm. Viral RNA in samples was measured by RT-PCR.

No air samples (or surface samples) were positive for SARS-CoV-2.

Public health measures in place included hand sanitized upon entry to the bus, mandatory masks and keeping the bus windows open. Authors state these precautions prevented SARS-CoV-2 circulation and detection in the investigation.

(est) = Estimated based on author affiliations and publication date.