Probable aerosol transmission of SARS‐CoV‐2 in a poorly ventilated courtroom

Abstract There is increasing evidence of SARS‐CoV‐2 transmission via aerosol; the number of cases of transmission via this route reported in the literature remains however limited. This study examines a case of clustering that occurred in a courtroom, in which 5 of the 10 participants were tested positive within days of the hearing. Ventilation loss rates and dispersion of fine aerosols were measured through CO2 injections and lactose aerosol generation. Emission rate and influencing parameters were then computed using a well‐mixed dispersion model. The emission rate from the index case was estimated at 130 quanta h−1 (interquartile (97–155 quanta h−1). Measured lactose concentrations in the room were found relatively homogenous (n = 8, mean 336 µg m−3, SD = 39 µg m−3). Air renewal was found to play an important role for event durations greater than 0.5 h and loss rate below 2–3 h−1. The estimated emission rate suggests a high viral load in the index case and/or a high SARS‐CoV‐2 infection coefficient. High probabilities of infection in similar indoor situations are related to unfavorable conditions of ventilation, emission rate, and event durations. Source emission control appears essential to reduce aerosolized infection in events lasting longer than 0.5 h.


| INTRODUC TI ON
The role of aerosols in the transmission of COVID is currently under debate. Superspreading events, in which aerosols may have been a significant factor, are regularly reported in the press. However, the cases reported in the scientific literature remain rare and we were able to identify few studies that explicitly attribute the emission to aerosols (see Table 1). In most of these cases, the evidence of aerosol contamination is also indirect, because contamination via surfaces or large droplets alone is not sufficient to explain the attack and reproduction rates observed. 1 A similar observation was made after the systematic analysis of 318 outbreaks in China, where transmission in confined and poorly ventilated spaces was attributed to aerosols. 2 The term aerosol generally refers to exhaled respiratory particles <5-10 µm that are both likely to remain suspended in the air for a long time and to penetrate deeply into the respiratory system. 3 Human activities such as breathing and speech produce exhaled respiratory particles mostly below the 5-10 μm range. 4 Zhu reports that a healthy individual emits 10-10 4 particles per liter of exhaled air, 95% of which are <1 μm aerosols. 5 During speech, the emission can reach 5 × 10 3 particles per minute. Coughing generates 10 3 -10 4 particles ranging between 0.5 and 30 μm, a majority of them below 2 μm. A sneeze | 1777 VERNEZ Et al. produces about 10 6 particles between 0.5 and 16 μm. 6,7 The emissivity increases with the energy supplied (breathing <speech < cough), with a strong variability between test subjects. 7 The sedimentation time of aerosols can reach several hours and their aeraulic behavior is close to that of a gas. Their displacement is essentially determined by the movement of the ambient air. 8 However, the largest droplets, typically greater than 50-100 μm, will follow ballistic trajectories due to the force of gravity and their initial kinetic energy (e.g., in case of coughing). The sedimentation mechanism will dominate and the droplets will settle on the ground after a few seconds, or ten seconds. 8 Between 10 and 50 μm, the droplets will have an intermediate behavior, which will depend on their intrinsic properties and their environment, especially due to the evaporation mechanism. For droplets of saliva, which contain non-volatile material, the evaporation mechanism is slower. A decrease in size during 3-4 min, followed by a stabilization at about half the initial droplet diameter, was observed experimentally with saliva droplets. 9 Simulations of the propagation of respiratory and aerosol droplets generated by speech over a wide range of temperatures (0-40°C) and relative humidity (0-92%) show that 95% of the droplets are deposited over a distance of less than 1.4 m. 10 However, some droplets may travel greater distances depending on the ambient conditions or initial situation, such as initial ejection speed and entrainment by the turbulent gas cloud when sneezing. The simulation of these phenomena shows that the distance traveled by droplets is less than 1 m for an air movement of 1 m s −1 (normal speech), whereas it can pass to 6 m, in the case of an air movement of 50 m s −1 . 11 These simulations are consistent with experimental tests, which show that droplets and aerosols produced by coughing and sneezing can reach distances of 7-8 m. 12 In a simulated classroom, CFD computation shows that a significant fraction (24%-50%) of particles smaller than 15 µm follow the airflow and are extracted within 15 min by the air conditioning system, while particles larger than 20 µm tend to settle on the floor, offices, and neighboring surfaces. 13 The theoretical transport capacity of a spherical droplet depends on its volume, which is proportional to the cube of its radius.
It is therefore possible to estimate, for a given viral load scenario, the probability of finding the virus in a droplet or aerosol immediately after emission. Analysis of oral fluid from infected patients shows viral RNA loads ranging from 10 6 to 10 9 copies ml −1 with a maximum of up to 10 11 copies ml −1 . 14,15 As reported by Stadnytskyi, for a viral load 7 × 10 6 , there is a 37% probability that a droplet of 50 μm (before dehydration) contains at least one virus. 16

Practical implications
• This case study contributes to the growing body evidence highlighting possible SARS-CoV-2 transmission through aerosol.
• The known exposure conditions in the courtroom allow us to reasonably exclude other transmission modes.
• Probabilities of transmission in room of similar sizes and various event durations and air renewal were estimated.

Location
Situation Reference  [17][18][19][20][21][22] However, other areas have also been investigated and airborne contaminations have been found in public places such as department store entrances 18 or in urban areas. 23 Most of these studies were conducted by PCR however, which is not a demonstra- In this study, the hypothesis of aerosol contamination in a poorly ventilated space is addressed through the case of a SARS-CoV-2 cluster occurring in a courtroom.

| Case description
In October 2020, Unisanté's occupational physicians were called upon by a Vaud state department following the appearance of symp- During the hearing, the door and window were always kept closed for confidentiality reasons. But the window was opened during the breaks. Based on weather data, the average daily relative humidity (outdoors) on the day of the hearing was 93.5%. The indoor relative humidity at the time was not known.
The contact tracing carried out afterward by the public health service established that: • P1 can be considered as the index case and that P2, P3, P4, and possibly P5 are likely secondary cases following the hearing.
• Except for P5, who had a contact with a positive case prior to the hearing, contact tracing did not identify other risk situations for secondary cases (bar, parties, occupational or private exposure).
In particular, P2, P3, and P4 had no contact with P1 outside of the hearing and no contact with other COVID cases during the window of contagiousness. can reasonably be considered secondary cases to P1. The situation is less conclusive for P5, which endured a very short incubation period and reported having been in contact with a positive case prior to the hearing.
Although the number of participants and contaminations related to this event is quite modest, the interest of this case lies in the precise temporal and situational information obtained on the hearing.
This information was supplemented with the contact tracing data for the index case. The analysis of these elements suggests that the aerosol transmission route may have played a determining role in this situation.
• During the hearing, the persons present remained seated in their assigned seats, there was no sharing of seats, and only a paper document was circulated between 3 persons, therefore transmission of the virus by contact with a contaminated surface or object does not seem credible to us. Moreover, the instructions for cleaning the surfaces were respected a priori. The work surfaces of the room's furniture can therefore be considered as clean before the hearing.
• Droplet transmission alone seems unconvincing because the index case spoke very little during the hearing and does not seem to have spoken loudly during breaks. Subsequent measurements in the room showed that the spacing between the chairs of the index case and P4, P2, P3, and P5 were 1.5, 1.5, 3, and 3.3 meters, respectively.
• The hearing lasted 3 hours in a room, which, apart from short breaks, had closed windows and no mechanical ventilation. In the absence of sufficient ventilation, the aerosols generated by the breathing and speech of the participants will concentrate in the room and contribute to increase the dose of quanta received.
The index case sample (nasal swab) was no longer available at the time of this evaluation. However, it was possible to retrospectively analyze the sample from a secondary case, which tested negative for S-gene dropout. It can be concluded that the cluster is not due to the UK variant.

| Field Measurements
Field measurements were conducted on December 21, 2020, in the absence of any hearing in the courtroom. Compared to the studied situation, the only difference during this field measurement was the presence of plexiglas™ dividers (65 × 110 cm) between each seat.
The air renewal was measured by injecting CO 2 from a pres-

| Modeling viral transmission through aerosols
The (1) F I G U R E 2 Symptom and test schedule for index case and all four probable secondary cases The integration of Equation (4) over the exposure time D allows expressing the average concentration in the compartment assuming a constant emission, no initial concentration in the room, and a negligible external concentration.
Using Equations (2) and (5), the emission rate can be estimated using the probability of infection p and the environmental exposure conditions in the courtroom.

| Implementation
The calculations have been implemented on Stata/IC 16.1 (StataCorp LLC, TX, USA). Table 2  Volumetric inhalation rates were taken from the Binazzi study.
The value used here is that of "reading aloud with a normal voice", which is lower than that of singing. Doremalen's study was selected to determine a plausible range of loss rate due to virus inactivation.
The half-life interval proposed by this study is 0.64-2.64 h, corresponding to a k of 0.26-1.08 h −1 .

| RE SULTS
The probable distribution of the resulting emission rate is shown in The results of the sampling carried out are presented in Table 3. Even with the presence of plexiglas™ (added after the hearing), which tends to limit horizontal airflows, these concentrations remain in the same order of magnitude, suggesting that the well-mixed model used is adequate.
Examination of the direct reading measurements made at locations P1 and P3 shows differences in the aerosol size distribution (see Supplementary material). At P1, the generated lactose aerosol presented a mass-based size distribution with a geometric mean diameter of 0.63 µm (GSD 1.72 µm), in agreement with the mean size of exhaled aerosols (0.7-1 µm), 29 while at P3, the mean diameter   Table 3.

TA B L E 2
The effect of the loss rate related to ventilation (λ v ) is shown in which are surprisingly few in number and highlight situations of confinement, insufficient ventilation, and/or high emissions rates. 27,31,32 Improving air renewal conditions in meeting spaces reduces the risk of transmission in general. However, our results suggest that an effective strategy to combat large infectious clusters would be to target "hot spots", where groups of individuals spend several hours in poorly ventilated spaces, as a priority.
The example given here is that of a 150 m 3 room and the results obtained are not readily transposable to larger or smaller volumes such as conference rooms or domestic rooms. It is interesting to note that 150 m 3 is typically in the order of magnitude of the volume of classrooms, the ventilation of which is a matter of debate. Our results suggest that while room ventilation is essential, it is difficult to control the risk of contamination with this parameter alone because of the residual probability of infection at high ventilation rates, brought by the variability of the other parameters (e.g., duration of exposure and emission rate). Keeping in mind that the number of transmitters and targets also increases linearly with the number of people in the room. Wearing a community or surgical mask, which contributes to mitigate emission at source, especially during voice bursts, coughing, or sneezing, seems essential.

| AUTHOR CONTRIBUTOR S
DV and SS planned the study. JJS, DV, and GS planned the field investigations. SS and CP investigated the case study JJS, and GS conducted field investigations. JJS, GS, and DV conducted the data analysis, DV did the modeling and wrote the first draft. All co-authors contributed to the revision and finalization of the manuscript. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted.

FU N D I N G I N FO R M ATI O N
The author received no specific funding for this work.

ACK N OWLED G EM ENTS
The authors would like to thank several people for their support

CO N FLI C T O F I NTE R E S T
No conflict of interest.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/ina.12866.

DATA AVA I L A B I L I T Y S TAT E M E N T
Detailed primary and secondary data used for this study are available upon request.