First up, the executive summary's context section in its entirety:

"The SARS-CoV-2 virus can spread in several ways: through zoonotic transmission, direct and indirect contact transmission, direct deposition transmission, and inhalation or airborne transmission. An increasing body of evidence [28]–[31] suggests that it is transmitted through infectious fluids released from an infected individual as particles of different sizes and quantities, such as during breathing, speaking, coughing and sneezing. While the largest particles travel downwards quite rapidly, the smaller ones remain suspended in the air for longer periods and can travel farther distances. When people are in close proximity, transmission of infectious particles can occur through direct inhalation (short-range) and deposition onto the mucous linings of the respiratory tract and ocular membranes of a susceptible host particularly in the absence of face covers and ventilation. ‘Long-range’ transmission can occur in enclosed settings when infectious particles accumulate over time in a given volume, where the concentration of virions is sufficient enough to cause infection once infectious particles are inhaled by a susceptible host."

Right off the bat we get to why I think this is so important and it's not JUST airborne transmission. A lot of time has been spent arguing methods of infection. This document acknowledges things like deposition onto the ocular membrane.

Why is this important? The science doesn't lie. Politicians and minimizers can ignore it, but, eventually they'll probably have to acknowledge it. In the meantime, you have another tool at your disposal to understand the science AND to help calculate your risk.

Would I rely on the percent given by the calculator as THE exact risk? Of course not. Do I think it's important that this document, and calculator exist? I do. I believe in science.

At this point I'd like to share a couple of the images. I'm not great at alt text, but I'm going to give it my best here. I am limited character-wise and was not able to simply copy-paste all of the text into the alt-text.

They go on to further define, and discuss, airborne transmission. I have to, again, point out that this is an important step for the WHO to acknowledge and discuss.

"Airborne or Inhalation transmission*: The process whereby aerosolized infectious respiratory particles (IRPs) are inhaled and enter the respiratory tract of a susceptible person, move through the upper and then lower parts of the respiratory tract, and can be deposited on the tissue at any point along the tract, potentially even reaching the distal alveolar region. This mode of transmission can occur when IRPs have travelled either a short or a long distance (range) after emission from an infected person or after resuspension of deposited particles from surface.

The Technical Advisory Group agreed to describe the inhalation transmission mechanism as a sequential five steps or components process with the short- and long- range transmission terms unfolding simultaneously (Figure 2)."

I'll put a brief description of each of the 5 components in the next post.

"The first component is the emission rate defined as the number of virus-laden particles exhaled by an infected person per unit of time."

"The second component affecting the inhalation transmission mechanism is the removal rate which can be defined as the total number of aerosolized virions removed from the air in a given time."

"The difference between the emission rate and the removal rate leads to the third component, the exposure. This component can be defined as the concentration of virions in ambient air, during a given time, at which a susceptible host comes in contact with."

"The fourth component is the cumulative or absorbed dose--meaning the total number of infectious particles inhaled and subsequently absorbed by a susceptible host during the exposure event."

"Finally, the pathogen infectious dose, the host immunological status and the specific SARS-CoV-2 variant transmissibility, contribute to the complex dose-response model which, in combination with the cumulative absorbed dose, define the host probability of infection which completes the fifth and final component."

I'll add their infographics of each of these components in the next post.

Here's the infographics on the first 4 of the components.

The fifth component and legend.

Next up is a pretty in-depth discussion of aerosolized particles. Salts, mucous and membranes! 😬

Honestly, my wife was a little grossed out by some of this, and she's in biology, so, if you're squeamish about it, might want to skip this one.

"Humans emit hundreds of aerosolized particles of different sizes during exhalation, and even more when speaking and singing [69]. Those aerosolized particles are water-based solutions of salts, containing mucus, proteins and may contain other material present on the surface of the respiratory tract including infectious pathogens. Thus, viruses on the surface of the respiratory tract can be released in the exhaled particles out of a person’s mouth and nose [70]–[72]."

"Particle aerosolization mainly results from an air- stream passing sufficiently quickly over the surface of a liquid to create separation. Several physiological phenomena contribute to aerosolization of respiratory fluid [73]. When exhaling, respiratory fluid blockages formed in the bronchioles burst and produce aerosolized particles released at the next exhalation – this is called bronchiole fluid film burst (BFFB)[74]. The vocal cord vibration during vocalization also aerosolizes the fluid by bathing the larynx, while the interactions between the tongue, teeth, palate, and lips aerosolize the saliva during speech articulation. Before being emitted, the particles undergo processes in the respiratory tract which change their size distribution. Furthermore, the respiratory tract contains relatively small-scale, curved viscous and viscoelastic films, which wrinkle during exhalation and thereby break up, leading to aerosol production. Emitted particles range in diameter from 0.01 and 1000 μm depending on the generation mechanism, respiratory and vocalization activity, age and site of origin [52], [74]–[76]. The size distribution is further affected by the quasi- instantaneous evaporation process particles have undergone after leaving the body. Particles of diameter smaller than 100 μm are likely to become airborne and remain suspended in the air from seconds to hours, because of their reduced size and settling velocity compared to larger ones [77]. The volumetric particle emission concentration can be estimated considering the specific anatomical processes originating in the aerosol, the bronchial region, the larynx, and the oral cavity. The aerosol concentration size distribution for speaking and coughing can therefore be modelled as a tri-modal lognormal distribution known as the Bronchiolar/Laryngeal/Oral (B.L.O.) tri-modal model [78]."

The next variable in the model here is viral load. It's hugely variable from person to person, and this is honestly a big variable in your chances of picking up COVID in a given situation. We've all heard of "super spreaders" and this is why.

"SARS-CoV-2 viral load

SARS-CoV-2 viral load detected by RT-PCR in nasopharyngeal swabs is widely distributed, ranging from 3 to 10 log10 copies/ml with a median of 6.78 log10 copies per ml [79]. The limited data on viral load in exhaled breath suggests high variability between infected individuals. While not every infector has detectable virus in exhaled breath, those who present detectable virus in breath range between 2-7 log10 copies for 15-60 minutes of expiratory activity [80]–[82]."

"The wide variation in viral load between individuals depends on factors such as age [85], [86], vaccination status [87], [88] and possibly exposure history, and variants of concern [89], [90]. Moreover, the highest mean viral load occurs quickly after symptom onset and at a higher magnitude in individuals with more severe COVID-19 symptoms [85], [91], [92]. The viral shedding episodes have a sharp upswing to reach the peak viral load, followed by a prolonged decay [85], [88]. The duration of viral shedding varies widely and correlates with viral load peak [93]."

Next up is a brief discussion of outward control, ie. if you're infectious and you're wearing a mask. I want to reiterate if I haven't made it clear here, these are not the entirety of each section. These are the parts that stood out to me as things I've wondered about, or seen other people asking about, or even just science I thought was cool. If you want to read more on any of this, it's in the document and references.

"Source control (outward)

Throughout the pandemic, the use of well fitting masks to cover the mouth and nose have been widely applied as a source control measure to reduce the emission of respiratory particles and thereby reduce the potential emission of infectious particles from infected people. Studies have supported that wearing a mask to protect others from potentially infectious particles is a highly effective infection control measure to limit the spread of COVID-19 [95], [96]. Well fitting masks provide a physical barrier to the emission of both large and small respiratory particles [97]–[99], and depending on the virus type, are also effective in reducing exhaled infectious particles [71], [84].

Both medical masks and respirators, even without fit-testing, can reduce the outward particle emission rates by 74% and 90% on average during speaking and coughing, respectively, compared to wearing no mask, corroborating their effectiveness at reducing outward emission [97]. In contrast, shedding of non-expiratory micron-scale particles from friable cellulosic fibres in homemade cotton-fabric masks may confound explicit determination of their efficacy at reducing respiratory particle emission [97]."

Next up is ventilation and air cleaning. In the COVID cautious community I think we discuss this a lot and this is simply talking about the variables and how they work. In the intro they discuss that in hospital rooms under airborne precautions the standard in 12 air changes per hour(ACH) and show, mathematically, how 3, 6 and 12ACH differ for probabilities of infection(page 4 for those following along).

"Ventilation rate and air cleaning devices

The purpose of ventilation in buildings is to provide healthy air for breathing by diluting pollutants originating in the building with clean air, and by providing an airflow rate to change this air at a given rate [40]. A well-designed, maintained, and operated ventilation system can reduce the risk of respiratory pathogens transmission, including SARS-CoV-2, in indoor spaces by diluting the concentration of potentially infectious particles through ventilation with outside air and filtration and disinfection of recirculated air. Natural ventilation can provide similar benefits. i.e., opening of windows and/or doors [9]. The removal rate due to ventilation and equivalent ventilation is obtained from the amount of air (m3 h-1) supplied to the space and the volume (m3) of the room [104].

For filters that are portable and self-contained, the
rate of particle removal from air passing through
the filter is expressed as the clean air delivery rate (CADR), which is approximately equal to the product of airflow rate and the contaminant removal efficiency [105]. For the purpose of this document, air cleaning and disinfection devices, using filter category MERV (minimum efficiency reporting value) 14 / ISO ePM1 70- 80%, high efficiency particulate air (HEPA) and higher filtration efficiency filter [9], as well as UVGI technology [106], are considered as equivalent ventilation and the Clean Air Delivery Rate (CADR) (m3/h) is added up to the ventilation rate."

Next is what they're calling the viable virus factor. I think we've all wondered how long is SARS-CoV-2 viable and in what scenarios, and there's a lot of variables, honestly. Here's what they based their model off of:

"Viable virus factor

Viable SARS-CoV-2 was identified in air samples from rooms occupied by COVID-19 patients in the absence of aerosol-generating health-care procedures [116] and in air samples from an infected person’s car [117].

...

While in laboratory experiments, SARS-CoV-2 stayed infectious in the air for up to 3 hours with a half-life of 1·1 hour [119], a systematic review looking at the range of ratio of viral copies in aerosol to plaque forming units (PFU) ratio returned values in between 6 to 0.343 log10 viral copies/litre of air and in between 2.15E+03 to 2.68E+04 TCID50/100 μm for viral titre [116], [120] and RNA to PFU [116] respectively."

I've certainly read numbers longer than 3 hours before, but don't have a reference handy. This might be one variable that I take issue with, but I'm not sure off hand.

Next up, masking yourself when not sick. I don't want to get too deep into the weeds on this. It's good that we're discussing the efficacy of masking and that it's recommended, AND that they discussed fit as being important. I've long wondered if we're ever going to get to a place where people not only mask, but care about the fit. Probably not on a large scale, but, this is about protecting yourself at this point. Masks work.

"Masks and Respirators (Inward)

The use of face coverings has been promoted as a
key measure to reduce the SARS-CoV-2 transmission throughout the pandemic. Specific types of face coverings including but not limited to respirators, surgical masks and cloth masks have been recommended for healthcare facilities [121], [122] and community settings [123] according to the level of risk and protection required. In the absence of mask mandates, the compliance of face coverings is an important parameter, hence the exact quantification of the risk reduction provided by the use of face coverings is highly dependent on the level of compliance. In addition, the mask composition and fit [124], [125] are also of high importance, where the standard requirements for filtration efficiency and fit are defined by NIOSH-42 CFR Part 84 [126], EN 149 [127] and ASATM F3502-21 [128]. In order to estimate the overall transmission risk, filtration efficiency values for this model have been extracted from experimental studies (Annex 1) that measured the inward and outward filtration efficiency of different types of masks for a given size of particles, with information on particles ranges, respiratory activity, particle velocity and airflow. Ultimately, however, the fit effect on the wearer’s face is probably the most significant parameter."

After masks, I'd like to add another graphic from the document, even though they had it a little earlier. This is about the "Size-dependent aerosol deposition mechanisms to sites in the respiratory tract." Want to keep them out of your respiratory tract? Masks.

A quick hit on the dose-response.

"Dose-response

Since the beginning of the pandemic, several animal studies looked at the dose-response relationships for symptoms, seroconversion and viral shedding through orotracheal [140], intranasal [141] and aerosol [142] exposure. Most recently, a human challenge study enabled the identification of the Median Tissue Culture Infectious Dose (TCID50) or the inoculum dose that induced infection in more than 50% of participants, corresponding to 10 TCID50 or 55 Focus-forming Unit (FFU) [143]."

What the heck is a focus-forming unit you say?

"A measurement of the concentration of live virus in a given amount of fluid. This is measured by spreading a known amount of the fluid over a layer of cultured cells which are infected by the virus, then counting the number of areas in the culture which look infected."

All of this leads us closer to where we're going, specifically, what the heck is my probability of infection and how did you decide to calculate it?

"Probability of infection

The interaction between the emission and the removal rates provides the infectious particles’ concentration, which, multiplied by the exposure time, enables estimation of the cumulative absorbed dose for a given scenario. The probability of infection is then appraised considering the host-pathogen interaction which includes the dose-response model, the specific SARS- CoV-2 variant considered, the host immunity and the sum of the short- and the long-range risk.

…

Multiple studies evaluated the percentage of asymptomatic SARS-CoV02 infections with results ranging from 1.4% to 78.3% [139]. For the above reason, the model aims to assess the risk of infection defined as the host probability to infection, proxy by seroconversion regardless of symptom onset."

Aha, asymptomatic infections. One of those variable that the COVID cautious drive ourselves crazy over. Just thought it was interesting. As I've long said, a big range, and also more common than you'd suspect.

But, hey, what about different variants? That's a whole thing. While I was doing this my wife was playing around with the calculator, and that's definitely one of the major limiting factors right now, but, let's see what they had to say about it.

"Variants of Concern

Of the six variants currently designated as VOI, five were considered in the analysis and among these only B.1.617.1 and B.1.525 demonstrated a statistically significant increase in the effective reproduction number of 48% (95% CI: 28–69) and 29% (95% CI: 23– 35), respectively.

Given the widespread co-circulation of VOC/VOI, the effective reproduction numbers of these variants were compared against each other in order to estimate the nature of future competitive growth between them. Notably, the pooled mean difference in the effective production number between the VOC B.1.1.7 and B.1.351 was small at 4% (95% CI: 0–8), while P.1 demonstrated an increase relative to B.1.1.7 and B.1.351 of 10% (95% CI: 3–17) and 17% (95% CI: 6–30). Given these estimates, the longer-term trends of competitive growth between these three VOC remain unclear. In contrast, the rapid observed growth of B.1.617.2 suggests a clear competitive advantage compared with B.1.1.7, B.1.351 and P.1, with estimated increases in the effective reproduction number of 55% (95% CI: 43–68), 60% (95% CI: 48–73) and 34% (95% CI: 26–43) respectively [145].

...

A systematic review shows that the effective reproduction number and basic reproduction number of the Omicron variant elicited 3.8- and 2.5-times higher transmissibility than the
Delta variant, respectively. The Omicron variant has an average basic reproduction number of 9.5 and a range from 5.5 to 24 (median 10 and interquartile range, IQR: 7.25, 11.88). The average effective reproduction number for Omicron is 3.4, with a range from 0.88 to 9.4 (median 2.8 and IQR: 2.03, 3.85) [147]."

Like I said, evolving science.

So, after all of that, you can put info into the calculator, and get a probability of infection if you have enough info.

Their example, in brief:

10 people in a 135m^3 room(to us USians, this would be about a 20x20 room with an 8' ceiling), no masks, 1 infectious person who speaks for 15 minutes, two separate times. Your chances of being infected in that room at ~15.6% and on average 1.4 people in the room will be infected.

If you know me, you know I love me some limitations in papers. There's A LOT in this. Evolving science. As far as I know the first attempt to put this all together. So many variables. Engineering meets biology. This is a lot. It's not exact. They have a whole lot listed, but I'll leave it at this:

Limitations

The aerobiology of infectious particles and the transmission dynamics to allow for a replication competent and infection competent virus to establish an invasive infection in humans is complex.

I was playing around with this calculator and so I put in a scenario that, pre-pandemic, was fairly common in my old workplace. We had ~12 people and there were times where we had to sit in a conference room from 8-5, with a lunch break for some training or certification.

The results were pretty interesting, in my opinion. I put in the size of the room, and that 1 person would be infected. As would be typical these days, I said zero masks.

It came up with a 45% chance that I would be infected were I in that room if I had short-range interactions with the infected person. 13% with long-range interactions.

You further get a graph showing Mean Concentration of Infectious Respiratory Particles(IRP) vs. Time of Day(aka, how long you've been in the room.)

The next graph it shows you is the Probability of Infection vs Viral Load. This is one of the biggest variables discussed above. While the estimate was 45%, interestingly based on how infectious the person who was COVID positive actually is, the 75th percentile is a 94% probability, for instance.

In this scenario at the end of the 8-5 work day you end up with an estimated 5 new cases, and 6 people who were exposed but not infected on average.

It then goes through the ways that you can reduce your risk, including wearing a well fitting mask, reduce interactions with people, and ventilating the space better.

Finally, it gives you the hierarchy of controls from most effective to least effective:

Elimination
Substitution
Engineering Controls
Administrative Controls
PPE

Giving you one last hint that maybe just you masking isn't the most effective thing, overall.

While not strictly the same document, the WHO this week released a related document bringing new clarity for a(hopefully) shared terminology of airborne infection going forward.

From the executive summary:

"Terminology used to describe the transmission of pathogens through the air varies across scientific disciplines, organizations and the general public. While this has been the case for decades, during the coronavirus disease (COVID-19) pandemic, the terms ‘airborne’, ‘airborne transmission’ and ‘aerosol transmission’ were used in different ways by stakeholders in different scientific disciplines, which may have contributed to misleading information and confusion about how pathogens are transmitted in human populations.

...

The scope of what type of pathogens were covered in this consultation and the resulting
descriptors used in this document are as follows:
• Pathogens, contained within a particle (known as ‘infectious particles’), that travel through the air, when these infectious particles are carried by expired airflow (they are known as ‘infectious respiratory particles’ or IRPs), and which enter the human
respiratory tract (or are deposited on the mucosa of the mouth, nose or eye of another person) and;

• Pathogens from any source (including human, animal, environment), that cause
predominantly respiratory infections (e.g., Tuberculosis [TB], influenza, severe acute
respiratory syndrome [SARS], Middle East respiratory syndrome [MERS]), but as
well as those causing infections involving the respiratory and other organ systems (e.g. COVID-19, measles).

The following descriptors and stages have been defined by this extensively discussed consultation to characterize the transmission of pathogens through the air (under typical circumstances):

• Individuals infected with a pathogen, during the infectious stage of the disease (the source), can generate particles containing the pathogen, along with water and respiratory secretions. Such particles are herein described as potentially ‘infectious
particles’.

• These potentially infectious particles are carried by expired airflow, exit the infec-
tious person’s mouth/nose through breathing, talking, singing, spitting, coughing or
sneezing and enter the surrounding air. From this point, these particles are known as ‘infectious respiratory particles’ or IRPs.

• IRPs exist in a wide range of sizes (from sub-microns to millimetres in diameter).
The emitted IRPs are exhaled as a puff cloud (travelling first independently from air
currents and then dispersed and diluted further by background air movement in the room).

• IRPs exist on a continuous spectrum of sizes, and no single cut off points should be
applied to distinguish smaller from larger particles, this allows to move away from
the dichotomy of previous terms known as ‘aerosols’ (generally smaller particles) and ‘droplets’ (generally larger particles).

• Many environmental factors influence the way IRPs travel through air, such as ambi-
ent air temperature, velocity, humidity, sunlight (ultraviolet radiation), airflow distri-
bution within a space, and many other factors, and whether they retain viability and infectivity upon reaching other individuals.

...

The descriptor ‘transmission through the air’ can be used to describe the mode of trans-
mission of IRPs through the air.
Under the umbrella of the ‘through the air’, two descriptors can be used:
•
‘Airborne transmission/inhalation’: Occurs when IRPs expelled into the air as described above and enter, through inhalation, the respiratory tract of another person and may potentially cause infection. This form of transmission can occur
when the IRPs have travelled either short or long distances from the infectious person. The portal of entry of an IRP with respiratory tract tissue during airborne transmission can theoretically occur at any point along the human respiratory tract, but preferred sites of entry may be pathogen specific. It should be noted that the dis-
tance travelled depends on multiple factors including particle size, mode of expul-
sion and environmental conditions (such as airflow, humidity, temperature, setting, ventilation).

•‘Direct deposition’: Occurs when IRPs expelled into the air following a short-range
semi-ballistic trajectory, then directly deposited on the exposed facial mucosal sur-
faces (mouth, nose or eyes) of another person, thus, enter the human respiratory tract
via these portals and potentially cause infection."

https://iris.who.int/bitstream/handle/10665/376496/9789240089181-eng.pdf