Quantitative assessment of the risk of airborne transmission of SARS-CoV-2 infection: Prospective and retrospective applications

Abstract

Graphical abstract

Superspreading Events Without Superspreaders: Using High Attack Rate Events to Estimate Nº for Airborne Transmission of COVID-19

“…One of the key differences between our work and prior work is our conclusion that so-called "superspreading" events can be adequately explained by index individuals with normal viral loads, while prior works analyzing some of the cases considered here have estimated the viral loads of index 22 Given this small exposure rate and the ubiquity of people touching their eyes, nose, or mouth, one immediately wonders if fomite transmission of sub-single-quantum doses of virions through touching of contaminated surfaces and subsequent inadvertent inoculation of the mucosal membranes by face touching might play a role in protection from infectious disease through natural variolation. A similar effect has been suggested to occur if a low volume of virions are transmitted between masked individuals when social distancing is being practiced.…”

“…[10,33] ; (b) range based on exhaled breath condensate (EBC) dilution measurements [34][35][36] ; Talking : (h) value used (2.7 × 10 -2 mL/hr); (i) range in Miller et al based on Morawska et al [10,20] ; (k) range in Stadnytskyi et al [33] ; Removal rate λ (1/hr) (k), (l) value used for cases in buildings and buses, respectively (2.12 h -1 and 3. Independent of the estimate of N 0 , and as Miller, Jimenez, Buannano, and Bazant have separately done, the case studies can be used to compute the number of aerosolized units of N 0 (quanta) released per unit time, which can in turn be used to estimate the quanta released by talking or breathing [20][21][22]28] . This calculation finds that talking releases ~460 N 0 per hour of aerosolized virus, whereas normal breathing releases ~10 N 0 per hour.…”

“…For purposes of this analysis, we therefore use a value of ρ = 10 7 copies/mL, which approximates the average viral density in the window of ~1 day around the peak. This is also the value that was used in Monte Carlo modeling of the individual infection risk for susceptible subjects confined in a room with an asymptomatic index patient [22] . As a side note, this value is also not far from the mean density measured by Wölfel et al , who found a mean viral density in sputum of 7.00 × 10 6 viral copies per mL (6.85 Log 10 copies/mL) [19] .…”

“…13 Note that in the Skagit choir case, the calculated quanta emission is of 2,347/h is roughly double the mean emission rate of 970 quanta/h reported by Miller et al [20] ; this difference is largely due to a mean assumed breathing rate in the Miller calculation of ~1.0 m 3 /h (range 0.65-1.38), compared to 0.5 m 3 /h here. based on the work of Buonnano et al , recommends a range of ~2-14 quanta/hour for oral breathing (depending on the activity) and 61-408 quanta/h for loud speaking [22,28] .…”

“…An example set of data and model fit (patient S14) from their work are shown in Figure 3(a) below: the model shows a very rapid increase of ~5 orders of magnitude over a few days to a peak viral load of order 10 7 copies/sample, a sharp drop of ~2 orders of magnitude over the next two days, and a somewhat slower decline 15 Using Figure 2 of Fears et al , the genome/PFU ratio is estimated to be ~20, ~1040, ~1140, ~189, and ~132 after 10, 30, 120, 240, and 960 minutes, respectively. [45] Buonanno et al appear use the "steady state" figure of 1.3 x 10 2 for c PFU (copies per PFU) in their analysis [22] . 16 The value of ~500 genome copies/PFU is in the range of that measured for SARS-CoV-1, where Vicenzi et al measured 360 genome copies/PFU [63] .…”