Observation and quantification of respiratory droplets by laser light scattering

National Institute Of Diabetes And Digestive And Kidney Diseases

Investigators

Linked publications, trials & patents

Abstract

Using video recordings of speech droplets that crossed through a planar sheet of high intensity laser light we were able to demonstrate that, at ca 1000 droplets per second of speaking, the number of particles generated by speaking is far higher than previously reported in the literature. The very large variation in intensity of the scattered light indicates that the size distribution of these particles is very heterogeneous. The composition of speech-generated respiratory droplets is mostly oral fluid, which has been documented to be rich in virus, in particular during the most infectious, early stage of infection (https://pubmed.ncbi.nlm.nih.gov/33767405/). Considering that the probability that a droplet contains a virion scales with its volume, and therefore with the cube of its diameter,large droplets carry far more virus than small droplets. However, this consideration is offset by the fact that large droplets fall to the ground much faster than small droplets that are known to rapidly dry out, turning into aerosol that can remain airborne for minutes or longer. Moreover, small droplet nuclei (< 5 micron) can reach the lower respiratory tract (LRT) and it is widely understood that infections that initiate in the LRT are likely to lead to increased disease severity. Our efforts therefore have focused mostly on this small to intermediate size range of droplet nuclei, large enough to have sufficient probability to contain a virion, but small enough to reach the LRT. We demonstrated that the number of speech droplets generated does not vary dramatically between different speakers, but different sounds lead to very different numbers and size distributions. Whereas vowels are generated primarily at the vocal folds and are invariably quite small (0.5-2 micron diameter in the dehydrated state) and modest in quantity (less than a few hundred per second), consistent with prior literature, sounds generated at the front of the oral cavity, including letters such as 'p', 'b', 'f', and 't' can be rich in particles that cover a wide size range, spanning from a few microns to more than 100 micron in diameter, with a single p generating between 100-1000 of such droplets. Dehydration kinetics of such droplets are considerably slower than for pure water droplets, and preliminary evidence points to the formation of lint-shaped aerosols upon dehydration, which can prolong the airborne lifetime relative to the collapse into a spherical aerosol. Our measurements showed that the use of cotton facemasks strongly elevates the humidity of inspired air, which will aid in preventing dehydration of the respiratory tract and thereby improve the functioning of the mucociliary clearance pathway. This observation therefore could explain the surprising literature reports that the use of cloth facemasks leads to lower disease severity, even though their protection against inhalation of small particles that can reach the lungs is known to be poor. Consistent with this conclusion is the statistical analysis of disease incidence and death in the USA as a function of salinity and humidity of the air, which shows a decreased death rate with increased salinity, while taking into account a wide range of possibly confounding factors. This decreased death rate, reflecting decreased disease severity, is attributed to the known positive effect of inhaled salt on MCC. In a separate study we evaluated to the importance of self-infection of the lungs through speech droplets, after an initial infection of the upper respiratory tract has taken place. A speaker is invariably at the center of their own speech aerosol cloud and there is no doubt that self-infection of the lungs can occur. However, it was not known what the importance of this not previously recognized pathway is relative to the widely accepted aspiration pathway. We explored the importance of self-infection through speech by surveying members of the Deaf community, engaged in various levels of vocalized speech, after becoming infected with SARS-CoV-2. Although the survey was small (102 participants) results indicate that disease severity correlates positively with the amount of vocalization and negatively with the use of masks, both consistent with significant self-infection through vocalized speech. Our findings provide a straightforward path to reduce the incidence of severe disease, after a known exposure, by minimizing vocalized speech and the use of facemasks.