Panel discussion with Moderator | Donald Milton | William Ristenpart
Robert T. Schooley | Bill Lindsley | Linsey Marr
Panelists (as pictured): Moderator, Dr. Donald Milton, Dr. William Ristenpart, Dr. Robert T. Schooley, Dr. Bill Lindsley, and Dr. Linsey Marr (biographies at bottom)
Given the large percentage of asymptomatic transmissions, and the much larger amounts of aerosols that are generated merely from talking or breathing, what do we know so far about the relative ratio of droplet vs aerosol transmission for SARS-CoV-2? What do we know about short range transmission?
Lindsley: It is still an open question for SARS-CoV-2. The majority of the work on other viruses has been done on influenza. This work has shown that aerosols have higher viral load than droplets, and it is possible to detect influenza in the air throughout healthcare facilities.
Given your work looking at face shields vs masks, do you have relative numbers on aerosol vs droplet production with influenza?
Lindsley: It is difficult to distinguish aerosol from droplet transmission at close range since when you are very close to someone you have a high aerosol concentration that drops off as you move farther away and the same is true for droplets. Don Milton's work showed the viral load difference that indicates the aerosol component is more important. You'd expect ventilation to have a larger impact on aerosols than droplets. Face shields are good blocking droplets, but not aerosols so it would be worth doing a study to evaluate if aerosol transmission occurs with people wearing face shields.
What do we know about where in the respiratory tract particles containing SARS-CoV-2 originate? We know for other viruses, but what about this one?
Schooley: It can be found in the nasopharyngeal secretions and the diagnostic focus in this area is likely related to the ease of accessibility. We also know there is more virus in sputum. There are ACE2 receptors in olfactory epithilium and that is how people lose taste and smell. But by far the largest amounts of viruses are found deep in the lungs where fine aerosols are generated. This virus has done a great job since upon infection the first thing it does is shut off the immune response, which allows people to go around talking and spreading viral-laden aerosols.
Bourouiba: Contrary to other pathogens, we have evidence that the virus is present in both the upper and lower respiratory tract. Given the uniform distribution, there is not a reason to think there would be a segregation in viral load based on the location from which aerosols/droplets originate. It is important to take into account where we measure since evaporation will concentrate the viral load and could lead to the impression that this is due to the inherent particle size. Having a more uniform measurement approach will reduce the risk that these variables become entangled.
How do the different activities lead to different amounts of aerosols (e.g. talking or breathing)?
Ristenpart: It is clear that the louder you talk, the more particles come out. Different sounds/syllables can change aerosol production by a factor of 2, but volume can change the levels of aerosols produced by 1-2 orders of magnitude. A possible intervention would be tell people to be more quiet. A 6db volume reduction halve the particle production.
Schooley: This is the reason for concerns about outbreaks in bars. People are loud in bars so we might want to design quieter bars where people do not have to shout at each other at close distances.
Ristenpart: There was an outbreak in a Washington Karaoke bar where there were plastic partitions in front of the singers, but this had no effect.
Milton: There was also an outbreak in a Zumba class and deeper breathing also leads to more particle generation. When you fully exhale, then inhale, you close and re-open pathways that form bubbles, which then break. That bubble leads to a sub-micron drop that is then exhaled.
Are there certain activities that will lead to larger amounts of viral emissions? Are there large differences between asymptomatic and symptomatic generation and what causes this?
Milton: From exhaled breath aerosols, 20% of the population produces 80% of the exhaled aerosol particles. We saw this with influenza. Some people emitted tens of millions of viral copies and others emitted practically nothing depsite having comparable levels of the virus in the nose. Preliminary data from my lab indicates we will see the same thing with COVID. It's not clear what drives this. Higher BMI is one of the factors that leads to higher shedding.
Bourouiba: The overall viral load was lower in the presymptomatic cases. However, once you have symptoms, the diminished lung capacity may reduce the ability to exhale forcefully so the force that is necessary for fragmentation might not be present.
Schooley: The lungs do not get sampled. In comparison, progression in HIV can be observed by directly sampling the blood. The nose is only a spillover from the lungs.
How much do we know about differences in mucus composition and how that affects production?
Bourouibia: The mucus composition affects fluid rheology, which affects the response to shearing, break-up, and piercing of bubbles that impacts particle size. Data on influenza shows that these properties vary a lot between patients and between presymptomatic and symptomatic patients. In addition to the size, the embedding of viruses into the particles will vary.
Viruses can cause infection when re-aerosolized from paper tissues as you showed. What do we know about how important this pathway might be for SARS-CoV-2?
Ristenpart: Just rubbing a tissue will generate thousands of viable articles. Guinea pigs can transfer to other guinea pigs via fur contamination. If you sneeze into your elbow, then you could see re-aerosolization. We don't know how much of a role this plays in SARS-CoV-2, but we need to start thinking about more and its role in transmission between humans.
You showed re-suspension from viruses on floors. Do we have any idea for influenza about the fraction of particles in a room that have been re-suspended?
Marr: We only looked at the amount that was re-suspended, but this should be an easy comparison to make so we will add it to our list of research topics.
People are surprised how far particles can travel indoors, even when the size is well above 5 microns. How far apart do you feel comfortable to remove your mask indoors?
Marr: I never remove my mask indoors in less in my own home or office when the building is fairly unoccupied. The 6 ft idea is to avoid the respiratory plume, but does nothing for the background aerosols. The correct guidlines should be wearing a mask indoors at all times when there other people present, not just when it is not possible to maintain social distance.
Milton: The karaoke outbreak with plexiglass shields shows that using these approaches in schools in front of the teacher or between kids is not going to be be very successful.
Ristenpart: The way to think about this for communicating to the public might be to consider, "is the shield going to block cigarrette smoke?", and the answer would be "no". The aerosol particles behave very much the same way.
Bourouiba: Even with the masks, you have leakage so you should still try to maintain social distance since the aerosol particles will accumulate.
Milton: This is where the next layer of ventilation comes in. Once it has leaked out, it's still in the room and we need to remove it from the room.
You mentioned thermal heating from the body causes these aerosols to rise upward.
Milton: We all have warm air rising from our body.
A lot of the data we have is on indoor settings. What do we know about outdoor distance travel of exhaled virus.
Milton: If you're at an outdoor restaurant with someone smoking and you're downwind of them, you will smell it, and that tells you something.
[our comment: olfactory senses are incredibly sensitive, able to detect absolutely miniscule concentrations, which would pose negligible risk from an infectivity perspective -- this comparison is flawed, and it is necessary to determine actual viral levels in outdoor environments, which may be sufficiently low to pose very low risk.]
Bourouiba: Within short distances, even in an open space there can be high concentration "pockets". The air flow determines the extent to which these pockets will be broken.
Milton: Exposure time matters. e.g. if you pass by someone for a few seconds or if you are sitting at an outdoor restaurant downwind of someone for an extended amount of time, these are completely different things.
In a narrow, unventilated hallway of an apartment building with 10 apartments per floor, if people wear masks in the hallway, will aerosols accumulate? Will aerosols accumulate without masks? Same question with regards to an elevator.
Marr: Depends on ventilation rates. It could build up, but the time people spend in the hallway or elevator could be low (both for emission and exposure). The air change rate in an elevator is higher than we think given the frequent opening and closing of the doors. The aerosol risk there is much lower than from a restaurant or a bar.
Milton: Looking back at the Amoy Gardens outbreak, the movement of air between apartments was what determined transmission.
Ristenpart: The call center outbreak in Korea also supports this.
N95 masks are the first line of defense against aerosols due to their excellent diffusional capture and electrophoresis capture, but they are in short supply. How do surgical masks compare for SARS-CoV-2?
Ristenpart: The surgical masks block 74-90% of the physical particle emissions at the micron scale. We also measure the leakage and cellulosic particle emissions that occur from the cotton masks that make it harder to isolate the particles that are directly from the exhalation.
Milton: Source control is different from personal protection. Ther surgical masks cannot do a great job at protecting the wearer from ambient aerosols. This is particular risky for healthcare workers when exposed to high concentration aerosols. From a public health perspective where concentrations are lower, if you just reduce emission by two thirds at the source, and then two thirds is blocked on the intake, then you are down to a ninth of the viral load. With ventilation this goes even further.
Schooley: In looking at models for what happens as a greater percentage of people wear masks some researchers suggested linear outcomes, but do you expect some synergistic/bi-directional effects that lead to non-linear improvements? It seems we might be low-balling the benefits of universal masking.
Marr: I'd agree with that since universal masking works as both source control and PPE. Even with only 50% efficacy of masks, everything is cut by a factor of 4 so it should be non-linear.
This can also lower the severity when people do get infected.
If you could build the perfect aerosol sampling machine/method to figure out where the virus is, what size particles is it in, what produces it, how does it spread in different environments, what would that look like?
Lindsley: PCR is used in many studies due to high sensitivity, but it doesn't measure infectivity, which we would want. In a perfect world, we would be able to measure infectivity without having to culture. Particle size is probably the most important factor for the behaviour in the air. The amount of the virus varies a lot between different places in the room so it would be good to see this distribution instead of just total levels, along with the ability to observe that over time after an event such as a cough.
Milton: It's difficult to be sampling and get enough material to analyze within different size fractions, while simultaneously observing changes over time. We would like to know things such as how many viruses are present in a droplet within each size fraction. It could be that a virus needs to hit a cell several times to initiate infection. A big question is how much infectious virus is being released into the air and we should be able to determine that with current instruments.
A paper from the University of Bristol indicated singing is no different than speaking, which is contrary to some of these talks, so would anyone like to comment?
Ristenpart: It depends on the definition of singing. Modulating volume changes the epiglottal resistance... I'm not sure what the technical distinction would be. Talking has more breaks.
Milton: Could say it the other way around, which is that talking is as bad as singing.
Marr: They found singing at the same volume as talking produced not significantly different levels of aerosols, but I think we might sing louder and for longer than we talk. In singing, everyone might sing in parallel, which is less common when talking.
Ristenpart: Our modeling shows that duration of vocalization matters e.g. in a continuous vocalization from a choir, the transmission probability goes up dramatically.
Bourouiba: Most studies on particle size amounts and distribution over time are not normalized based on the duration of the event (e.g. comparison between talking, coughing, or singing) so it is difficult to make accurate comparisons.
There are many techniques that have been developed over decades for measuring and sampling aerosols in the aerosol science community. Would these approaches work for sampling viruses as well or other there other specialized requirements such as to preserve infectivity?
Lindsley: Most samples collect dry, on a filter or on a tube, and viruses generally don't like being dry and will die in a sampler. What you want is to collect the sample in a liquid (e.g. the condenser Dr. Milton uses).
Does anyone know about transmission of SARS-CoV-2 in dental offices or aerosols that are generated in dental processes?
Bourouiba: There is not to my knowledge SARS-CoV-2 specific data, but there is a large body of work quantifying emissions of aerosols from various dental procedures. It was possible to culture bacteria from the mouth on opposite sides of the room. There is a lot of work that needs to be done for dental offices to re-open in a safe way.
According to Chen et al. 2020, why does the CDC still say droplet transmission dominates when there is no specific evidence for it and it has never been demonstrated directly for any respiratory disease in the history of medicine?
Milton: Part of it goes back to the analogues to Koch's postulates that were proposed. It is hard to culture from the air and in fact no one has ever cultures measles from the air and it's generally impossible to culture tuberculosis from the air of a patient room. It seems the medical community may be waiting for these postulates to be addressed before buying into it.
Schooley: Hamsters have been shown to transmit between cages without droplets. Outgoing surgical masks were shown to be more effective than for intake. There are a lot of things we would like to do in humans, but would hopefully never do in humans. Does it really matter which is more important between aerosols and droplets? A lot depends on context. We need to take into account both from a public health perspective in planning interventions. It is not wise for the CDC to say it is primarily droplets merely for the purpose of re-assuring the public. The public wants dichotomous, definite answers, and when you don't give them that they say "see, you don't know what you're talking about."
Ristenpart: The hamster study did not actually prove direct aerosols since it could have been aerosolized fomites.
Schooley: That's true.
Given everything we know so far, is there sufficient evidence that we recommend universal masking, and upgrade and ensure we have better ventilation in buildings?
[all panelists nodding, many thumbs ups, "absolutely"s]
Any other comments for last few minutes of time left?
Bourouiba: In the prior SARS outbreaks there was also some dispute about the transmission mechanisms, but at some point, in the heart of the pandemic, the precautionary principle must come into play and we need to act even if we do not have 100% certainty on all the details.
See the key FAQ pages provided by a team of scientists on COVID-19 transmission and preventive approaches:
- 1. General questions about COVID-19 transmission
- 2. General questions about aerosol transmission
- 3. Protecting ourselves from aerosol transmission
- 8. Ventilation
- 9. Filtering, and “air cleaning”
Bill G. Lindsley is currently a research biomedical engineer at the National Institute for Occupational Safety and Health (NIOSH), which is part of the Centers for Disease Control and Prevention (CDC). Dr. Lindsley studies the role of infectious airborne particles (aerosols) in the transmission of diseases and the efficacy of measures to protect health care workers from these aerosols. He designed the NIOSH two-stage cyclone aerosol sampler which has been used to conduct aerosol sampling for a wide range of applications, including collecting airborne influenza virus and SARS-CoV-2 virus. His group has collected respiratory aerosols from influenza patients and examined the amount of virus contained in these samples. Dr. Lindsley also designed the NIOSH cough aerosol simulator, which is being used to study the efficacy of face masks and face shields as protective devices for the wearer and as source control devices to reduce the expulsion of cough-generated aerosols into the environment. Dr. Lindsley received his BS in Mechanical Engineering from the University of Maryland, College Park and his PhD in Bioengineering from the University of California, San Diego.
William Ristenpart is Professor of Chemical Engineering at the University of California Davis. His research focuses on complex transport phenomena in a variety of applications, including electrocoalescence of charged droplets, shear-induced deformation of red blood cells, and extraction dynamics of coffee grounds. A recipient of the NSF CAREER award for work on charged droplets, since 2014 Prof. Ristenpart has been supported by the NIH to investigate the transport of pathogens through the air, with recent results revealing the relationship between vocalization loudness and expiratory particle emission, and that influenza virus is transmissible via “aerosolized fomites.” His group is currently investigating facemask efficacy as well as the potential role of dust in transmission of COVID-19. He received his Ph.D. from Princeton University and did his post-doctoral work at Harvard University.
Robert T. Schooley is an infectious disease specialist and an expert in RNA virus infections and treatment. He currently serves as a distinguished professor in the division of infectious diseases at UC San Diego School of Medicine, where he has developed a drug discovery program for HIV, HCV and coronaviruses. As a professor of medicine, he leads the Universidade Eduardo Mondlane-UC San Diego Medical Education Partnership Initiative and supervises postdoctoral fellows. Infectious disease specialists care for patients with infections or diseases caused by viruses, bacteria, fungi and parasites. These include hepatitis viruses, tuberculosis, influenza, and HIV/AIDS, in addition to infections of the sinuses, heart, brain, lungs, gastrointestinal system, urinary tract, pelvic organs and bones. His research interests include HIV, influenza, coronaviruses, global health and international medicine, and the diagnosis and management of infections that cause death and morbidity in resource-limited settings. Dr. Schooley is particularly interested in the origin and development (pathogenesis) of HIV and HIV therapy, and was one of the first researchers to describe the humoral and cellular immune responses to HIV infection. Prior to joining UC San Diego, Dr. Schooley was head of the Division of Infectious Diseases at University of Colorado and director of the Colorado Center for AIDS Research. During his tenure at Colorado, Dr. Schooley was chair of the National Institute of Allergy and Infectious Diseases’ AIDS Clinical Trials Group. Before that, he served as associate professor of medicine at Harvard Medical School. Dr. Schooley is extensively published, having edited numerous books and authored hundreds of articles and book chapters. He serves on the editorial board of several medical journals and currently serves as Editor-in-Chief of Clinical Infectious Diseases. He is a fellow of the Infectious Disease Society of America and Royal Society of Medicine (UK), and member of numerous professional societies, including the American Society for Clinical Investigation, the Association of American Physicians, and the American Society of Tropical Medicine and Hygiene. In 2013, Dr. Schooley was honored with the Best Doctors in America and America’s Top Doctors award. Dr. Schooley received his medical degree from Johns Hopkins School of Medicine.
Donald K. Milton is a Professor of Environmental and Occupational Health at the University of Maryland School of Public Health. He is a Diplomat of the American Boards of Internal and Preventive (Occupational) Medicine and a Fellow of the International Society for Indoor Air Quality and Climate. He has served on the editorial boards of Applied Environmental Microbiology, Indoor Air, and BMC Public Health, on the NIOSH NORA Indoor Environment Team, and chaired the ACGIH Bioaerosols Committee. Dr. Milton’s research focuses on infectious disease aerobiology and exhaled breath. His work lead to the recognition that influenza patients can shed infectious virus into aerosols without coughing and that surgical mask worn by infected cases can reduce the aerosol release of influenza and seasonal coronaviruses. Dr. Milton earned a B.S. in Chemistry from the University of Maryland Baltimore County, an M.D. from Johns Hopkins, and a Dr.P.H. (Environmental Health) from Harvard School of Public Health.
Linsey Marr is the Charles P. Lunsford Professor of Civil and Environmental Engineering at Virginia Tech. Her research group applies interdisciplinary approaches to study pollutants in indoor and outdoor air. She is especially interested in emerging or non-traditional aerosols such as engineered nanomaterials and viral pathogens. Her research on the airborne transmission of infectious disease has focused on influenza, Ebola virus disease, and Legionnaire’s disease. Dr. Marr is a recipient of an NSF CAREER award and an NIH New Innovator award. In 2018, she was named a Fellow of the International Society of Indoor Air Quality and Climate. She is an Associate Editor of Microbiome and also serves on the editorial advisory boards of Aerosol Science & Technology, Environmental Science: Processes & Impacts, and Environmental Science & Technology Letters. She is a member of the National Academies’ Board on Environmental Science and Toxicology and recently served on the committee on Grand Challenges in Environmental Engineering for the 21st Century. Dr. Marr received a B.S. in Engineering Science from Harvard College and a Ph.D. in Civil and Environmental Engineering from the University of California at Berkeley and completed her post-doctoral training in Earth, Atmospheric, and Planetary Sciences at the Massachusetts Institute of Technology.
Lydia Bourouiba is an Associate Professor at the Massachusetts Institute of Technology, where she founded and directs the Fluid Dynamics of Disease Transmission Laboratory. Her research leverages advanced fluid dynamics experiments at various scales, biophysics, applied mathematics to elucidate interfacial flow and fluid fragmentation processes driving mixing, transport, and persistence of particles and microorganisms driving multiscale epidemiology and disease transmission. Dr. Bourouiba founded the Fluids and Health Conference, to be expanded into a Gordon Research Conference that she will chair in 2022, creating an international forum for exchange on frontier research and challenges in health, where fluid dynamic concepts are at the core, including infectious diseases disease transmission and related policy. Dr. Bourouiba is the recipient of many awards, including the Tse Cheuk Ng Tai’s Prize for Innovative Research in Health Sciences, the Ole Madsen Mentoring Award, and the Smith Family Foundation Odyssey Award for high-risk/high-reward basic science research. Dr. Bourouiba received her PhD from McGill University.