Updated: Jul 18, 2020
What is driving this new viral surge? At the end of May, the virus seemed relatively under control in the USA, with most places seeing decreasing cases, and others seeing a steady state of cases or slight upticks, but no real surges. However June has brought this southern and western surge, starting in Phoenix, Houston, and Dallas, Florida and then South Carolina, other cities in Texas, Las Vegas, and now evident in Louisiana, Tennessee and other places. As this surge was emerging, I couldn't help but realize it was happening in places where it was all of a sudden too hot to go outside, and people would be spending most of their time indoors. This has made me wonder what this surge means for our understanding of viral spread, that there might be additional factors we should take into account as we adapt our mitigation strategies to most effectively prevent these types of surges.
Aerosol droplets: the overlooked mode of transmission? The CDC and WHO have consistently stated that the main route of transmission of SARS-CoV-2 is through respiratory droplets generated by the cough or sneeze of an infected person, suggesting that the primary mode of infection is from standing near someone with respiratory symptoms or coming in contact with a surface contaminated by viral particles and then touching one's mouth, nose, or eyes. This has driven our 6 ft social distancing and disinfection focused mitigation strategies. There have been some articles and discussion of the potential of the virus to spread from aerosol particles, which are much smaller than respiratory droplets and capable of traveling much further than 3 or 6 feet, but this has been largely downplayed by most experts. The CDC and WHO formal guidelines and discussion for transmission have not mentioned aerosol particles at all.
Respiratory vs. aerosol droplets. This May 2020 PNAS article makes the case that airborne transmission is the dominant route for spread of SARS-CoV-2 based on available epidemiological and experimental evidence, and also gives a nice description of how the virus is spread, including a definition of aerosol and respiratory particles. Infected individuals can shed viral particles from coughing, sneezing, or even normal talking and breathing. Some of these particles are shed as large respiratory droplets that will travel no more than 3-6 feet before falling to the ground, especially in more hot or humid environments. These have been thought to be the primary mode of spread, and motivated the social distancing guidance. Some other particles are shed in much smaller aerosol droplets that can remain in the air for an extended period of time and potentially travel great distances. A recent PNAS paper tracks the travel and virus-carrying potential of droplets exhaled while talking. The risk of these particles being one of the primary modes of transmission has been largely downplayed.
Minimum concentration or exposure needed for infection. Just because an individual is exposed to some viral particles does not mean they will necessarily become infected. There needs to be a certain volume of virus and that virus needs to be effectively transmitted to fertile ground in the respiratory tract for infection to occur. Erin Bromage, a biology and infectious disease professor at University of Massachusetts at Dartmouth, explained how these dynamics work in a blog post in May that went completely viral (no pun intended), with >13 million views in the first week and now >18 million views. He explained how volume of virus, which consists of viral concentration as well as time of exposure, is a crucial factor in determining spread. The skepticism of the primacy of aerosols as a route of SARS-CoV-2 spread is driven by an assumption that the particles are too small to contain a substantial volume of virus, and/or dissipate too quickly to accumulate enough to cause infection.
Accumulation of aerosols under the right (wrong) conditions. However, in crowded or enclosed indoor locations, especially with inadequate ventilation, the viral particles in aerosol droplets could in principle accumulate in sufficient numbers to pose a significant infection risk. One good analogy is cigarette smoke, which shares similar aerosol dynamics. Outdoors in the open air, someone could be smoking and I might not notice unless I am directly downwind and/or close to the person. Indoors, however, the smoke is a lot more noticeable. If I am close by, I notice it the most, but there is a chance I could smell it even if I am way across the room. If the place is enclosed (with low ceiling), crowded, and with poor ventilation, the smoke might encompass the whole room and cause people to choke and their eyes to burn. Thus, even if aerosolized viral particles contain a small number of viral molecules, in the right environment, such as an indoor enclosed setting with poor ventilation, it is possible for these viral particles to accumulate like cigarette smoke and expose those in the room to potentially infectious volumes of virus, especially if the infected person is in the room for a long time consistently exhaling aerosolized particles into the air.
Why didn't virus go away in the summer? I have wondered for the past several weeks whether this might be a major factor in this massive viral surge experienced in the southern and western US and now extending northward. Our PolicyLab COVID-19 tracking models have found temperature to have a clear bimodal effect, meaning that risk of transmission is higher in BOTH low and high temperatures, and lower in moderate temperatures (paper in press and under embargo; will update link in 2 weeks when published paper online). This may be one key feature of our model that has enabled it to successfully project these surges in Houston and other places in the south and west.
When I saw this bimodal effect I didn't believe it. I knew from the literature that respiratory viruses tended to go away in the summer -- that the hotter temperatures and humidity caused the respiratory droplets (i.e. the larger exhaled particles) to drop to the ground quicker so the virus couldn't spread, and that SARS-CoV-2, with its lipid exterior, is likely to not hold up in heat and UV light like other lipid-sheathed respiratory viruses (like influenza). I put some hope that the virus would subside in the summer because of these facts, and thought locations with high temperatures should be less at risk for an outbreak. Thus, the bimodal effect in the model was perplexing and counterintuitive to me. But the data were very clear -- integrating incidence data across hundreds of counties across the USA every day since the beginning of the pandemic, the model learned that a string of hot days would put county at increased risk of an uptick a week or two later. What could explain this effect?
Maybe the key causal factor underlying the model's temperature effect is not the biochemical properties of the virus when exposed to heat, humidity, and UV light, but the behavior of people when the conditions outside become very hot and humid. Living in Houston for 20 years and with my cousin moving to Phoenix a few years ago, I know full well that by June it is far too hot to spend much time outdoors, and at that time of year people travel out of state or start spending most of their time indoors in the air conditioning. Note that Houston and Phoenix were the first two cities to experience undeniable surges characterized by exponential growth, followed by Dallas, Miami, Orlando, and Las Vegas, which happen to be other notable cities in the country for which June really brings the heat. Is it possible that the temperature effect is telling us that the virus spreads more readily when people spend more time indoors, whether from cold or heat? There are papers suggesting that the virus spreads much more efficiently indoors than outdoors, e.g. this research from Japan showing the viral spread is 19x more likely indoors than open-air environments. Could it be aerosol spread is a major factor in what is going on?
Aerosol spread key factor in famous outbreak case studies: El Pais published an excellent article analyzing 3 SARS-CoV-2 outbreaks -- how they happened and how they could have been avoided. This online article uses clever animated graphics to illuminate the factors contributing to the outbreak and I highly recommend taking the time to read it. They described three outbreaks, one in an office, one in a restaurant, and one in a bus. The office was a high rise call center and 79/137 of the people in one room/wing of the floor were infected, only 5 people on the other side of the floor were infected, and virtually no one in the rest of the high rise was infected despite sharing elevators and common areas with them. In the restaurant, 10/20 people in a 3 table row were infected, but none others, and this restaurant had a poorly designed AC system that involved a gradient along those 3 tables that repeatedly recirculated air in that closed area. On the crowded bus with recirculating air, one infected person led to 23 people out of about 70 becoming infected, located all over the bus, with no spatial pattern relative to the infected person. In all of these cases, it appears aerosol transmission, not large respiratory droplets or contaminated surfaces, was the driving factor, and exacerbated by people spending long periods of time in crowded enclosed settings with inadequate ventilation, and also no masks. These provide only anecdotal evidence, but the factors leading to these outbreaks could be replicated in other settings.
Potential role of air conditioning: This raises the question of whether air conditioning systems could be playing a role in these surges as people move indoors. Air conditioning systems take warm air from the room, run it over evaporator coils that remove the heat from the air and transfer it outside, and the cooled air is then recirculated into the room. Could this recirculation of air lead to an accumulation of aerosolized viral particles and a potential outbreak? Any system relying on 100% recirculated air and poor ventilation and filtration would certainly be vulnerable, but proper ventilation and filtration in a well=designed HVAC system should ameliorate this problem.
In the past, buildings were built less airtight so a considerable amount of external air entered the buildings through walls and doors, providing adequate natural ventilation for even a completely recirculation-based A/C system. With improving construction and insulation, however, it has become necessary to add controlled ventilation to pull outside air into the system. The American Society for Heating, Refrigeration, and Air-conditioning Engineers (ASHRAE) has published standards for commercial (ASHRAE 62.1) and residential (ASHRAE 62.2) buildings to ensure adequate ventilation to prevent contaminants from accumulating in a home, and the commercial standards have been broadly adopted for new construction. However, these codes do not ensure that the specified minimum ventilation rates are maintained in practice. In most states, no such requirement exists, and ventilation rates are infrequently measured and difficult to measure accurately. Increasing recirculated air reduces energy costs, energy consumption, and CO2 emissions, so the only downside is potential inadequate ventilation and accumulation of contaminants. If not enough external air is drawn into an airtight system, it is possible for contaminants, including SARS-CoV-2, to accumulate. Any steps to ensure adequate ventilation and increasing inflow of external air into the system may help reduce these risks.
CO2 meters can be used to assess whether ventilation is adequate for the number of people in an enclosed space. The reasoning is that if exhaled CO2 is accumulating resulting in higher levels at a given location then it is also conceivable that exhaled aerosol particles containing SARS-CoV-2 from an infected person could also be accumulating at that location. Many modern, large scale HVAC air handling units already use these sensors to monitor ventilation and regulate the inflow of external air. In a room, multiple sensors may be necessary to average over spatial variability and get an accurate measure of CO2 levels in the room. Also, monitoring changes over time over time could be useful, as an increasing slope in CO2 measurements may indicate a building is becoming too crowded for its ventilation capabilities.
A/C systems also contain filtration systems that remove particles and contaminants from the air. Filters are given a "minimum efficiency reporting value" or MERV rating, with higher values indicating they filter out smaller particles, with HEPA filters having the highest values. Incorporation of filters with higher MERV may help reduce the risk of recirculating aerosolized particles in a room. HVAC companies are also assessing the use of germicidal ultraviolet light (UV-C) to further kill viruses and other contaminants in the AC system.
ASHRAE has published recommendations for safe operation of HVAC systems to minimize the potential for airborne transmission of SARS-CoV-2 that involves increasing fresh air intake, addressing filtration (including higher rated filters and UV-C where practical), and avoiding overcrowding. Many large employers with health and safety departments are already actively applying these strategies in their facilities, but we should all be aware of these key factors so we can ask the right questions where we work, study, worship, and relax.
Aerosol spread increasingly recognized as major factor: Last week, a group of 239 scientists published a letter in Clinical Infectious Diseases suggesting airborne and aerosol transmission should be recognized as a key mode of SARS-CoV-2 transmission. The letter was co-written by a WHO consultant and reviewed and signed by scientists from more than 30 countries, including Drexel department chair of environmental engineering Charles Haas. who along with assistant professor of engineering L. James Lo and mechanical engineer and lifelong friend Jonathan Thatcher provided me with feedback as I was preparing this post. This letter recommends improving ventilation, maximizing filtration, and controlling crowding at indoor, enclosed locations. In response, the WHO updated their March 29th brief on transmission to acknowledge the possibility of aerosol spread, although they still downplay it significance, highlighting the lack of rigorous scientific evidence and suggesting that apparent aerosol spread might be explainable by close contact with large respiratory droplets or contaminated surfaces, the major routes of infection they have been emphasizing. From my perspective, It can't understand why they are so hesitant to acknowledge what might be a major factor to take into account when constructing targeted mitigation strategies. Maybe they are afraid of the fear, panic, and loss of control people might feel if they consider that the virus might be airborne, but I think that would be unnecessary as I think more knowledge empowers us and enables policymakers to construct the most effective and least obtrusive measures of viral mitigation.
Take home message: What does this mean, and what should I do? Given airborne transmission, does this mean I need to live in fear that I can become infected anytime even when practicing social distancing and avoiding touching potentially contaminated surface? No. Remember that you would need to be exposed to a certain volume of virus to become infected so there is no need to fear the air we breath. But it means that we need to be aware of the settings in which aerosol particles can potentially accumulate to infectious levels: indoor, enclosed settings, with poor ventilation and filtration, with lots of people interacting without masks for long periods of time. This combination of factors is the recipe for some of the super spread events we have seen.
So what should we do? First, spend as much time outdoors as possible. It is important to stay socially connected to people, but it is much safer to gather outside than inside, so hold gatherings outdoors whenever possible. When indoors, avoid crowded, enclosed areas as much as you can, and most importantly limit the amount of time you spend in these settings. If in a store or restaurant, don't linger. Wear a mask whenever indoors and around other people or in an enclosed space where aerosol particles might conceivably accumulate. These guidelines are nothing new -- we've been told these things for months, but the acknowledgement of aerosol spread may provide extra clarification on which factors are primary and which are more secondary. I have seen news articles mention we should be opening our windows for increased ventilation. If you have very poor ventilation and filtration in your room, this might not be a bad idea, but most HVAC systems probably do a reasonable job, and the risk of accumulating virus may be greater in more crowded settings with more people.
What about airplanes? It may seem like they would be especially prone to outbreaks, yet have you heard of any super-spread events from flights? While social distancing and hygiene are still important factors to remember while flying, aerosol particles may not be a major issue given airplanes use HEPA filters and a mix of 70% external/30% recirculated air to keep the air fresh. Careful ventilation and filtration in other enclosed settings can similarly decrease the risk of aerosol spread.
Schools and other institutions should check for adequate ventilation in their A/C and heating systems and add improved filtration wherever possible. This may be a crucial consideration for safe reopening of schools, and the proper funding to assess these systems and make any necessary alterations and improvements need to be provided -- let's get on it, Congress!!
I also wonder if CO2 sensors could be a potential tool to monitor whether ventilation is sufficient for the amount of people in a given enclosed space, and monitoring these levels over time might be more effective than imposing fixed percent capacity guidelines that ignore the size of room or their ventilation and filtration systems.
We need to remember with this pandemic, we are in it for the long haul ... it is a long-term war, not a short-term battle. With this in mind, the essential task is to identify and practice key targeted mitigation strategies that limit viral spread and prevent surges yet allow us to live our lives to the fullest extent possible. Hopefully the acknowledgement of aerosol spread as a potential key factor can help us refine our mitigation strategies to be more effective.
Update: I just found out some a colleague of mine at Penn, Michael Kohansk, L. James Lo, and Michael Waring, who is an associate professor of architectural engineering at Drexel and one of the signatories of the letter to WHO on aerosol spread, just published a peer reviewed scientific paper entitled Review of Indoor Aerosol Generation, Transport and Control in the Context of COVID‐19 in the journal International Forum for Allegy and Rhinology. I highly recommend reading the article.