On March 10, 2020, the day before the World Health Organization declared COVID-19 a pandemic, 61 members of the Skagit Valley Chorale gathered for choir practice at a Presbyterian church in Mount Vernon, Washington. The choir members took what they thought were appropriate precautions: according to the Los Angeles Times, they refrained from hugs and handshakes, and many used hand sanitizer. Sitting in chairs spaced six to ten inches apart, they sang a few numbers, then split into separate groups in different parts of the church before reassembling. The practice lasted two and a half hours.

Apparently unknown to the rest of the choir, one member attended the March 10 practice while experiencing cold-like symptoms. Days later, this individual tested positive for SARS-CoV-2. Within three weeks, 52 other choir members who were present on March 10 had become ill. Ultimately three were hospitalized, and two died.

The Skagit Valley Chorale practice was one of the first COVID-19 “superspreader” events to be widely reported. The details supported what many infectious disease researchers and OEHS professionals had suspected: that SARS-CoV-2 is transmitted primarily through aerosolized particles. But it would be many months before this fact was incorporated into WHO and CDC guidance for limiting viral transmission.

By April, a few weeks after most businesses and schools in the U.S. shut down, Brian Schimmoller was working on a project intended to help employees at his company, Signature Science, a scientific consulting firm based in Austin, Texas, return to work safely. This was the beginning of the COVID-19 Exposure Assessment Tool, a free, editable PDF that OEHS professionals can use to determine the relative risk of group projects in the age of COVID-19. Users of the CEAT—pronounced “seat”—enter information such as the community’s transmission rate, the event’s duration, the room’s dimensions, the ventilation rate, the type of filtration, the group’s vaccination rate, the types of masks group members will wear, and their anticipated breathing rate. The CEAT uses these and other variables to calculate a single value that represents the risk of transmission relative to a documented high-risk exposure scenario.

Recently, The Synergist spoke with Schimmoller to learn how the CEAT’s many layers came together. We also sought perspectives on the tool from Ken Martinez, CIH, chief science officer of the Integrated Bioscience and Built Environment Consortium (IBEC), which is working with AIHA to promote the CEAT’s use, and from Ben Heck, CIH, CSP, CAC, who has used the CEAT extensively at NASA’s Ames Research Center.

FIRST PRINCIPLES Schimmoller studied meteorology at Cornell and served four years as a weather officer in the U.S. Air Force. Later, he attained a master’s degree in environmental engineering, with a focus on air pollution, from the University of Texas. He then went to work for Radian Corp., contributing to various projects that incorporated the measurement and modeling of air pollution, both indoor and outdoor. In 2001 he helped found Signature Science, where he is now president and CEO.

Signature Science offers physical, life, forensic, and data science-related technical solutions and technologies. Its employees specialize in bioinformatics, genomics sequencing, CBRNE (chemical, biological, radiological, nuclear, and high-yield explosives) detection systems, and laboratory quality assurance, and collaborate to develop technical solutions to address national security and law enforcement challenges.

“Since our employees work in labs, they needed to come in to work at some point,” Schimmoller said, recalling the lockdowns and his initial work on what would eventually become the CEAT. “So, I was thinking about, ‘How can we look at this in a risk-based way using a first-principles approach?’”

Schimmoller’s original intent was to create a nomogram—a type of graphical calculator that presents the relationships among several variables. Nomograms were once prevalent in engineering manuals and reference books. By drawing lines between values portrayed on a nomogram, users could determine, for example, the correct sizing of a blower or pump, or estimate air pollution concentrations downwind of a stack. One advantage of a nomogram is that it would have been usable by anyone, even someone without a computer or smart phone.

“We went down the nomogram path for a while and eventually realized that this is just going to be too complex,” Schimmoller said, referring to his partner in this initiative, Molly Isbell, director of Quality Assurance and Data Science at Signature Science. “You just can’t have that many variables on a single chart.”

With a nomogram ruled out, Schimmoller focused on creating a PDF that uses JavaScript for calculations. A PDF was preferable to a web application because it offers greater security: users can store PDFs on their own devices, mitigating concerns that their data would fall into the wrong hands.

The PDF format also had important design implications. “I have a tendency to want to build dashboard user interfaces, where the user has everything in front of them,” Schimmoller said. “With a dashboard, you can see more directly the way things affect the result—you know, ‘If I dial this down, this is what happens.’”

Schimmoller began with the premise that the tool would piece together all the information needed to calculate whether an individual would receive an infectious dose of SARS-CoV-2. But, initially, there was a problem: no one knew what an infectious dose was. So he focused instead on calculating relative risk against a baseline exposure scenario of concern that was defined by CDC and OSHA in April 2020.

The baseline scenario used in CEAT was meant to represent a scenario that would have been considered high risk according to the initial guidance from CDC and OSHA. This scenario considered an individual exposed to an infectious person for 15 minutes in a 9-foot by 9-foot room with a ceiling height of 9 feet. The two individuals are approximately 3 feet apart, and neither is masked; the uninfected individual is breathing at a rate of 13 liters per minute, a value that the EPA Exposure Factors Handbook assumes for activity of light intensity; and the ventilation rate in the room is 6 air changes per hour (ACH), which is CDC’s recommendation for healthcare facilities. Other assumptions in the baseline scenario include the aerosol settling rate and the virus degradation rate. (A full explanation of the baseline scenario and the CEAT model appears in a preprint article and supplementary material published by medRxiv. An updated, peer-reviewed version of the article was published in September in the journal Science Advances.)

With the baseline scenario defined, all other scenarios could now be compared to this scenario. Eventually, when infectious dose research was published in the literature, Schimmoller was able to incorporate that information as well, and the mechanics of the tool started to take shape. Users would enter information about their planned event, and the PDF would calculate the event’s risk relative to the baseline scenario. If their planned scenario yielded a result of “high risk,” users could also plan for a safer event by changing the variables to see how each affected the result. For example, moving the event to a larger room or a facility with better ventilation would yield a more acceptable relative risk. So would limiting the number of people or requiring more effective masks.

OUT OF THE BOXES The dominant model of the airborne spread of infectious diseases indoors is the Wells-Riley model, which was developed by Richard Riley of Johns Hopkins in the 1970s to describe an outbreak of measles in an elementary school in upstate New York. To describe a transmission event, a user enters the number of people who contracted disease during the event, their inhalation rate, the number of initially infected people at the event, and the room ventilation rate. With these inputs, the Wells-Riley model can calculate the quanta per hour, which is an estimate of the amount of viral emissions from the initially infected people. Once these values are estimated from one or more known transmission events, these quanta per hour can be used to calculate the infection rates for other future theoretical events.

A key simplifying feature of the Wells-Riley model is its assumption that the air in a room is always well mixed—that the viral concentration is equal in all parts of a room. But this assumption doesn’t account for proximity to a viral source. Closer to the source, there is the potential for greater concentration of viral aerosols—and greater risks of transmission. At sufficient distance, however, the well-mixed assumption is typically a valid approximation.

The CEAT attempts to address the complexity of proximity, building off the near-field/far-field model as described by Mark Nicas in chapter 6 of the AIHA publication Mathematical Models for Estimating Occupational Exposure to Chemicals. Intended to describe chemical emissions, this model envisions a room as comprising two “boxes”: the near field, which contains the emission source, and the far field, which is the rest of the room. The model allows users to predict concentration rates both for individuals who are close to the source and those further away.

Schimmoller relied on the near-field/far-field model as a starting point but had to make some adjustments to account for the additional complexities of groups of people and viral transmission. “That [near-field/far-field] scenario assumes that you have a single source and a single receptor,” he said, but infectious disease transmission occurs in groups of people that may have multiple sources of infection and multiple receptors. Schimmoller also found it mathematically convenient to arrange these sources and receptors equidistantly within a space, but the rectangular boxes assumed by the near-field/far-field model wouldn’t allow it: in a rectangular area, people on the diagonal are further apart from each other than they are from the people next to them, in front of them, and behind them.

The solution, essentially, was to reimagine the shape of the near-field box. “It really doesn’t matter what shape that box is,” Schimmoller explained—it could be a cone, cylinder, hemisphere, triangle, hexagon, or any other shape. “We came up with a way of doing it on a triangular grid, where the grid allowed us to add some unique geometric advantages for the computations. You could use those triangles to build little near-field volumes within the larger [far-field] box that represents the room.” This innovation “allowed us to generalize the equations so we could quickly do the math and change the sizes of the clusters of people and all the different volumes so we can evaluate multiple sources and multiple receptors,” Schimmoller said.

By November 2020, other COVID-19 risk assessment tools started to appear, but they assumed a well-mixed room. Schimmoller, Martinez, and Heck all found them to be too simplistic to be relied upon for bringing employees back to work during the pandemic.

“A big part of what I think is unique about the CEAT model is its ability to evaluate an exposure based on proximity of the people to each other,” Schimmoller said. “And that’s one of the things that we found was lacking from similar models.”

Another major piece of the CEAT model was generating an estimate of viral concentration. Taking a cue from Nicas’ work, which characterizes wind speed in a closed room, Schimmoller’s team reviewed studies of the effects of ventilation rates on the mixing of airborne contaminants in indoor environments and applied an “eddy diffusivity” technique that accounted for the turbulence produced by ventilation.

With the concentration mechanism established, completing the CEAT was a matter of adding variables such as mask use, room size, filtration, number of people, and so on. To test the CEAT’s accuracy, the team plugged in information from 20 documented cases of COVID-19 transmission, including the Skagit Valley Chorale practice. The CEAT’s predicted infection rate for these cases strongly correlated with actual infections, and Signature Science released a beta version in December 2020.

INCREASING REACH Martinez’s professional background includes many years of experience in biologics and infectious agents. With NIOSH, he participated in several emergency responses, including the 2004 SARS outbreak in Toronto. He was familiar with Signature Science through the Department of Homeland Security’s BioWatch program, a counter-terrorism effort that focuses on countering biological agents employed as weapons of mass destruction, and he provided feedback to Schimmoller on early versions of the CEAT. Asked to describe his reaction the first time he saw it, Martinez said, “I was very excited. It includes so many variables that are critical to a risk assessment tool, [but] by presenting it as a PDF to the user, it looks very simple. You can just tweak things here and there.”

At that time, IBEC was a brand-new nonprofit dedicated to helping mitigate transmission of SARS-CoV-2 and educating the public about safe protocols during the pandemic. IBEC and AIHA would partner on several educational efforts, and AIHA supported IBEC’s delivery of a series of virtual summits on topics related to the pandemic.

Then, in 2021, AIHA received a grant from CDC/NIOSH that funded efforts to educate businesses and healthcare facilities on how to better prepare their indoor environment against airborne diseases. AIHA sought IBEC’s help to identify ways to achieve this goal, and one of Martinez’s recommendations was for the two organizations to publicize the CEAT.

The Signature Science team was receptive to this idea. They had developed the CEAT without any outside funding, for their own use, and largely on their own time, but they believed in its potential to help other businesses and appreciated the opportunity to increase its reach. IBEC, AIHA, and Signature Science subsequently signed an agreement allowing AIHA to use the CEAT as one of its deliverables to satisfy requirements of the CDC/NIOSH grant.

APPLYING THE CEAT By then, the CEAT had already been in use for several months at NASA’s Ames Research Center in Mountain View, California, where Ben Heck is the senior industrial hygienist for contract staff. Heck supervises a team of four other industrial hygienists who are responsible for approximately 4,000 workers.

Heck—who, like Schimmoller, is a coauthor of the medRxiv preprint paper—became aware of the CEAT through his participation in the COVID-19 International Research Team (COV-IRT), a consortium formed early in the pandemic to inform understanding of the disease. Staff from Signature Science also participated in COV-IRT.

In the medRxiv paper and in a presentation at AIHce EXP 2022, Heck described how Ames started using the beta version of CEAT in December 2020. Initially, Ames OEHS staff used the CEAT to determine the maximum number of people who could be allowed into a location while maintaining what Ames leadership defined as acceptable risk. The CEAT’s ability to isolate the contributions of individual variables to the reduction of relative risk also helped Ames prioritize distribution of PPE by allowing staff to determine which projects would benefit most from N95 filtering facepiece respirators. After California’s lockdown ended in June 2021, Ames started using the CEAT to determine occupancy loads for meeting rooms.

“I’d say that’s like ninety percent of what we’re using it for now,” Heck explained. “When people are reserving [conference rooms], they’re usually told they have to determine occupancy size using this tool.”

The Ames employee pool is unusual compared to most workplaces: many Ames personnel have PhDs, and some are literal rocket scientists. With their science background, these employees have shown a degree of enthusiasm for the CEAT that workforces at other employers might not. Still, not everyone at Ames is knowledgeable enough about ventilation rates, for example, to fill out the CEAT entirely on their own, so Heck started a Microsoft Teams group through which he shares updated versions of the tool with some variables already entered.

Ames also encourages supervisors and meeting hosts to be trained on how to use the CEAT. Earlier in the pandemic, Heck conducted real-time training through Teams on how to fill out each step of the tool. He has since made a video of himself delivering the training that new staff can watch.

Most work at Ames is funded by the project, and staff need to be on site to perform it; you can’t build a NASA rocket in your kitchen. In the early days of the pandemic, when the OEHS staff were determining which workers could come to the Ames campus, they were effectively deciding which projects could move ahead.

“It was really stressful on us in the beginning,” Heck recalled. “We were trying to make it clear that we were not trying to [say], ‘This person is going to have a job, and this person isn’t going to have a job,’ or that they’re going to miss their launch date for a rocket because of us.”

The CEAT helped by giving OEHS staff an objective measure by which to make these determinations. “We’re able to say, ‘Here’s what we’re using as our standard, you have access to it, you can do it yourself,’” Heck explained. “I think that increased our reliability when a lot of people were really scared that they were going to lose funding for their projects.”

FUTURE PROMISE The CEAT holds promise not only for helping employers navigate what remains of the COVID-19 pandemic but also for responding to future pandemics. Schimmoller explained that, with some tweaks to the CEAT model, the tool could be used to predict relative risk for other infectious diseases such as tuberculosis, measles, and pandemic influenza. He also speculated that the CEAT model could be repurposed for hazards other than bioaerosols. “Looking at the way we did the near-field/far-field might be useful for other people doing near-field/far-field concentration estimates of other hazards and chemicals,” he said.

But for the foreseeable future, the CEAT should get plenty of use from employers trying to keep their staff safe from COVID-19.

“People want it to be over,” Martinez says of the pandemic. “Many believe that it is over. But COVID-19 is still taking a hundred thousand people per year in deaths.

“The bottom line is that we believe risk assessment is the first thing people need to do to understand their environment with regard to the pandemic. And the CEAT is going to be one of those tools.”