Computer model offers insights on COVID-19 super-spreading

New research compares transmission patterns of SARS-CoV-2 with those of influenza
A pastor preaches before a largely empty church to avoid transmission of COVID-19
To avoid COVID-19, Americans are avoiding crowds in enclosed spaces like this New Jersey church, a precaution supported by a new Fred Hutch computer model that shows how the disease is easily passed by "the wrong person, in the wrong place, at the wrong time" during surprisingly brief periods of contagiousness. Photo by Elsa / Getty Images

As the number of COVID-19 cases in the United States streaks toward 5 million, it may surprise you to learn that 80% of those who test positive for the disease never infect anyone else with the virus.

Instead, new research suggests the disease weaves its way, rapidly, into the population because of a confluence of events. They are largely driven by contact with people who happen, briefly, to be highly contagious in the wrong place, at the wrong time.

It is a phenomenon known as super-spreading, and experts at Seattle’s Fred Hutchinson Cancer Research Center are using complex computer simulations of viral transmission among individuals to gain a better understanding of how this virus, SARS-CoV-2, with an unknowing assist from super-spreaders, manages to wreak so much havoc in populations.

In research newly posted in medRxiv, a team led by Fred Hutch postdoctoral scientist Dr. Ashish Goyal, infectious diseases physician Dr. Joshua T. Schiffer and epidemiologist Dr. Bryan T. Mayer reports some surprising insights using a computer model that compares the behavior of COVID-19 with influenza.

While existing data shows that people who are infected with SARS-CoV-2 may shed tiny amounts of the virus for weeks, the model suggests that they usually only shed enough to be contagious for a brief period — likely only one or two days.

'The ethical thing to do as an individual is to walk around with the assumption that you're infectious and contagious, and that it's your responsibility to protect the public.'

— Dr. Joshua T. Schiffer

On the other hand, the model suggests that for many, but not all infected people, the two-day contagious period may have ended before a COVID-19 patient comes down with symptoms, and that before then, the person may already have spread the virus to dozens of others.

Schiffer emphasized that the model cannot rule out the possibility that a small minority of infected people may be contagious for an extended period during the symptomatic phase of illness.  

“The ethical thing to do as an individual is to walk around with the assumption that you’re infectious and contagious, and that it’s your responsibility to protect the public. That doesn’t change at all,” he said.

Super-spreading in COVID-19 vs. influenza

MedRxiv is a “preprint server,” an online journal that presents research for discussion prior to receiving peer review — the time-consuming professional critiques that are being bypassed due to the urgency of COVID-19. As such, Schiffer stresses that his research is preliminary and subject to revision, but important enough to be presented to colleagues in the field for their consideration.

The team’s model indicates that both influenza and COVID-19 share this brief contagious period in common. Yet SARS-CoV-2 may be more prone to super-spreading, their research suggests, because of airborne transmission, which makes it much more dangerous in crowded, confined spaces.

The six-foot rule for social distancing springs from evidence that COVID-19 spreads in droplets that can travel only short distances before settling to the ground. Airborne transmission envisions the virus can be carried in tinier particles of moisture that hang in the air longer and can drift across a room.

Suspicions that COVID-19 spreads through airborne transmission remain controversial even after the World Health Organization reluctantly acknowledged last month there was some evidence of it.

Dr. Joshua T. Schiffer
Dr. Joshua T. Schiffer is an infectious disease physician at Fred Hutchinson Cancer Research Center. Photos by Robert Hood / Fred Hutch News Service

To gain a better understanding of COVID-19’s spread, the team built their computer model to simulate the transmission patterns of viruses, based on similar methods that Mayer and Schiffer have used to track the behavior of several herpesviruses. The goal was to simulate the infectious behavior of COVID-19 using thousands of scenarios based on different assumptions about factors that might influence transmission.

Among those assumptions were:

  • Infectious dose: how many millions of viral particles are needed to actually infect a person.
  • How steeply infection risk increases among people as the assumed dose goes up.
  • The average number of times an infected person has a potential transmission exposure with an uninfected person per day
  • A measure of whether that average number of exposures is constant or varies greatly from day to day.

After loading different numbers for these and other variables into their computers, the researchers ran approximately half a million simulations that — using the magic of computer algorithms — spat out thousands of predicted courses of infection. Think of them like strands of spaghetti tracking across a chart.

Next, they plotted the course of actual transmissions based on epidemiological data — more spaghetti — and found that only a few of the simulated scenarios matched that of the real-world data. At that point, the researchers could infer which of their assumptions about the virus were likely to be correct.

When people are shedding virus at their peak

The scientists also ran transmission scenarios for influenza and found some similarities and some differences that have implications for understanding how to prevent this new disease.

“Let’s say you have two people walk into a crowded, closed room, with poor ventilation,” Schiffer explained, “and one of those people has influenza, and one has SARS-CoV-2. Both are unfortunately shedding at the highest viral load possible. Our model shows the person with influenza will likely expose far fewer people to their virus within that crowded environment than the person with SARS-CoV-2.

“That’s what drives the results in our model. And to me, the most likely explanation for that would be a predisposition towards aerosolization — meaning the virus is physically dispersed over a larger area and perhaps for a longer duration of time with SARS-CoV-2 than with influenza.”

Another inference from the model: “Super-spreader events come from when people are shedding virus at their peak,” Schiffer said. That underscores the importance of knowing, if possible, how soon after exposure a person might reach their two-day window of contagiousness. In other words, most people who are infected with SARS-CoV-2 do not infect anybody, but many people have the potential to be super-spreaders.

“They have to show up in a crowded place, and they have to do so when they are shedding at a high viral load,” he said.

Dr. Dan Reeves, a research associate in Schiffer’s lab and co-author of the paper, said that there are some potentially reassuring findings from the computer simulations. One is that contagiousness lasts only a couple of days, so repeat testing after a positive result may not be necessary. Also, any early treatment that can reduce the viral load (the amount of virus measured from a swab sample) might have “an outsized impact on transmission reduction.”

Finally, said Reeves, because masks can help reduce the amount of virus, even slightly, during an exposure, “wearing a mask really will help!”

Sabin Russell is a former staff writer at Fred Hutchinson Cancer Center. For two decades he covered medical science, global health and health care economics for the San Francisco Chronicle, and he wrote extensively about infectious diseases, including HIV/AIDS. He was a Knight Science Journalism Fellow at MIT and a freelance writer for the New York Times and Health Affairs. 

Are you interested in reprinting or republishing this story? Be our guest! We want to help connect people with the information they need. We just ask that you link back to the original article, preserve the author’s byline and refrain from making edits that alter the original context. Questions? Email us at

Related News

All news
Coronavirus outbreak puts 'open science' under a microscope Quick release of data could stop an epidemic, disrupt how research is reported February 13, 2020
The present and future of science intersect in Seattle Highlights from the annual meeting of the American Association for the Advancement of Science February 19, 2020
Persistent HIV infection works a lot like cancer, study shows Ecology-inspired analysis suggests why reservoirs of HIV linger in treated patients November 19, 2018

Help Us Eliminate Cancer

Every dollar counts. Please support lifesaving research today.