To overview

Large heterogeneous gatherings cause most COVID-19 spread

New research supported by the Carlsberg Foundation shows that we can reduce general infection transmission far more by limiting contacts between people who do not regularly meet than by limiting contacts in well-defined social groups consisting of people who often meet. The study has just been published in the journal PNAS – Proceedings of the National Academy of Sciences.

All over the world, attempts have been made to manage the COVID-19 pandemic by introducing bans on gatherings and implementing unprecedented lockdowns of social institutions, workplaces, shopping and cultural life. It is well documented that such measures only reduced infection transmission moderately in connection with Spanish influenza around a century ago. During the COVID-19 pandemic, however, these mitigation measures have proven to be a surprisingly effective means of preventing the spread of infection.

Over the past year, Semper Ardens researcher Professor Lone Simonsen and fellow researchers have tried to understand why this is the case. Given that these draconian measures also have huge social costs, there is a need for a more precise understanding of which exact measures are most effective in terms of keeping infection down.

Now Professor Simonsen and Professor Kim Sneppen of the Niels Bohr Institute are ready with their proposition of what is driving corona infection and how best it can be controlled. They present the results in a new agent-based model study carried out with fellow researchers Bjarke Frost Nielsen and Robert J. Taylor. The study “Overdispersion in COVID-19 increases the effectiveness of limiting nonrepetitive contacts for transmission control” is published today in the journal PNAS – Proceedings of the National Academy of Sciences.

“This is really interesting novel insight that Lone Simonsen and her colleagues present here,” says Professor Flemming Besenbacher, Chairman of the Board of Directors of the Carlsberg Foundation. “The past year of corona has shown us that we need new research that can give us a much greater understanding of COVID-19 infection transmission in order to be able to implement restrictions that are optimal for preventing infection transmission. Particularly in light of the huge costs that the restrictions entail for our society. This study gives a new solid starting point for better management of both the COVID-19 pandemic and other future pandemics.”

Big events with large numbers of participants carries the greatest risk of superspreading

Overall, the study shows that we can limit infection transmission far more by limiting contacts between people who do not regularly meet than by limiting contacts in well-defined social groups consisting of people who often meet.

“It has long been known that the COVID-19 pandemic is characterised by superspreading,” says Professor Simonsen. “Several studies found that perhaps 10% of infected people account for 80% of new infections. Our agent-based model study demonstrates for the first time that this phenomenon is a real Achilles’ heel for the SARS-2 virus”.

“We demonstrate with this study that the pandemic can be effectively controlled by shutting down big events and other situations in the public space where an infected person can come into contact with a large number of new people. This is something that could never be done with pandemic influenza, which is not driven by superspreading.”

It’s all about k

In the study, the researchers operated with three categories of gathering: “close” (a small, unchanging group of mutual contacts, as might be found in a household); “regular” (a larger, unchanging group, as might be found in a workplace or school); and “random” (drawn from the entire model population and not repeated regularly).

When researchers want to calculate how the infection is spreading, they must include consideration of the dispersion parameter k. When k is low – i.e. when there is greater heterogeneity in a group of people who do not meet regularly – the risk of infection by superspreading increases. By contrast, when k is high – i.e. when there is less heterogeneity in a group and people meet regularly – there is not a big risk of superspreading.

“When k is low, the difference between infected people is high; most infect very little, while a few are responsible for the majority of infection,” says Professor Simonsen. “COVID-19 has a k of around 0.1, which equates to 10% of infected people causing around 80% of new infections. Consequently, infection can be effectively limited by restricting our behaviour in public spaces so that no one will be able to infect a lot of people. And in this way the epidemic can be controlled. The important thing here is that it’s not enough to just limit the number of people you meet, but also that you only meet with a few people throughout the period when you are infectious, which lasts about one week.”

The researchers behind the study point out, however, that because variant B.1.1.7, now dominant in Denmark, may have a lower tendency to superspreading than the original variant, it will require even greater social restrictions than previously to control the pandemic.

Study with huge implications

The new study comes at an important time when we are in the process of easing the last lockdown of Denmark. If we consider the results, the greatest benefit is obtained by limiting large gatherings of people who do not normally meet rather than by restricting access to small gatherings of people who often meet.

“The study has huge new implications,” says Professor Sneppen. “There is a new factor that we need to include in the assessment of future pandemic threats. It’s called k and defines the degree of superspreading. This is a phenomenon that also characterised the SARS outbreak in 2003, and it was probably a factor that contributed to our success in halting it that same year.”

Go to the article “Overdispersion in COVID-19 increases the effectiveness of limiting nonrepetitive contacts for transmission control”

Read more about Professor Simonsen’s research project on COVID-19 and historical pandemic models 



To overview