Social bubbles reduce risk of COVID-19 superspreader events

Any given person sick with the flu spreads it, on average, to the same number of people. But for some diseases, certain infected individuals—so-called superspreaders—infect far more people than others do. Along with measles and Ebola, COVID-19 seems to be one such disease; the top 10% most infectious people produce 80% of infections.

Bjarke Frost Nielsen of the Niels Bohr Institute at the University of Copenhagen and his colleagues have used a superspreader model for COVID-19 to quantify how reducing the size of active social networks may help curtail the epidemic. (For more on COVID-19 modeling, see Physics Today, June 2020, page 25 , and the article by Alison Hill, Physics Today, November 2020, page 28 .) In their model, each individual has their own infectiousness—the probability of infecting another individual per unit time. The individuals are in a network, and at every time step, each person interacts with a random other person in their immediate social circle.

The researchers found that without superspreaders (left graph) the number of different people each individual interacts with didn’t strongly influence the number of infections. But with superspreaders (right graph), reducing the size of individuals’ social circles from unrestricted (purple) to 10 contacts (orange) dramatically shrinks both the peak and total number of infections. Decreasing the duration and number of social interactions within that social group was not necessary to reduce infections.

The data above are for an evenly distributed network, but when the social ties are clustered, the influence of reduced social contacts is more pronounced. The finding supports the efficacy of social bubbles. (B. F. Nielsen, L. Simonsen, K. Sneppen, Phys. Rev. Lett. 126, 118301, 2021 .)