Kvinna med munskydd
Photo: Mostphotos


In recent weeks, after twenty years as a mathematician, Tom Britton has become extremely popular. He is interviewed by newspapers and magazines, writes debate articles and is hired as a lecturer at other universities. What everyone wants to know is of course: How serious will it be?

How large will the spread of Corona be in Sweden?

Matematikern Tom Britton
Tom Britton

“The answer depends very much on what preventive measures are being taken, by ourselves and by the authorities. It is estimated that the reproduction number R, the number infected by a person, is 2.5 before any restrictions or measures have been implemented in a society. Then 60-70 percent will be infected. If nothing is done, this is how many people will become infected in Sweden. But if we manage to reduce the number of reproductions with different measures, of course fewer will be infected. The magic value is R = 1. As soon as we get the reproduction number below 1 with different measures, it will quickly disappear. China succeeded in this by means of very comprehensive measures, and very quickly, so that less than 1 percent will be infected in Wuhan. I am not convinced that we will be as effective in Sweden.”

What can be learned from previous outbreaks of infection?

“Very much. The latest pandemic was the swine flu, but there are also differences. On the one hand a vaccine was quickly found then, and there was some natural immunity too.”

Have you been hired by the Public Health Agency of Sweden to help?

“I am acquainted with several people at the Public Health Agency of Sweden and I have had close and daily contact with their modeling group, where one of my former doctoral students works.”

How would you describe your research area?

“My main area is mathematical models for the spreading of infections, and also statistical methodology for how to determine a model’s parameter values, such as R, from outbreak data. It is called "taxes" in statistical language.”

Have you researched any other infectious agents?

“In principle, I never research on individual epidemics or diseases, but work with more general models that can be adapted to several diseases. Of course, it is not a one-size-fits-all but different types of illnesses have specific models. In recent years, I have been specifically involved with HIV, which uses sequence data from viruses in infected patients to learn about the spread of infection.”

How can mathematics be used to calculate the spread of infection?

“Mathematical models can model most phenomena, obviously more or less good depending on what it is about. The main principle is to simplify reality and to include only the most essential in the phenomenon of interest, and then to analyze the model and see what happens.”

What does this mean in the corona case?

“For the coronavirus outbreak, this is about estimating R by estimating the so-called generation time, which is the typical time between getting infected and infecting others. If we manage to estimate the generation time, for example through contact tracking, and observe the growth rate of the epidemic, then we can deduce what R is approximately. With this knowledge of R we can calculate how many people will be infected if nothing is done. And above all, it provides guidance on how much we must push back the spread of the infection to bring it down to below 1 so that the outbreak subsides.”