Exit scenarios in a simulation model
Computer scientists at FAU produce model of the dynamic of the coronavirus pandemic
A way out of the coronavirus lockdown is dominating current political and public debate. There is currently much speculation about the effects easing the lockdown will have and reliable prognoses are almost impossible. A team of researchers at FAU led by Prof. Dr. Reinhard German has now completed detailed modelling of the Covid-19 pandemic and has derived potential strategies for a controlled easing of restrictions. The group has published its results in a pre-print: ‘Modeling Exit Strategies from COVID-19 Lockdown with a Focus on Antibody Tests’.
Without drastic measures to contain the spread of the Covid-19 pandemic, there are fears that the German healthcare system will become overburdened thus endangering the lives of many people. What is also certain is that a lockdown lasting several months would have a considerable impact on the economy and society as a whole. Politicians and scientists are therefore working together to find strategies to ease the current restrictions in a controlled manner while protecting the health of the population and that of those in high-risk groups in particular. Whilst there is much speculation about the efficacy of relaxation measures, there are hardly any reliable prognoses based on valid figures and models.
Models simulate pandemic
Under the leadership of Prof. Dr. Reinhard German, a team of researchers at the Chair of Computer Science 7 (Computer Networks and Communication Systems) at FAU has now developed two simulation models in a very short space of time that can be used to reproduce the progress of the Covid-19 pandemic and assess the effects of controlled easing of current restrictions. ‘The first model is based on system dynamic calculations and essentially uses key indicators such as the basic reproduction number, incubation period, or severity of the progression of the disease, which are also used by the Robert Koch Institute,’ explains Prof. German.
‘The second model is a so-called agent-based simulation. Here, we look at single individuals and can make precise statements about their behaviour.’ For example, in the agent model, the scientists simulated in which locations certain groups of people can meet and infect each other, such as their families, during leisure activities, at work or during stays in hospital.
Social distancing restrictions until 2023 without vaccine
The FAU team was able to confirm the prognosis of the Robert Koch Institute for the progress of the pandemic while assuming the same conditions. If the lockdown were eased without further restrictions to social contacts, the German healthcare system would probably not be able to cope with the resulting situation.
‘In this scenario, the peak would only have been delayed. We would have to be prepared for up to 400,000 patients in intensive care during this peak and the number of deaths would be uncontrollable,’ says Prof. German. ‘Measures involving hygiene, such as wearing face masks, would flatten this curve a little, but the situation would still be threatening.’
The computer scientists have come to the conclusion with these modelling assumptions that repeated short-term restrictions to social contact might be required over a longer period of time – until March 2023 – to prevent overburdening the healthcare system. ‘This adaptive strategy would be a compromise between returning to normal life and preventing risk to the healthcare system until we have achieved herd immunity. This time period could be shortened if a vaccine becomes available before that,’ explains Prof. German.
Antibody tests and apps are useful
FAU’s models are currently the only ones that also take the influence of antibody tests into consideration. Restrictions to social contact could be lifted for people who have antibodies after being infected with the virus and who are thus probably immune, an aspect that is particularly relevant to vulnerable people and people who work in key areas.
‘As few as 50,000 antibody tests per day in Germany would identify 4.4 million people who have had the virus without showing any symptoms, in addition to those who know they have had the virus and have recovered. These patients could be excluded from the restrictions to social contact,’ explains Reinhard German. ‘Doubling the capacity for testing could increase this figure to more than 5.4 million people. All this information is, of course, subject to the very limited epidemiological data available and the assumptions in the models.’
The team regards the apps for digital tracing currently being discussed as a potential means of reporting contact with a person who has since shown symptoms of the virus. ‘The relevant apps can help us to trace the route of infection, improve our understanding of the dynamics of the pandemic and thus increase the efficient use of antibody tests,’ says German.
Further information
Prof. Dr. Reinhard German
Chair of Computer Science 7 (Computer Networks and Communication Systems) at FAU
reinhard.german@fau.de