A paper published in the journal “Science” at the end of December could substantiate the current discussion about school closings in Germany with facts: An international team led by Jan M. Brauner, Sören Mindermann and Mrinank Sharma from the University of Oxford has run out of time In the course of reported infections and death dates, the relative effectiveness of various anti-corona measures was calculated. The researchers concluded that closing schools and universities and limiting meetings to a maximum of 10 people had the greatest effect on proliferation have the virus. The closure of shops, restaurants and pubs that are not absolutely necessary, on the other hand, only has a medium-sized effect, curfews only have a minor effect.

“If you simulate what is happening, you have to make relatively many assumptions,” says Jan Brauner, one of the three main authors of the study. “Our approach, on the other hand, is purely data-driven”. The researchers used infection data from 41 countries – mainly in Europe – between the 20 . January 2020 and the 30. May 2020. The core of the method is a hierarchical Bayesian model. First, the scientists calculated the course of the actually infected from the number of reported corona deaths and the number of reported infections.

From the comparison between the extrapolated further course of the infection and death numbers with the actual The researchers were then able to use data to calculate the real number of reproductions and thus the relative effectiveness of the anti-corona measures. Since the course of the measures was different in the individual countries, the researchers were also able to make statements about the relative effectiveness of individual measures

.

The researchers then classified a total of seven measures into three impact areas (high, medium and low ) a. The closure of schools and universities and the restriction of meetings to a maximum of people can therefore reduce the reproductive number of the virus – i.e. the average number of people who an infected person is infected – reduce by up to 40 percent. The closure of shops, restaurants and bars that are not absolutely necessary by 25 percent. Additional curfews only show a minor effect of a maximum of ten percent.

However, the aggregation and consolidation of the data also has disadvantages. “The statements about the measures are quite general,” says Brauner, “but you can still see clear trends”. In order to be able to derive concrete recommendations for politics, one must take into account the context in the individual countries – such as the age structure. It is also advisable to “consult as many facts as possible”, including data from simulations and epidemiological studies.

The scientists are currently working on understanding “the second wave”. Brauner says that he cannot say whether the statements from the study, which applies to the first corona wave, can be transferred to autumn and winter. “We don’t have the data for that.” On the one hand, there would certainly be changes in people’s behavior that must be taken into account. On the other hand, effects like the new, more contagious Corona variant B 1.1.7 have to be taken into account. (wst)