What influences COVID-19 infection rates: A statistical approach to identify promising factors applied to infection data from Germany

Moritz Mercker; Uwe Betzin; Dennis Wilken

In summary, we present a new statistical approach which is specifically designed to identify factors influencing local infection rates. Since only relative changes in detected infections are considered, the approach is robust with respect to the percentage of undetected infections, as long as the latter does not locally change in time. The approach has been applied to spatio-temporal COVID-19 infection data in Germany. Nevertheless, it can be adapted to other countries at any state of infection, provided that both, infection numbers and factors affecting them, are provided in a sufficient spatio-temporal resolution. The analysis of daily COVID-19 infections in Germany reveals that (1) currently, infections are in average significantly decreasing; (2) there is a strong spatial variability in the local recent infection rates, especially varying in the centre of Germany (mostly including the two states Thüringen and Hessen); (3) there are strong evidences that both, an early emergence of first infections and high local infection densities, lead to strong recent decays in infection rates. The latter may indicate that psychological effects (such as the awareness of danger due to infection cases in the circle of friends and acquaintances) lead to a more cautious behaviour, which again reduces the local risk of additional infections.