Inferring COVID-19 spreading rates and potential change points for case number forecasts

Jonas Dehning; Johannes Zierenberg; Frank Paul Spitzner; Michael Wibral; Joao Pinheiro Neto; Michael Wilczek; Viola Priesemann

We combine the SIR model with Bayesian parameter inference and augment the model by a time-dependent spreading rate. The time dependence is implemented via potential change points reflecting changes in the spreading rate driven by governmental interventions. Based on three distinct measures taken in Germany, we detect three corresponding change points from reported COVID-19 case numbers. Already on April 1 we had reported evidence for the first two change points, and predicted the third one [30] . Now, with data until April 9, we have evidence for all three change points. First, the spreading rate decreased from 0. 40 This corresponds well to the strict contact ban, which was announced on March 22. While the first two change points were not sufficient to switch from growth of novel cases to a decline, the third change point probably brought this crucial reversal.