In Germany, growth of covid-19 cases declined after a series of three social distancing interventions, detectable at a two-week delay following each intervention. However, it was only after the third—a far-reaching contact ban—did cases decline significantly. These results from a modeling study designed to better estimate the impact of various levels of social distancing on virus spread indicate that the full extent of social distancing interventions was necessary to stop exponential growth in Germany, the authors say.
Further, the two-week delay in understanding an intervention’s impacts warrants caution in lifting restrictions, the authors of the study say; doing too much too early could leave policymakers and others “effectively blind” to a worsened situation for nearly two weeks. In early days of a pandemic, reliable short-term forecasts of the impacts of interventions like social distancing are key to decision makers. When reliability of such forecasts is challenged initially, when case numbers are low, Bayesian modeling can help.
Read more: Pakistan manufacturing covid-19 oxygenators
Jonas Dehning and colleagues combined a Susceptible-Infected-Recovered transmission model with Bayesian parameter inference to better estimate the impact of social distancing on SARS-CoV-2 virus spreading rate in Germany, where three interventions—starting with a 7 March cancelation of large public events and ending with a far-reaching contact ban late in the month—were implemented over three weeks. Using data on deaths through 21 April, the authors report evidence of three change points, each detectable two weeks after an intervention, and reflecting slowed spread of the virus. Only by the third change point, however, initiated by contact ban, did they see a crucial decline in new cases daily.
Further simulations suggest that the consequences of delaying social distancing by as little as five days can have severe effects, the authors report. They say that their findings of a two-week delay indicate it is important to consider lifting restrictions only when the number of active cases is so low that a two-week increase will not pose a serious threat to healthcare infrastructure.
The authors maintain that while their study was applied to Germany, the approach can be adapted to other countries or regions.