The first three months of the COVID-19 epidemic: Epidemiological evidence for two separate strains of SARS-CoV-2 viruses spreading and implications for prevention strategies

Knut M. Wittkowski

Without repeated broad testing, however, that peak prevalence of infections cannot be directly observed, but it is known to be followed about one week by peak number of diagnoses (new cases). This is too late to make a decision, but the SIR model shows that the peak in diagnoses is preceded by two weeks by the "turning point" in cases where the curve of the new cases changes from increasing ever faster to increasing ever more slowly (day 76). The turning point can be estimated from the observed cases in time to making a decision. (It is also about 50% of the peak number of new cases, which one might be able to predict.) Hence, peak prevalence (of infections) follows the turning point/half peak (in number of cases) by about a week. The window of opportunity for starting an intervention is the week following the turning point in number of diagnoses (new cases) per day.