Tracking a new disease at the very beginning of an outbreak is not an easy task. Adequate tools for diagnosis may not be available very early on. Further, a significant number of individuals may already be infected before the implementation of adequate tracking measures in the population. Factors that cannot be accounted for may also affect the reporting of new cases, such as testing efforts, bulk reporting or political interference. All these put into question the reliability of disease estimates based on case reports. However, some of these estimates are crucial to evaluate the risk of a pandemic and to plan countermeasures. The growth rate before countermeasures are put into place is one such measure. In this presentation, I will show how we used travel data and a simple sampling model to infer the growth rate of COVID in China while avoiding most of the factors hampering its reliability.
