Generalized logistic growth modeling of the COVID-19 outbreak in 29 provinces in China and in the rest of the world

摘要: Background: the COVID-19 has been successfully contained in China but is spreading all over world. We use phenomenological models to dissect development of epidemics and impact drastic control measures both at aggregate level within each province. experience from analyze calibration results on Japan, South Korea, Iran, Italy Europe, make future scenario projections. Methods: we calibrate logistic growth model, generalized model Richards reported number infected cases Jan. 19 March 10 for whole China, 29 provinces four severely affected countries Europe as a whole. The different provide upper lower bounds our predictions. Results: quantitatively document phases outbreak with detailed analysis heterogenous situations across provinces. Based Chinese experience, identify high risk Japan estimated total confirmed 25 being 1574 (95% CI: [880, 2372]), 5669 [988, 11340]) by June. For expect approach ceiling, 7928 [6341, 9754]), 20 days. estimate 0.15% [0.03%, 0.30%]) Italian population be positive scenario. would 114867 people days, negative probable scenario, corresponding 0.015% European population. Conclusions: extreme containment implemented were very effective some instructive variations other countries, it almost inevitable see continuation coming months. are serious no short-term end expected. There significant concerning upcoming July 2020 Summer Olympics Tokyo. Iran9s situation highly uncertain unclear scenarios, while Korea approaching outbreak. Both USA early stages outbreak, posing health economic risks world absence measures.