The impact of containment measures and air temperature on mitigating COVID-19 transmission: non-classical SEIR modeling and analysis

Early non-pharmaceutical interventions (NPIs) are crucial to prevent and control of COVID-19 pandemic. We established a stochastic non-classical SEIR NPIs model (ScEIQRsh) which can quantify the three kinds of NPIs measures simultaneously to mimic the clustered intra-family or intra-acquaintance spreading pattern of COVID-19 under the effective integrated NPIs in Mainland China. Model simulation demonstrated that measures to diminish contactable susceptible (Sc), such as home confinement, travel constraint, social distancing etc. and measures to avoid delay of diagnosis and hospitalized isolation (η) were more effective but consumptive than contact tracing (κ, ρ). From fitted model by MCMC method, the proportion of asymptomatic infectors was 14.88% (IQR 8.17%, 25.37%). The association between air temperature and the fitted transmission rate (β) of COVID-19 suggests that COVID-19 pandemic would be seasonal with the optimal temperature range of 5°C-14°C and peak of 10°C for spreading, and vaccine is indispensable to ultimate prevention COVID-19.