Predicting Epidemics and the Impact of Interventions in Heterogeneous Settings: Standard SEIR Models are too Pessimistic
研究发现，忽略传播和控制的异质性会高估传播率和疫情规模，导致干预效果预测过于悲观；而异质性越高，后续疫情潜力越低。
Abstract The basic reproduction number (R0) is an established concept to describe the potential for an infectious disease to cause an epidemic and to derive estimates of the required effect of interventions for successful control. Calculating R0 from simple deterministic transmission models may result in biased estimates when important sources of heterogeneity related to transmission and control are ignored. Using stochastic simulations with a geographically stratified individual-based SEIR (susceptible, exposed, infectious, recovered) model, we illustrate that if heterogeneity is ignored (i.e. no or too little assumed interindividual variation or assortative mixing) this may substantially overestimate the transmission rate and the potential course of the epidemic. Consequently, predictions for the impact of interventions then become relatively pessimistic. However, should such an intervention be suspended, then the potential for a consecutive epidemic wave will depend strongly on assumptions about heterogeneity, with more heterogeneity resulting in lower remaining epidemic potential, due to selection and depletion of high-risk individuals during the early stages of the epidemic. These phenomena have likely also affected current model predictions regarding COVID-19, as most transmission models assume homogeneous mixing or at most employ a simple age stratification, thereby leading to overcautious predictions of durations of lockdowns and required vaccine coverage levels.