School of Veterinary Medicine, University of Pennsylvania, Kennett Square, PA 19348, USA.
Epidemics. 2011 Jun;3(2):61-70. doi: 10.1016/j.epidem.2011.02.003.
A stochastic, spatial, discrete-time, SEIR model of avian influenza epidemics among poultry farms in Pennsylvania is formulated. Using three different spatial scales wherein all the birds within a single farm, ZIP code, or county are clustered into a single point, we obtain three different views of the epidemics. For each spatial scale, two parameters within the viral-transmission kernel of the model are estimated using simulated epidemic data. We show that simulated epidemics modeled using data collected on the farm and ZIP-code levels behave similar to the actual underlying epidemics, but this is not true using data collected on the county level. Such analyses of data collected on different spatial scales are useful in formulating intervention strategies to control an ongoing epidemic (e.g., vaccination schedules and culling policies).
宾夕法尼亚州家禽养殖场禽类流感疫情的随机、空间、离散时间 SEIR 模型。使用三种不同的空间尺度，即将单个农场、邮政编码或县内的所有鸟类聚集到一个点，我们获得了疫情的三种不同视角。对于每个空间尺度，使用模拟疫情数据估计模型中病毒传播核内的两个参数。我们表明，使用在农场和邮政编码级别收集的数据模拟的模拟疫情与实际的基础疫情相似，但使用在县一级收集的数据则不然。对在不同空间尺度上收集的数据进行此类分析有助于制定控制正在进行的疫情的干预策略（例如，疫苗接种计划和扑杀政策）。
