Despite much effort, influenza continues to be one of the major public health concerns. To understand its transmission dynamics, extensive epidemiological models have been developed in various spatial contexts, from small-scale community to larger-scale area such as city and country. However, there is a significant lack of rigorous models designated for characterizing specific transmission patterns in fine-scale spatial contexts, particularly, in public indoor venues such as shopping malls, airports, and grocery stores. In fact, the transmissions in such settings could be very critical as the infected individuals can quickly spread the disease area wide. In modeling the transmission dynamics in public venues, one of the biggest challenges is to appropriately capture the “random” contacts occurred between individuals within a dynamic and highly mobile population. Focusing on the public indoor environment, this study conceptualized an agent-based modeling framework for investigating human-to-human transmissions via an explicitly represented contact network. By applying this framework, a computer simulation model was developed to mimic how the influenza can be transmitted in a real shopping mall in the U.S. The impacts of three contributing factors – timing, number and role (shopper or mall employee) of the introductory cases – was examined under different hypothetical disease scenarios using computer experiments. Furthermore, the effectiveness of employee-targeted vaccination and public social distancing strategies were evaluated with different vaccine efficacies and mall hour reductions, respectively. Our findings are valuable for local health authorities to take actions as they prepare and control the seasonal flu as well as the potential epidemics.

尽管付出了很多努力，但流感仍然是主要的公共卫生问题之一。为了了解其传播动态，已经在各种空间环境中开发了广泛的流行病学模型，从小规模社区到大范围区域（例如城市和乡村）。然而，特别是在大型购物中心，机场和杂货店等公共室内场所，严重缺乏严格的模型来指定特定传输模式的特征，这些模型用于在小规模空间环境中表征特定的传输模式。实际上，在这种情况下的传播可能非常关键，因为受感染的个体可以迅速将疾病传播到整个地区。在模拟公共场所的传播动态时，最大的挑战之一是要适当地捕捉动态和高度流动的人群中个体之间发生的“随机”联系。专注于公共室内环境，本研究概念化了一个基于代理的建模框架，用于通过明确表示的联系网络调查人与人之间的传播。通过应用此框架，开发了一种计算机模拟模型，以模拟如何在美国真实的购物中心中传播流感。介绍性案例的三个因素-时间，数量和角色（购物者或购物中心员工）的影响-使用计算机实验在不同的假设疾病情景下进行了检查。此外，以员工为目标的疫苗接种和公共社会疏远策略的有效性分别通过不同的疫苗效力和购物时间减少进行了评估。我们的发现对于地方卫生部门在准备和控制季节性流感以及潜在的流行病时采取行动非常有价值。