Integrated Science Lab, Department of Physics, Umeå University, 90187, Umeå, Sweden.
Embedded Intelligent Systems Lab, Department of Computer Science, Electrical and Space Engineering, Luleå University of Technology, 97187, Luleå, Sweden.
Sci Rep. 2021 Nov 18;11(1):22512. doi: 10.1038/s41598-021-01857-4.
Identifying the critical factors related to influenza spreading is crucial in predicting and mitigating epidemics. Specifically, uncovering the relationship between epidemic onset and various risk indicators such as socioeconomic, mobility and climate factors can reveal locations and travel patterns that play critical roles in furthering an outbreak. We study the 2009 A(H1N1) influenza outbreaks in Sweden's municipalities between 2009 and 2015 and use the Generalized Inverse Infection Method (GIIM) to assess the most significant contributing risk factors. GIIM represents an epidemic spreading process on a network: nodes correspond to geographical objects, links indicate travel routes, and transmission probabilities assigned to the links guide the infection process. Our results reinforce existing observations that the influenza outbreaks considered in this study were driven by the country's largest population centers, while meteorological factors also contributed significantly. Travel and other socioeconomic indicators have a negligible effect. We also demonstrate that by training our model on the 2009 outbreak, we can predict the epidemic onsets in the following five seasons with high accuracy.
识别与流感传播相关的关键因素对于预测和减轻疫情至关重要。具体来说，揭示疫情爆发与社会经济、流动性和气候等各种风险因素之间的关系，可以揭示在疫情蔓延中发挥关键作用的地点和旅行模式。我们研究了 2009 年至 2015 年期间瑞典各城市发生的 2009 年甲型 H1N1 流感疫情，并使用广义反感染法（GIIM）评估了最重要的致病风险因素。GIIM 代表了网络上的传染病传播过程：节点对应于地理对象，链路表示旅行路线，分配给链路的传输概率指导感染过程。我们的研究结果证实了现有的观察结果，即本研究中考虑的流感疫情是由该国最大的人口中心驱动的，而气象因素也有显著影响。旅行和其他社会经济指标的影响可以忽略不计。我们还证明，通过在 2009 年的疫情中训练我们的模型，我们可以高度准确地预测接下来五个季节的疫情爆发。
