基于社会网络理论的流感传播特征及防控措施效果评价

发布时间：2018-03-01 06:31

本文关键词： 流感流行病学社会接触网络接触层次 SIR 干预措施效果　出处：《华中科技大学》2010年博士论文　论文类型：学位论文

【摘要】： 研究目的借助目前复杂网络及社会网络理论领域研究各种经典分析方法和指标,从动态和静态两个角度分析流感在人群接触网络中的传播特征,从微观行为与宏观结构两个层面构建基于社会网络的流感传播模型,利用计算机进行模拟和定量分析,提出预防控制措施和管理模式,从而有效地预防和控制流感在人群中的传播。研究方法 1、调查学校青少年在社会网络中与其他人的接触情况和流感防控知识知晓情况。调查问卷内容主要包括：个人基本情况、日常生活中的接触行为、流感基本知识、态度、行为和需求。接触行为数据分年级和群组排列,计算不同群组活动变量的平均值(AVG)、变异系数(CV)等。本调查数据应用EPI DATA 3.1完成数据的录入和整理,运用EXCEL 2003和SPSS13.0进行统计分析。 2、调查发生流感疫情的湖北省随州市某小学四年级某班全体学生。调查问卷主要包括两部分内容：个人基本情况和学生间的接触行为。构建一个相对封闭的班级内部学生接触的整体网络,进行计算机录入分析,获取网络特征参数。用UCINET网络分析集成软件进行整体网络分析,接触行为特征数据采用EPI DATA3.1录入并核对,用EXCEL 2003和SPSS 13.0进行统计分析。 3、选取湖北省人口作为研究对象,构建个体和群体两个层面的模型。个体模型基于拓展后的经典SIR模型,群体模型是基于真实表现个体之间接触的社会接触网络(SCN)模型。仿真模拟开始于10万个人的社会接触网络和一个受感染的个体,使用InfluSim分析在没有提供疫苗的情况下,药物和非药物措施减轻流感大流行的效果。研究结果 1、不同年级间各群组平均接触时间的分布比较均衡,没有比较大的起落变化。在列出的十个群组中,家庭群组具有最高的平均接触时间,其次是学校课堂群组。各年级级别中,小学生中接触层次比初中生和高中生高。但是各群组接触层次时间值的总和以高中年级最高,小学最低。特别是少数几个学生的接触层次时间值非常高,这些接触层次时间值很高的人有可能成为“超级传播者(super spreader)"。 2、流感发生前后班级内学生之间接触发生了很大变化,流感发生后网络结构整体上比流感发生前要松散得多。流感发生前班级内日常接触网络比流感发生后班级内部接触网络的网络距离小,聚类系数大,网络的分散程度小,凝聚程度高。流感发生前后各节点的度和相对中心度的均值分别为9.1和6.5,大部分节点的度有不同程度的减少。流感发生后班级内部接触网络中主要接触人数下降了2个；平均接触时间由每人每天2.6小时降为1.9小时,与主要接触者的接触时间总和由流感发生前的17.2小时下降为8.4小时。 3、没有干预措施时人群感染率为83.0%,采取药物治疗、社会距离和综合措施后感染率分别为69.0%、67.0%和54.0%。没有干预措施时流感流行引起死亡人数为111人,采取干预措施后,死亡数大大降低。不同干预措施对暴露者、隐性感染者和严重病例者人数也有一定影响。没有干预措施时流行持续时间为79天,采取治疗措施、社会距离和综合措施后,流感流行持续时间延长,增加率分别为19.0%、44.3%和75.9%。各年龄段的感染率也大大降低,分别降低到2.3%、5.7%、3.8%、3.3%、2.9%和1.8%。研究结论 1、应用SCN理论描述人群中接触行为的特征,找出对流感传播具有重要作用的群组或个体,有助于合理设计社会接触距离措施。并且,提高流感主要侵害对象的认知,能促进防控措施的顺利实施。 2、流感发生后班级内部整体网络结构变得松散了,个体节点的位置发生了一些变化,平均每个学生密切接触的学生人数减少,接触时间与接触层次时间值也显著下降,表明流感防控措施产生了一定效果。人群接触行为也发生了改变,有利于控制流感在人群中的快速传播。 3、在流感高速传播时,及时应用抗病毒药物(即使是有限的药品)和快速实施减少社会接触的措施将显著延缓流行高峰的到来,并且大幅度降低其高峰期的感染人数,降低流感流行的凶险程度。 1、应用社会接触网络(SCN)理论分析流感在人群中传播的特征,找到流感防控的重点关注对象,利于有针对性地采取防控措施,达到效益的最大化。 2、利用整体网络分析方法分析流感疫情发生前后接触网络拓扑结构和个体接触行为的变化,为定量评价不同流感防控干预措施的效果奠定基础。3、结合SIR和SCN构建合适的流感传播模型,模拟流感在人群中传播的过程,定量评价不同的流感防控干预措施的效果。
[Abstract]:research objective
With the help of the complex network and social network theory research in the field of various classical methods and indicators, from two aspects of static and dynamic analysis of influenza transmission characteristics in the network contact in the crowd, from the two aspects of micro behavior and macro structure construction based on social network analysis model of influenza transmission, using computer simulation and quantitative, put forward prevention control measures and management mode, so as to effectively prevent and control the spread of flu in the crowd.
research method
1, contact and knowledge of influenza prevention and control awareness of school adolescents and the other people in the social network. The contents of the questionnaire include: basic personal information, contact behavior in daily life, influenza basic knowledge, attitude, behavior and needs. The contact behavior data and grade group arrangement, average value of different group activities variable (AVG), the coefficient of variation (CV). The survey data using EPI DATA 3.1 to complete the data input and processing, using EXCEL 2003 and SPSS13.0 for statistical analysis.
2, investigation on the occurrence of influenza epidemic in Suizhou city of Hubei Province, the fourth grade of a primary school in a class all students. The questionnaire mainly includes two parts: basic information and personal contact behaviors between students. The whole network to build a relatively closed internal class students contact, computer input analysis, access to the network parameters. The whole network analysis using UCINET network analysis software, using EPI DATA3.1 data characteristics of contact entry and check, were analyzed by using EXCEL 2003 and SPSS 13.
3, population in Hubei province were selected as the research object, the construction of two aspects of individual and group model. Individual classic model of the expanded SIR model based on population model is social contact network contact between the real performance of the individual based (SCN) model. Social contact network simulation began in 100 thousand and an infected people the individual does not provide analysis in the case of vaccine using InfluSim, drug and non drug measures to reduce influenza pandemic results.
Research results
1, the distribution of different grades of each group average contact time is relatively balanced, no landing relatively large changes. In the ten group are listed in the family group has the highest average contact time, followed by the school classroom group. Each grade level, level higher than junior high school students and high school students. But the sum of contact the value of the group level exposure to the highest grade high school, primary school. Especially the lowest level of contact time of a few students' value is very high, the high level of contact time value may become "superspreaders" (super spreader).
2, between before and after the occurrence of influenza in class contact has changed greatly, the whole network structure after the occurrence of flu flu than before to loose much. Influenza occurred before the class daily contact network of small distance than the flu after class internal contact network, clustering coefficient, degree of dispersion of small networks, agglomeration high. The mean degree and the relative center of each node degree before and after the occurrence of influenza were 9.1 and 6.5, most of the nodes have different degrees of reduction. The flu after class internal contact network main contact number decreased by 2; the average contact time by 2.6 hours per day reduced to 1.9 hours, the total contact time and the main contacts by influenza occurred before 17.2 hours reduced to 8.4 hours.
3, no intervention measures for infection rate of 83%, take medication, social distance and comprehensive measures after the infection rates were 69%, 67% and 54.0%. without intervention measures for flu epidemic caused 111 deaths, after the intervention, the number of deaths is greatly reduced. With the intervention of exposure, the number of hidden serious cases of infection and also have certain effect. There are no interventions when the epidemic lasted for 79 days, treatment measures, social distance and comprehensive measures after the flu epidemic duration, increase rate is respectively 19%, 44.3% and 75.9%. of all ages the infection rate is greatly reduced, reduced to 2.3%, 5.7% respectively. 3.8%, 3.3%, 2.9%, and 1.8%. research conclusion
1, apply SCN theory to describe the characteristics of people's contact behavior, find out groups or individuals that play an important role in the spread of influenza, and help to design social contact distance measures reasonably. Moreover, improving cognition of the main victims of influenza can promote the effective implementation of prevention and control measures.
2, after the occurrence of influenza classes within the whole network structure became loose, some changes have taken place in the individual node position, the average number of students per student in close contact to reduce the contact time and the contact time, the level was also significantly decreased, showed that influenza prevention and control measures have a certain effect. People contact behavior also changed, with the rapid spread of to control the flu in the crowd.
3, when influenza is spreading rapidly, timely application of antiviral drugs (even limited drugs) and rapid implementation of measures to reduce social contact will significantly delay the arrival of the peak of epidemics, and significantly reduce the number of infections at its peak, and reduce the risk of influenza pandemic.
1, we apply the theory of social contact network (SCN) to analyze the characteristics of influenza transmission in the population, find the key objects of influenza prevention and control, and take targeted measures to prevent and control, so as to maximize the benefits.
2, change analysis before and after the flu epidemic in the contact network topology and individual contact behavior by the whole network to evaluate different influenza control measures for the quantitative results lay the foundation of.3, construction of influenza appropriate propagation model based on SIR and SCN, to simulate the process of influenza spread in the population, the quantitative evaluation of different influenza control measures the effect.

【学位授予单位】：华中科技大学
【学位级别】：博士
【学位授予年份】：2010
【分类号】：R184

【相似文献】