【摘要】 目的 通过综合流行病学调查与校园社交网络分析，探讨水痘感染风险的关键因素。方法 采用基于超宽带（Ultra-Wideband,UWB）定位技术的无线可穿戴设备，针对某小学2024年5月水痘疫情暴发事件开展回顾性现场研究。以疫情波及班级为研究对象，利用高精度接触数据重构社交接触网络，系统采集接触人数、接触频次、接触总时长等关键参数。运用描述性分析方法分析学生的接触特征分布规律，继而构建贝叶斯logistic回归模型，解析水痘发病与传播关键危险因素。结果 本次调查的小学水痘暴发疫情罹患率为21.7%(10/46)，现场记录45名学生和1位老师的14 009次接触数据，接触人数中位数57人，单次接触中1 min内的接触占比为59.8%,1 019个接触对的累积接触时间中位数为28.1 min。近距离（≤1 m）单次接触中位时间6.6 min较远距离（>1 m）1.5 min长。贝叶斯多因素logistic回归模型表明，接种水痘疫苗（OR=0.031,95%CI:0.003～0.236）、近距离平均单次接触时间每增加1个标准差（1.938 min）(OR=4.621,95%CI:1.485～18.922)、度中心性每增加1个标准差（9.968）(OR=3.910,95%CI:1.494～12.555)即可增加水痘感染风险。结论 研究创新应用UWB技术量化了社交网络中心性对水痘传播的影响：疫苗接种显著降低感染风险，而个体近距离接触时长与网络度中心性高使风险提升，为构建基于接触网络的精准预警系统提供科学依据。更多还原

【Abstract】 Objective To investigate the key risk factors for varicella(chickenpox) infection by integrating epidemiological survey data with social contact network analysis. Methods A retrospective field study was conducted following a varicella outbreak in May 2024 at a primary school. We utilized wireless wearable devices equipped with UltraWideband(UWB) positioning technology to reconstruct the social contact network of the affected class. High-precision data on key parameters, including the number of contacts, contact frequency, and total contact duration, were systematically collected. Descriptive analysis was employed to characterize the distribution of student contact patterns. Subsequently, a Bayesian logistic regression model was constructed to identify critical risk factors associated with varicella incidence and transmission. Results The varicella attack rate during this primary school outbreak was 21.7%(10/46). A total of 14 009 contact instances were recorded among 45 students and one teacher. The median number of contacts per individual was 57. Of all single interaction events, 59.8% had a duration of less than one minute, and the median cumulative contact duration for 1 019 contact dyads was 28.1 minutes. The median duration of a single close-proximity(≤1 m) contact(6.6 minutes) was substantially longer than that of more distant(>1 m) contacts(1.5 minutes). The Bayesian multivariate logistic regression model indicated that varicella vaccination was a significant protective factor(OR=0.031, 95% CI: 0.003-0.236). Conversely, each standard deviation increase in the average duration of close-proximity contact(1.938 min)(OR=4.621, 95% CI: 1.485-18.922) and in degree centrality(9.968)(OR=3.910, 95% CI: 1.494-12.555) was associated with an increased risk of infection. Conclusion This study innovatively applied UWB technology to quantify the influence of social network centrality on varicella transmission. The findings demonstrate that vaccination significantly reduces the risk of infection, whereas prolonged duration of individual close-proximity contact and higher degree centrality within the network elevate infection risk. These results provide a scientific foundation for the development of precise, contact network-based early warning systems.更多还原