运用空间计量模型侦测广西结核病发病的气象影响因素

Detecting the correlation between atmosphere factors and tuberculosis incidence using spatial econometric model

目的探索广西壮族自治区传染性结核病的时间和空间变化趋势,为科学防控结核病传播模式研究提供依据。方法从传染病报告系统、气象局和环保局收集2016年广西壮族自治区报告的活动性结核病数据和气象数据。用“Moran′s I”(莫兰指数)全局和局部空间自相关统计量侦测结核空间聚类;用空间回归模型筛选结核病发病的气象影响因素。结果2016年共报告活动性结核病47701例。近7年度报告发病率呈下降趋势(χ^2=159.76,P<0.01)。2016年经验贝叶斯标化全局莫兰指数为0.262,表明结核病病例报告存在空间聚集性。东部地区表现为低发聚集。经过大气影响因素的拟合和模型拟合度筛选,空间误差模型(R2=0.763,对数释然比为-59.444,赤池信息准则=136.888,回归标准误=14.334)为最优模型,低年降雨量和低气温是结核病集中发病的危险因素。结论广西壮族自治区结核病总体疫情较高,低年降雨量和低气温是结核病相对高发的主要原因。

Objective To detect the spatiotemporal mode and atmosphere influence factors of tuberculosis (TB) in Guangxi Zhuang autonomous region,providing a basis for scientific research on prevention and control of tuberculosis and its transmission patterns. Methods Collect active tuberculosis data and meteorological data reported in Guangxi in 2016 from the infectious disease reporting system,the meteorological bureau and the environmental protection agency.The "Moran′s I" global and local spatial autocorrelation statistic was used to detect the spatial clustering of tuberculosis;the spatial regression model was used to screen the meteorological influencing factors of tuberculosis. Results In 2016,47 701 cases of active tuberculosis were reported.The incidence rate has decreased in the past 7 years (χ^2=159.76, P <0.01).Moran′s I global statistic was 0.262 in 2016,indicating the spatial cluster from global perspectives.The eastern region was characterized by low incidence.After fitting the atmospheric influence factors and screening the model fit,the spatial error model ( R 2=0.763;Log Likelihood=-59.444;AIC=136.888,S.E of regression=14.334) was the optimal model,low annual rainfall and low temperature were the risk factors for the concentration of tuberculosis. Conclusion The overall epidemic situation of tuberculosis in Guangxi is high.Low annual rainfall and low temperature are the main reasons for the relatively high incidence of tuberculosis.