Ripley's L指数与最近邻空间热点分析在流行病学标点地图分析中的应用  

Ripley′s L index and the Nearest Neighbor 'hot spots' analysis in epidemiological spots map analysis

作　　者：高杰[1] 王志强[2] 邵琦[1] 薛皓[3] 许桂春 李学刚 王洁贞[1] 薛付忠[1] 

机构地区：[1]山东大学公共卫生学院流行病与卫生统计学研究所,济南250012 [2]山东省疾病预防控制中心传染病防治所,济南250014 [3]山东大学医学院,济南250012 [4]菖南县卫生防疫站,山东莒南250000

出　　处：《山东大学学报（医学版）》2009年第3期89-93,97,共6页Journal of Shandong University:Health Sciences

基　　金：国家自然科学基金资助课题(30170527)

摘　　要：目的探讨Ripley’s L（d）指数与最近邻空间系统聚类分析在流行病学标点地图分析中的应用。方法采用实验流行病学的方法，以ArcGIS9．0为数据管理与分析平台，将Ripley’s L（d）指数分析与最近邻空间系统聚类分析结合，综合反应疾病空间异质性及其动态特征。结果实验疫区内，宿主鼠类第一聚集区平均半径为4．29m；最强聚集区平均半径为14．43m；最大聚集区平均半径为86．26m。各村聚集“热点”数差别较大，其一阶波动范围为3，8个，二阶波动范围为0-1个；小家鼠第一聚集区平均半径为4．86m；最强聚集区平均半径为21．14m；最大聚集区平均半径为92．57m。各村聚集“热点”数差别较大，其一阶波动范围为1-12个，二阶波动范围为0-2个；褐家鼠第一聚集区平均半径为5．00m；最强聚集区平均半径为32．71m；最大聚集区平均半径为93．86m。结论在空间流行病学领域，将Ripley’sL（d）指数分析与最近邻空间系统聚类分析相结合，能够为阐明宿主鼠类的空间分布特征，控制HFRS传染源提供统计学依据。Objective To explore the apphcation of Ripley＇s L function and the nearest neighbor hierarchical clustering ＇hot spots＇ analysis in epidemiological spots map analysis. Methods ArcGIS9.0 was used for data management and analysis. The experimental epidemiology method in combination with Ripley＇s L function analysis and nearest neighbor spatial clustering analysis compositely reflected the disease spatial heterogeneity and its dynamic characters. Results The average radius of the host rat was 4.29 meters, of the strongest cluster district was 14.43 meters, and of the biggest cluster district of host rat was 86.26 meters. The numbers of the ＂hot spot＂ in different villages greatly differed. The undulation range of the first order was 3-8 and of the second order was 0-1. The average radius of Mus norvegicus was 4.86 meters, of the strongest cluster district was 21.14 meters, and of the biggest cluster district was 92.57 meters. Numbers of the ＂hot spot＂ in different villages greatly differed. The undulation range of the first order was 1-12 and of the second order was 0-2. The average radius of Rattus norvegicus was 5.00 meters, of the strongest cluster district was 32.71 meters and of the biggest cluster district was 93.86 meters. Numbers of the＂hot spot＂ in different villages gready differed. The undulation range of the first order was 3-11, and there was no cluster hot spot of the second order. Conclusion In the spatial epidemiological field, Ripley＇s L index analysis in combination with the nearest neighbor spatial clustering analysis can provide statistical evidence for clarifying the spatial distribution of the host rat and controlling the HFRS infection source.

关 键 词：Ripley’s L(d)指数 聚类分析 空间流行病学 

分 类 号：R181.2[医药卫生—流行病学]