2015—2017年浙江省登记涂阳肺结核患者的空间分析

目的 分析浙江省登记涂阳肺结核患者的空间分布规律,探测聚集区域,为进一步研究结核病危险因素及制定防控策略提供理论依据。 方法 收集2015—2017年浙江省的89个县(市、区)登记涂阳肺结核患者的登记资料,将其与浙江省电子地图相匹配,利用ArcGIS 10.0 作为数据管理和呈现的平台,构建空间数据库。利用GeoDa 1.6.0软件分别计算全局Moran’s I指数和局部Moran’s I指数,以检验浙江省涂阳肺结核的空间自相关性;并采用SaTScan 9.3软件进行时空聚集分析,探讨浙江省涂阳肺结核登记分布的空间自相关性及聚集范围。 结果 2015—2017年浙江省共登记涂阳肺结核患者30292例,各县(市、区)的年均登记发病率为20.69/10万(4.73~45.61/10万)。浙江省各县(市、区)涂阳肺结核患者登记的分布具有全局空间自相关性(Moran’s I=0.429,Z=5.834,P=0.001);在局部空间自相关分析中,将25个P<0.05的具有局部空间正相关(高-高聚集模式+低-低聚集模式)与局部空间负相关(低-高聚集模式+高-低聚集模式)的县(市、区)筛选出来,其中高-高聚集模式[12个县(市、区)]+低-低聚集模式[11个县(市、区)]占92.0%(23/25),提示局部空间分析结果主要表现为空间正相关。利用SaTScan软件进行时空聚集分析,共探测出包含18个县(市、区)的4个聚集区域,每个聚集区域差异均有统计学意义[对数似然比(LLR)值分别为211.54、57.66、51.70、44.47;P值均<0.001]。 结论 浙江省登记涂阳肺结核分布情况存在明显的空间自相关,具有较强的空间异质性。

Spatial analysis on smear-positive pulmonary tuberculosis in Zhejiang province between year 2015 and 2017

Objective To analyze the spatial distribution pattern of smear-positive pulmonary tuberculosis patients registered in Zhejiang province, and determine the gathered areas, providing theoretical basis for further study of the risk factors of tuberculosis and prevention and control strategies. The registration data of smear-positive pulmonary tuberculosis in 89 counties in Zhejiang province from 2015 to 2017 was collected and was merged with a vector map to build spatial databases by ArcGIS (version 10.0). Global Moran’s I and local Moran’s I were calculated by GeoDa (version 1.6.0) software respectively, as well as spatial-temporal analysis were studied by SaTScan (version 9.3)software to detect the spatial autocorrelation and cluster range of the distribution of smear-positive pulmonary tuberculosis in Zhejiang province. From 2015 to 2017, a total of 30292 cases of smear-positive pulmonary tuberculosis were registered in Zhejiang province, and the mean annual registered incidence rate of each region was 20.69/100000 (range: 4.73 to 45.61/100000). The distribution of smear-positive pulmonary tuberculosis of each region in Zhejiang province showed global spatial autocorrelation (Moran’s I=0.429,Z=5.834,P=0.001). While at the local scale, 25 regions (P<0.05) with positive local spatial correlation (high-high pattern and low-low pattern) and negative local spatial correlation (low-high pattern and high-low pattern) were sort out for analysis, in which 12 regions showed high-high pattern and 11 regions showed low-low pattern, accounting for 92.0% (23/25), indicating the main result of local spatial autocorrelation analysis was positive. Spatial-temporal analysis using SaTScan software detected four cluster areas including 18 regions, and each cluster area showed statistical significance (P<0.001) with log likelihood ratio (LLR) of 211.54,57.66,51.70,44.47, respectively. There are apparent spatial autocorrelation and relative strong spatial heterogeneity existed in the distribution of smear-positive pulmonary tuberculosis in Zhejiang province.