SaTScan与FleXScan软件空间扫描统计量法在肺结核疫情空间聚集性研究中的应用比较

李婷; 何金戈; 杨长虹; 李京; 肖月; 李运葵; 陈闯; 吴建林

Abstract

李婷,何金戈,杨长虹,李京,肖月,李运葵,陈闯,吴建林.SaTScan与FleXScan软件空间扫描统计量法在肺结核疫情空间聚集性研究中的应用比较[J].Chinese journal of Epidemiology,2020,41(2):207-212

SaTScan与FleXScan软件空间扫描统计量法在肺结核疫情空间聚集性研究中的应用比较

A comparative study on SaTScan and FleXScan software for spatial clustering analysis regarding the incidence of pulmonary tuberculosis

Received:April 04, 2019

DOI：10.3760/cma.j.issn.0254-6450.2020.02.013

KeyWord: 肺结核时空扫描统计量聚集分析

English Key Word: Tuberculosis Spatial scan statistic Cluster analysis

FundProject:美国国立卫生研究院国际合作（1R01AI125842）

Author Name	Affiliation	E-mail
Li Ting	Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610041, China
He Jin'ge	Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610041, China	hejinge@163.com
Yang Changhong	Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610041, China
Li Jing	Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610041, China
Xiao Yue	Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610041, China
Li Yunkui	Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610041, China
Chen Chuang	Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610041, China
Wu Jianlin	Sichuan Provincial Center for Disease Control and Prevention, Chengdu 610041, China

Hits: 6207

Download times: 2063

Abstract:

目的探测2018年四川省肺结核发病的空间聚集性，识别高风险区域；比较SaTScan与FleXScan软件扫描统计量法在肺结核空间聚集性探测方面的应用效果。方法基于中国疾病预防控制信息系统传染病报告信息管理系统中四川省181个县（区）2018年肺结核疫情数据和人口数据，建立地理信息数据库，分别采用SaTScan 9.4.1和FleXScan 3.1.2软件中的Poisson模型探测肺结核发病聚集区域，比较2种方法探测到的聚集区域的位置和范围，通过ArcGIS 10.5软件进行可视化。结果 2018年四川省肺结核报告发病率为57.34/10万（47 601例），呈明显的聚集性分布。SaTScan和FleXScan软件分别扫描到8个和10个具有统计学意义（P < 0.05）的空间聚集区域，对数似然比（log-likelihood ratio，LLR）分别为24.62~2 416.05和1.48~2 618.96。2种方法的扫描结果中一级聚集区均覆盖了大、小凉山地区，即彝族聚居区，二级聚集区共同覆盖了部分川西高原少数民族地区。二者在扫描出的聚集区域形状和范围上有所差异。SaTScan的聚集区覆盖了部分实际发病情况并不高的县（区），而FleXScan能将其从聚集区中区分出来，探测出更准确的不规则形状聚集区。结论四川省肺结核疫情存在明显的空间聚集性，大、小凉山地区和川西高原少数民族地区是高风险和重点防控区域。FleXScan更利于精准区分高聚集性区域中的“冷点”地区，更适合在四川省肺结核聚集性探测中应用。

English Abstract:

Objective To detect the spatial clustering and high risk areas of pulmonary tuberculosis (PTB) in Sichuan province in 2018 and, to compare the effects of application on both SaTScan 9.4.1 software and FleXScan 3.1.2 software to detect the PTB spatial clusters. Methods Geographic information database was established by using the incidence data of PTB and demographic data reported in the ‘China disease prevention of infectious disease reporting information management system’ in Sichuan province in 2018. Spatial clustering analysis was conducted using the Poisson model in software SaTScan 9.4.1 and FleXScan 3.1.2 to detect the high risk areas of PTB by software ArcGIS 10.5. Differences of clusters locations and scopes in the two scanning methods were compared. Results The reported incidence rate of PTB was 57.34/100 000 (47 601 cases) in Sichuan province in 2018, presenting an obvious clustering distribution. SaTScan and FleXScan scanned 8 and 10 clusters showed statistically significant differences (P < 0.05), with log-likelihood ratio (LLR) as 24.62-2 416.05 and 1.48-2 618.96, respectively. Results from scanning of the two methods showed that the most likely clusters appeared in the Daliangshan and Xiaoliangshan of Liangshan Yi ethnic aggregation areas. The other clustering areas would include some minority areas in the western Sichuan plateau, detected by both two methods differences in the shape and scope of the clustering were detected by both methods. The clustering scopes detected by SaTScan covered some counties, in which the actual incidence was not high. FleXScan could distinguish the clusters and detect more irregular shaped clusters. Conclusions Obvious spatial clusters of PTB distribution were found in Sichuan in 2018. Areas of Daliangshan, Xiaoliangshan and the minority areas in Western Sichuan plateau appeared at high risk, suggesting these were the key areas for prevention and control. FleXScan seemed more conducive in accurately distinguishing the "cold spot" areas in the highly aggregated areas, and more suitable for the application of spatial clustering detection for TB, in Sichuan province.

View Fulltext Html FullText View/Add Comment Download reader