Spatio-temporal Clustering Characteristics Of Pulmonary Tuberculosis In Shanxi Province And The Study On Socio-economic Factors Based On The Geo Detector

Objective:This study aim to establish the geographic information database of pulmonary tuberculosis of Shanxi Province with the co unty(district)as spatial unit,conduct spatial-temporal clustering study,find out the high-risk clusters,radius of the clusters and the risk intensity of pulmonary tuberculosis in Shanxi Province,furthermore,discuss the socio-economic factors that may affect tuberculosis based on the surveillance data of pulmonary tuberculosis of 119 counties(districts)in Shanxi Province from 2008 to 2015.The study eventually provide scientific and rational basis for the accurate decision-making of regional prevention and control of tuberculosis in Shanxi Province.Methods:This study uses the surveillance data of pulmonary tuberculosis of 119 counties(districts)in Shanxi Province from 2008 to 2015,and collects other relevant data from Shanxi Statistical Yearbook and the National Earth System Science Data Sharing Infrastructure to describe the epidemiological characteristics of pulmonary tuberculosis in Shanxi Province by using three-dimension trend analysis and seasonal decomposition method with the county(district)as spatial unit.At the same time,the global spatial autocorrelation analysis of pulmonary tuberculos is in Shanxi Province was carried out to reveal the type of spatial association of tuberculosis in Shanxi Province,and local spatial autocorrelation analysis was used to detect the hotspot regions of tuberculosis in Shanxi Province.Then spatial-temporal scanning analysis was conducted to detect the high-risk clusters of tuberculosis,radius of clusters and the risk intensity.Finally,the relationship between Real GDP per capita,population density and tuberculosis,and interactions between factors were quantitatively detected by geo-detector,and the Real GDP per capita was calculated based on constant price.Arc GIS 10.2.2 and other related software were used to visualize the results.Results:1.The reported incidence of tuberculosis of 119 counties(districts)in Shanxi Province from 2008 to 2015 generally showed a downward trend.The trend in geographic location presented that the reported incidence is higher in north and south,lower in the middle,western is higher than the eastern and higher in west and east,lower in the middle.Although tuberculosis is sporadic throughout the year,tuberculosis in Shanxi Province still has certain seasonal characteristics from 2008 to 2015,it mainly occurs from January to July,and a distinct seasonal rise in spring,which is the peak of reported incidence.2.The results of global autocorrelation showed that there is significantly positive spatial autocorrelation of tuberculosis’ reported incidence in 119 counties(districts)of Shanxi Province from 2008 to 2015,which means aggregative.The results of local spatial autocorrelation detected 24 hotspots and 27 cold spots during 8 years.The hotspots mainly include Guangling,Jingle,Kelan,Loufan,Xiangfen,Yicheng,Quwo,Houma,Jiangxian,Lanxian and Lingqiu.The counties(districts)above are areas with high reported incidence of tuberculosis and should be paid more attention.3.The results of spatial-temporal scanning analysis showed that there are 7 significantly clusters from 2008 to 2012,including the most likely cluster and other 6 secondary likely clusters.The most likely cluster(LLR=417.71,P<0.0001)includes 23 counties(districts)and covers all counties(districts)of Yuncheng in southwest of Shanxi Province and over half of counties(districts)of Linfen.The center of cluster is Linyi,radius of the most likely cluster is 149.37 kilometers.Time frame from September 1st 2008 to August 31 st 2010 and the relative risk is 1.31.Secondary likely clusters were: a.The cluster with a center Datong Mining area(LLR=179.39,RR=1.82);b.The cluster with a center Jiaocheng(LLR=168.05,RR=1.54);c.The cluster with a center Hequ(LLR=120.55,RR=1.45);d.The cluster with a center Guangling(LLR=106.07,RR=1.43);e.The cluster with a center Daixian(LLR=97.51,RR=3.25);f.The cluster with a center Zuoquan(LLR=58.66,RR=2.22).4.The results of geo-detector showed that Real GDP per capita and population density are negatively correlated with the reported incidence of tuberculosis(P<0.001).There is interaction between two factors,the combined effect of Real GDP per capita and population density on tuberculosis is higher than the sum of their respective independence which is enhancement effect.The better economic level and larger population density the lower reported incidence of tuberculosis.Conclusion:In this study,spatial-temporal statistical analysis and geo-detector were used to study the tuberculosis surveillance data of Shanxi Province from 2008 to 2015 in an ecological perspective.The distribution of tuberculosis in Shanxi Province presents clustering in time,space and space-time.The peak incidence of tuberculosis is spring.The hotspots mainly concentrated in the north,north central,midwest and southwest of Shanxi Province.O ne of the clusters with the largest relative risk is 3.25,the center of the cluster is Daixian.The largest scope is the cluster which center locates on Linyi County,the radius of cluster is 149.37 kilometers,the relative risk is 1.31.The Real GDP per capita and population density have interaction,which is enhancement.The better economic level and larger population density the lower reported incidence of tuberculosis.This study can provide a scientific and rational basis for the accurate decision-making of regional prevention and control of pulmonary tuberculosis in Shanxi Province,and it also can provide analysis ideas and methodological reference for the investigation of the influences of the disease.