Application of spatial data analysis on clustering of bacillary dysentery in Gansu

Objective To understand the spatial distribution of bacillary dysentery cases in Gansu in 2013 and its spatial autocorrelation and clustering areas. Methods The incidence data of bacillary dysentery in 87 counties in Gansu in 2013 were collected from National Disease Reporting Information System to analyze the spatial autocorrelation by using Geoda 1.60 and conduct spatial scan statistics by using SaTScan 9.1.1.0. The results were visualized by using ArcGIS 10. 0 software. Results In 2013, 8191 bacillary dysentery cases were reported, the incidence was 31.81/100 000. The bacillary dysentery distribution showed a spatial autocorrelation in all the study areas (Moran's I=0.4555,Z=6.51,P=0.001). By local spatial autocorrelation analysis, the highly autocorrelation of bacillary dysentery cases was observed in Qingyang in eastern Gansu and 9 countries in southern Gansu, the hot spot areas, and the low autocorrelation of bacillary dysentery cases was observed in 11 counties in west-central Gansu, the cold spot areas. The spatial scan detected the major clustering areas in 5 counties near Lanzhou(LLR=137.10,RR =2.38)and in 9 counties in Qingyang(LLR=428.60,RR=2.40). Conclusion The bacillary dysentery cases were not distributed randomly in Gansu in 2013, the spatial autocorrelation and obvious clustering were observed.