Spatial Statistics Of Tuberculosis Incidence In China And Epidemic Characteristics Analysis

Tuberculosis prevention and control has always been a public health issue of high concern in China and even the world.It is not only related to human life and health safety,but also related to national stability,social security and international trade.This makes tuberculosis one of the world’s most significant public issues.It is of great significance to make public health policy and tuberculosis prevention and control measures to analyze tuberculosis data and grasp its distribution and change rule.Firstly,we conducted a descriptive statistical analysis of TB-related epidemic data from 31 regions in China from 2011 to 2019.（1）The incidence of tuberculosis showed a downward trend,and the annual decline rate of tuberculosis in 18 regions exceeded the national average.（2）The number of tuberculosis patients is changing from young people to middle-aged and old people,and the number of tuberculosis patients among teenagers is decreasing.（3）The incidence of tuberculosis in males was significantly higher than that in females,with a male to female ratio of about 7:3.（4）The occupation distribution is mainly farmers（62%）,housework and unemployed（13%）.（5）Western China belongs to high risk area,central and southern China belongs to medium high risk area,and most coastal and north China belong to low risk area.Secondly,through spatial statistical analysis,this thesis found that most of the global Moran’s I of tuberculosis incidence were above 0.4 and showed a highly significant increasing trend,that is,areas with high tuberculosis incidence were often adjacent to other areas with high incidence.Areas with low tuberculosis incidence tend to be adjacent to areas with similar rates;From 2011 to 2019,the number of high-high-concentration areas in China decreased from 12 to 7,and the number of low-low-concentration areas increased from 15 to 17.Cluster areas are shrinking and spatial differentiation is becoming more and more obvious.The incidence rate is low in the eastern region,high in the southwestern region,and the central region shows a changing trend under the influence of east and west regions.The types of clusters converge towards their nearest regions.Finally,the time trajectory of the month of tuberculosis incidence and the incidence of tuberculosis in each region of China were analyzed by cluster analysis,both 12 months and 31 regions were divided into 4 categories.The prediction result of ARIMA-SVR combination model is better than that of ARIMA（1,0,0）（0,1,2）₁₂.The MSE of ARIMA-SVR combination model is more than two times of that of ARIMA（1,0,0）（0,1,2）₁₂.The MAPE of ARIMA-SVR combination model is 2.56%,while that of ARIMA is only 1.42%.Finally,the ARIMA-SVR combined model was used to predict and analyze the error of the total number of tuberculosis cases in 2020.The results showed that the error of predicting the total number of tuberculosis cases in the whole year was less than 1%.