| Objective:The purpose of this study is to describe and analyze the incidence of notifiable infectious diseases in 15 poverty counties of Jilin Province from 2008 to 2018.And this study analyzed the epidemic characteristics of notifiable infectious diseases at different times,different regions,and different populations,and the time trend of reported incidence.This study used different models to predict the incidence of notifiable infectious diseases and to evaluate the model fitting effect to get the best model.This study provided scientific epidemiological references for the formulation of scientific public health policies for the prevention and control of notifiable infectious diseases in 15 poverty counties in Jilin Province.Methods:The data of the notifiable infectious diseases in this study were from the Center for Disease Control and Prevention of Jilin Province.The population data of 15 poverty counties in Jilin Province come from China's basic information system for disease prevention and control.Descriptive epidemiological methods were used in this study.The IBM SPSS software(version 24.0)was used to analyze the epidemic characteristics of notifiable infectious diseases in 15 poverty counties of Jilin Province from 2008 to 2018,including the incidence,incidence order and composition ratio of infectious diseases in different years,regions and populations.Using Joinpoint regression software(version 4.7.0.0)to establish a joinpoint regression model to describe the time trend of reported incidence,including the annual percentage change rate of reported incidence and the average annual percentage change rate of age-specific reported incidence of men,women and the total population.The prediction of the monthly incidence of notifiable infectious diseases in 15 poverty counties of Jilin Province from 2008 to 2017 is realized by using the Eviews 8.0 software to build the SARIMA model,Matlab 2016 a to build the BP neural network model and NAR neural network model,SARIMA-BPNN combination model and SARIMA-NARNN combination model.Compare and evaluate the fitting effect of the five models,and get the best model.Results:1.In 2008-2018,a total of 155215 reported notifiable infectious diseases were reported including 110526 cases of Class B infectious diseases and 44689 cases of Class C infectious diseases in 15 poverty counties in Jilin Province.Respiratory tract diseases(53142 cases)were the most,accounting for 34.24%.The top 5 most frequently reported diseases are tuberculosis B,hand,foot and mouth disease,hepatitis B,brucellosis,and hepatitis C,with a total of 120524 cases,accounting for 77.65% of all reported infectious diseases.The disease with the largest number of cases in infants,young children and adolescents is hand-foot-mouth disease.The diseases with the largest number of cases in the youth,middle-aged and elderly populations are viral hepatitis and tuberculosis.Among them,97389 were male(62.74%),57826 were female(37.26%),and the ratio of male to female cases was 1.68: 1.Overall,the cumulative age-specific incidence rate for the age group of 0-9 years is the highest,the rate is 93.51 / 100000.The key groups for prevention and control of infectious diseases are children,students and farmers.Notifiable infectious diseases have obvious seasonality,with a peak incidence from May to August,and account for 41.23% of all cases.The highest incidence was in Taobei District(26124 cases,16.83%),and the lowest incidence was in Changbai County(1921 cases,1.24%).2.From 2010 to 2018,in 15 poverty counties in Jilin Province,the standardized incidence of the total population decreased at a rate of 10.67% annual change(P <0.05),and the standardized incidence of the male population decreased at a rate of 11.00% annual change(P <0.05),and the standardized incidence of female population declines at a rate of 10.17% annual percentage change(P <0.05).The normalized incidence of the total population,men,and women remained stable from 2008 to 2010.In the total population,among the age group of 5-49,55-59 and 70-79,the average annual percentage change rate of the age-specific incidence rate of other groups showed a downward trend(P <0.05).Among men,the age group of 30-34 and 70-74 decreased the slowest,with the average annual percentage change rate of 6.4% and 5.8%.Among women,the age group of 25-29,45-49,and 55-59 decreased the slowest,with the average annual percentage change rate of 6.5%,6.3%,and 6.4%.3.Among the five models used in this research,the model with the best prediction effect is the SARIMA-NARNN model,followed by NARNN model,SARIMA-BPNN model,BPNN model,and SARIMA model.In the SARIMA-NARNN model,the RMSE value is 0.17,the MAE value is 0.14,and the MAPE value is 8.13%.The values of the three evaluation indicators are the smallest.The SARIMA-NARNN model is the optimal model,and the predicted monthly incidence(/ 10000)in 2018 is 1.63,1.63,1.71 1.46,1.69,1.58,1.75,1.65,1.59,1.43,1.63,1.63,respectively.Conclusions:1.In 2008-2018,a total of 155215 reported notifiable infectious diseases were reported in 15 poverty counties in Jilin Province.The top 5 most frequently reported diseases are tuberculosis B,hand,foot and mouth disease,hepatitis B,brucellosis,and hepatitis C.The disease with the largest number of cases in infants,young children and adolescents is hand-foot-mouth disease.The diseases with the largest number of cases in the youth,middle-aged and elderly populations are viral hepatitis and tuberculosis.The incidence is higher in men than in women.Overall,the cumulative age-specific incidence rate for the age group of 0-9 years is the highest.The key groups for prevention and control of infectious diseases are children,students and farmers.Notifiable infectious diseases have a seasonal trend,with a peak incidence from May to August.The area with the highest incidence was Taobei District.2.The incidence of reported notifiable infectious diseases showed a downward trend from 2010 to 2018,and remained stable from 2008 to 2010,in 15 poverty counties in Jilin Province.3.The SARIMA model,BPNN model,NARNN model,SARIMA-BPNN model,and SARIMA-NARNN model can predict and fit the reported monthly incidence of notifiable infectious diseases in 15 poverty counties in Jilin Province.But the SARIMA-NARNN combination model is the optimal model. |
PDF Full Text Request