Analysis Of Epidemic Characteristics Of Chinese Rabies From 2007 To 2016 And Forecasting Of Epidemic Trends

Objective The prevalence and epidemic characteristics of rabies in China in 2007-2016were analyzed;based on the monitoring data of 2016,the treatment and host immunity was discussed.According to the historical data in the Chinese Disease Surveillance Information Report Management system,the time series model was used to predict the epidemic trend and the incidence of rabies in 2017.It provides the basis for developing the strategy and measures of rabies prevention and control in our country.Methods The 2007-2016 human rabies epidemic data from the China Disease Prevention and Control center,"Disease Surveillance information Reporting management system."The 2016 monitoring data came from 15 national monitoring points nationwide and 2provincial monitoring points in Anhui province.Use Excel 2013 software to collate and summarize data such as outbreak data and individual case investigations reported by each monitoring point,post-exposure outpatient treatment and animal monitoring information.（1）The epidemiological characteristics of rabies in China in 2007-2016 were described by three distributions,and the indexes of incidence,morbidity,mortality,and composition ratio were analyzed.（2）The data usage rate,composition ratio and the median number of the cases reported by each monitoring point,such as the case investigation,the outpatient treatment after exposure,and the monitoring information of the host material,are described.（3）Organize Rabies monthly incidence data from January 2007 to December 2015.The first step is model creation-using SPSS 19.0 to establish time series with year and month as variables,an ARIMA model was created through model identification.The second step is model selection—using the stationary R²,root means square error（RMSE）,and Normalized Bayesian Information Criterion with the goodness of fit statistics,and selecting alternative models.The third step is model verification—the use of alternative ARIMA models to predict the number of cases of Rabies in China in 2016,compared with the actual incidence data from January to December 2016.Relative error indicators were used to evaluate the prediction of the model and select the effect smallest the error.The model is the optimal model.The fourth step is model prediction—adjusting the model for monthly incidence data in 2016 to predict the number of rabies cases from January to December 2017.Results（1）China reported a total of 16 910 cases of rabies in 2007-2016 and 16 376deaths.From 2007 to 2016,the incidence of rabies in China decreased year by year,from3300 in 2007 to 644 in 2016.Geographical distribution:During 2007-2016,rabies cases were reported in 31 provinces of China,the high incidence of rabies was in Guangxi Zhuang（2461 cases）,Guangdong Province（1961 cases）,Guizhou（1804 cases）and Hunan（1482 cases）,which accounted for 45.58%of total incidence.Time distribution-seasonal patterns of rabies exist.In August,September,and October,the number of cases accounted for 31.12%（5 263/16 910）of the total cases,and the lowest number of cases in February accounted for 5.72%（968/16 910）.Population Distribution:there were 11 863 males and 5047 in the females incidence in 2007-2016.The ratio of male to female was 2.35:1.The age distribution was mainly concentrated in the 5-15-year-old and 40-69 age groups,accounting for 12.75%（2 156/16 91 0）and 54.74%（9 257/16 910）respectively.In the occupational distribution,the high population accounted for 69.28%of the farmers（11715/16 910）,followed by 11.23%（1 899/16 910）of the students.（2）2016 monitoring point monitoring results:In 2016,a total of 228 death cases questionnaires were collected,62.44%of cases（133/213）were not treated with wounds,and 29.58%（63 cases）were treated on their own,only 7.46%（17/228）has been treated by a medical institution.Of these,14 received post-exposure vaccination and three were injected with immunoglobulin.Rabies Clinics:there were 1,281,340 outpatients,935,869（73.04%）of the outpatients treated wounds,77.66%in the whole course of vaccination,and28.89%of the patients in the class III exposure who received passive immunization.The average density of dogs at the monitoring sites was 7/100,and the average immunization rate of dogs was 37.64%（1792 369/4 762 112）.（3）The 2007-2016 time series is a non-stationary sequence with long-term trends and seasonal fluctuations.Firstly,nine ARIMA models satisfying the conditions for the establishment of the model are created by model identification.Second,according to the smooth R2,the bigger the better,the smaller the standard BIC is,the better,and the smaller the root mean square error is,the better the principle is,choose ARIMA（0,1,1）（1,1,0）₁₂,ARIMA（2,1,0）（0,1,1）₁₂ and ARIMA（2,1,0）（2,1,0）₁₂ as alternative models.After the model is verified,ARIMA（0,1,1）（1,1,0）₁₂ is finally selected as the optimal model based on the principle that the relative error of the predicted value is minimum（2.33%）.Finally,after the 2016 data was added to ARIMA（0,1,1）（1,1,0）₁₂,an adjustment model was obtained,and the number of incidence in 2017 was predicted to be 535 cases.Conclusions In the 2007-2016,the incidence of rabies in China continued to decrease.The incidence of the provinces from 23 to 28,the high hair provinces in Guangxi,Guangdong,Guizhou and Hunan province,rabies all the year-round onset,summer and autumn high,the incidence of a male and peasant-dominated population.In 2016,the treatment rate of the wound outpatient department was 73.04%,the vaccination rate was 77.66%,and the immune rate was 37.64%,so we should further standardize and strengthen the exposure of the exposed people to improve the immunity of dogs.The ARIMA（0,1,1）（1,1,0）₁₂ model can fit the long-term and seasonal trends of rabies onset.The results of back-generation fitting and short-term prediction are ideal,and the number of incidence in 2017 is expected to decline to 535 cases.