Epidemiological Characteristics Of Human Brucellosis And The Prediction Research In Liaoning,2006-2016

Objective:1.To describe the epidemiological characteristics of human brucellosis from2006 to 2016 from the part of population distribution and spatial and temporal distribution,by collecting human brucellosis surveillance data.2.Based on the time trend and seasonal trend of data,the ARIMA model was constructed to explore its application effect in predicting the human brucellosis incidence in Liaoning province.Methods:1.The surveillance data of human brucellosis were collected and analyzed according to gender,age,occupation,region and year by using SPSS 20.0 to describe the characteristics of human brucellosis surveillance data in Liaoning province;2.Establish an ARIMA model by using SPSS 20.0.The steps included sequence stabilization,model identification,parameter estimation and model diagnosis and prediction.A time series was established from January 2012 to December 2016,and the incidence of January to June 2017 was predicted in the short term,which was compared with the actual number of reported cases,and the prediction effect was discussed.Results:1.A total number of 15314 cases was collected in Liaoning province 2006-2016,and the peak year was 2015 with a total of 2948 cases.Cases were mainly concentrated in the age group of 41 to 50 and 51 to 60 years old,the incidence of male was more,the sex ratio was 3.17:1.In addition,farmers accounted for 78.5%of all reported cases,followed by the housework and unemployed people?including retirees?.Incidence areas concentrated in the cities of huludao,jinzhou,shenyang,fuxin and chaoyang.The cases were mainly concentrated in spring and summer from March to July of each year,followed by February and August,and there was an obvious seasonality.2.After the logarithmic transformation and difference,the sequence diagram was leveled off,and then four kinds of white noise sequence models were filtered through comparing the BIC which to determine the optimal model of ARIMA?0,1,0??1,1,0?₁₂.Applying the model in Liaoning province by using surveillance data from 2012 to 2016 to predict January to June 2017.The results showed that the observed value and the fitting values were within the predictive value of 95%confidence interval,prediction effect was good,and the accuracy was high relatively.Conclusion:1.There was a rising tendency in the prevalence of human brucellosis,and there existed seasonality from March to July.Besides,farmers in semi-rural and semi-pastoral areas were high risk groups.2.The cities of huludao,jinzhou,shenyang,fuxin and chaoyang were the high risk areas of the epidemic disease in Liaoning province,which was of great significance to study the risk factors and intervention measures of the disease.3.The ARIMA model has a good effect on short-term prediction of human diseases for its temporal and seasonal trends.