Epidemiological Characteristics And Time Series Prediction Of Scarlet Fever In Xinjiang

Objective:To understand the epidemiological characteristics of scarlet fever in Xinjiang from 2010 to 2017 and predict its incidence trend,so as to provide evidence for further prevention and control of scarlet fever and related research.Methods:The data of scarlet fever incidence in Ili Kazakh Autonomous Prefecture（Ili）and Bayingolin Mongol Autonomous Prefecture（Bazhou）of Xinjiang from 2010 to 2017 were selected and analyzed descriptively.Meanwhile,the annual incidence of scarlet fever in Xinjiang and China was compared dynamically.The monthly scarlet fever cases in Xinjiang from 2010 to 2016 was used as the training set and the monthly scarlet fever cases in 2017 was used as the test set.The ARIMA product seasonal model was constructed by the training set,and then verified by the test set.Results:From 2010 to 2017,2345 cases of scarlet fever were reported in Ili.The highest cases of incidence were 420 in 2017,and the lowest cases of incidence were 193 in 2010.The highest incidence rate was 15.05/100,000 in 2017,and the lowest incidence rate was 6.87/100,000 in2010.The incidence peak was occurred in April to July and September to December in each year,taking the proportion of the whole yearly reported cases’78.00%and above.The 1 to 10 years old group took the most part in age distribution of scarlet fever.A total of 2061 cases（87.89%）were reported in this group,and the group’s sex ratio of males to females was 1.5:1.The population classification characteristics were mainly children in kindergartens,students,and scattered children,with a cumulative 2298 cases（98.00%）of reported incidence.Yining City had the highest periodically cumulative cases of reported incidence,accounting for 57.19%of the cumulative cases of reported incidence in the whole prefecture.Yining City also had the highest yearly average incidence of 33.82/100,000.From 2010 to 2017,259 cases of scarlet fever were reported in Bazhou.The highest cases of incidence were 99 in 2017,and the lowest cases of incidence were 15 in 2013.The highest incidence rate was 7.00/100,000 in 2017,and the lowest incidence rate was 1.14/100,000 in 2010.The incidence peak was occurred in April to June and September to December in each year.The 1 to 10 years old group took the most part in age distribution of scarlet fever.A total of 240 cases（92.66%）were reported in this group,and the group’s sex ratio of males to females was 1.2:1.The population classification characteristics were mainly children in kindergartens,students,and scattered children,with a cumulative 256cases（98.84%）of reported incidence.Korla City had the highest periodically cumulative cases of reported incidence,accounting for 48.62%of the cumulative cases of reported incidence in the whole prefecture.Ruoqiang County had the highest yearly average incidence of9.70/100,000.The optimal model for predicting the number of scarlet fever cases in Xinjiang was ARIMA（1,1,1）（0,1,1）₁₂.The average absolute percentage error of the training set of the model was 34.81,and the MAPE of the test set was 20.52.The results showed that the model was effective in predicting the incidence of scarlet fever in Xinjiang.Therefore,this model was selected to predict the number of scarlet fever cases in Xinjiang in 2018.Conclusion:The incidence of scarlet fever was on the rise in Xinjiang.ARIMA product seasonal model can be used for early warning of scarlet fever in Xinjiang.It was suggested that prevention and control measures should be strengthened in schools and kindergartens to effectively control the epidemic of scarlet fever.