四川省2008-2018年流行性乙型脑炎流行特征及ARIMA模型应用

Analysis of epidemiological characteristics of Japanese encephalitis in Sichuan Province from 2008 to 2018 and application of autoregressive integrated moving average model

目的描述四川省2008-2018年流行乙型脑炎(Japanese encephalitis,JE)流行病特征,分析其变化规律,并构建自回归移动平均(autoregressive integrated moving average,ARIMA)模型探讨该模型在预测JE发病趋势中的应用。方法采用描述流行病学分析2008-2018年四川省JE流行概况,利用2008年1月-2017年12月四川省JE分月监测资料拟合ARIMA模型,并应用2018年1-12月报告发病数进行模型检验。结果四川省2008-2018年JE疫情呈下降趋势,全省疫情主要集中于川东和川南,发病高峰为每年7-9月,儿童为高危人群,但近年来青少年及成人发病有上升趋势。ARIMA(1,0,0)(2,1,0)12能较好拟合JE发病时间序列趋势。结论构建ARIMA模型可用于四川省JE疫情报告发病数的短期预测。

Objective The aim is to describe the epidemiological characteristics of Japanese encephalitis( JE) in Sichuan Province from 2008 to 2018,to build time series autoregressive integrated moving average ( ARIMA) model,and to discuss the model application in the prediction of JE incidence trends.Methods Descriptive epidemiological analysis was used to analyze the epidemic situation of JE in Sichuan Province from 2008 to 2018.Monthly surveillance data of JE in Sichuan Province from January 2008 to December 2017 were used to fit ARIMA model.The number of reported cases from January to December in 2018 was used to test the model.Results The epidemic situation of JE in Sichuan Province from 2008 to 2018 showed a downward trend,and eastern and southern areas were the highly prevalent areas. The incidence peak was from July to September every year.Children were the high-risk group,but the incidence of adolescent and adult was on the rise in recent years.ARIMA( 1,0 ,0 )( 2,1 ,0 ) 12 could appropriately fit the time series.Conclusion ARIMA model can be used for short-term prediction of the reported incidence of JE in Sichuan Province.