江西省乙型肝炎发病趋势的时间序列和预测模型分析  

Analysis on the time series and prediction model in forecasting the incidence of hepatitis B in Jiangxi

作　　者：毛向群[1] 熊小庆[1] 涂秋凤[1] 余平[1] 杨健平[2] 

机构地区：[1]江西省疾病预防控制中心,江西南昌330029 [2]江西省赣州市疾病预防控制中心

出　　处：《中国预防医学杂志》2013年第6期435-438,共4页Chinese Preventive Medicine

基　　金：江西省科技厅科技支撑计划(2009BSA11800)

摘　　要：目的利用乘积季节模型预测江西省乙型肝炎(乙肝)的发病趋势,为江西省乙肝预警预测奠定基础,同时为制定防控措施提供科学依据。方法利用最小二乘原理,应用自回归求和移动平均模型与随机季节模型相结合的乘积季节模型,对江西省1990-2009年乙肝月发病数进行时间序列分析并建立预测模型,用2010年相应数据验证预测效果,并对2010年以后江西乙肝发病趋势进行预测。结果利用1990-2009年资料构建ARIMA(1,1,1)(0,1,1)12模型,所建立的预测效果良好,实际值均在预测值95%可信区间内,预测2011-2014年江西省乙肝发病人数呈上升趋势。结论采用ARIMA乘积季节模型预测江西省乙肝发病情况,拟合及近期预测效果均较好。预测效果符合江西省乙肝发病现状及目前采取的乙肝防治措施。Objective To forecast the incidence of hepatitis B in Jiangxi province by the application of product seasonal model, in order to lay the foundation for hepatitis B early warning and prediction, and to provide sci entific basis for hepatitis B control. Methods Based on the principle of least square, a product seasonal model which coupled the Auto Regressive Integrated Moving Average （ARIMA） with the stochastic seasonal model was applied to analyze the time series of monthly number of hepatitis B cases from 1990 to 2009 in Jian gxi and to establish a predicting model, which was further verified for the predictive effect by the correspond ing data of 2010 and was used to forecast hepatitis B incidence after 2010. Results The prediction model established based on the data from 1990 to 2009 was proved to be effective in forecasting the incidence of hepatitis B, since the actual values were in the 95% confidence interval of predicted values, and it also predicted the upward trend of hepatitis B incidence during 2011 to 2014. Conclusions A product seasonal model is suitable for the forecasting incidence of hepatitis B in Jiangxi province, and the predictive effect fits the current hepati tis B status and current prevention measures.

关 键 词：乙型肝炎 时间序列分析 乘积季节模型 预测 

分 类 号：R181.3[医药卫生—流行病学]