应用SARIMA模型分析湖北省肺结核时间分布特征及发病趋势

发布时间：2018-04-01 08:40

本文选题：肺结核　切入点：SARIMA模型　出处：《现代预防医学》2017年03期

【摘要】：目的探讨季节性自回归积分滑动平均模型(Seasonal Autoregressive Integrated Moving Average Model,SARIMA)在分析湖北省肺结核报告病例时间分布特征中的应用,预测其发病趋势。方法利用全国"传染病报告信息管理系统"中2005年1月至2015年12月湖北省肺结核各月病例数,分析肺结核报告病例的时间分布特征,建立SARIMA预测模型进行趋势预测。结果肺结核报告病例数整体处于下降趋势且具有明显的季节特征,3-6月和12月为年周期的两个高峰期;在肺结核病例时间分布特征中拟合得到的SARIMA最佳预测模型为(1-B)(1-B12)xt=(1-0.732B)(1-0.409B~(12))μt,模型的显著性检验得到P0.05,显示拟合模型具有统计学意义。结论建立的SARIMA模型能较好地预测湖北省肺结核报告病例数的时间趋势,为疫情监测和疾病防治提供参考依据。
[Abstract]:Objective to explore the application of seasonal Autoregressive Integrated Moving Average model SARIMA in the analysis of the temporal distribution of reported tuberculosis cases in Hubei Province. Methods using the national infectious disease report information management system, the number of tuberculosis cases from January 2005 to December 2015 in Hubei Province was analyzed, and the time distribution characteristics of the reported cases were analyzed. SARIMA forecasting model was established to forecast the trend. Results the total number of reported cases of pulmonary tuberculosis was in a downward trend and had obvious seasonal characteristics: March to June and December were the two peak periods of the annual cycle. The best prediction model of SARIMA obtained by fitting the time distribution of pulmonary tuberculosis cases is 1-B ~ (12) B ~ (12) B ~ (1) ~ (12) 渭 t ~ (-1), and the significance test of the model is P0.05, which shows that the fitting model has statistical significance. Conclusion the established SARIMA model can better predict the lung of Hubei province. Time trends in the number of TB cases reported, To provide reference for epidemic surveillance and disease control.
【作者单位】：湖北省疾病预防控制中心;华中科技大学医学院;
【基金】：湖北省卫生计生科研基金资助(WJ2016JT-002)
【分类号】：R521;R181.3

【相似文献】