中国乙类呼吸道传染病发病波动与长短期预测

Fluctuation analysis and long-term and short-term prediction of class B respiratory infectious diseases in China

目的分析中国乙类呼吸道传染病发病的波动规律并进行长期趋势预测和短期预测,为制定防控策略提供科学依据。方法利用CensusX-12季节调整法和Hodrick-Prescott(HP)滤波法对中国乙类呼吸道传染病发病的月度时间序列数据进行分解,将时间序列中不规则变动、季节因素、趋势因素和循环因素分离,研究其波动规律,同时结合回归模型和Holt-Winter季节指数平滑模型实现长期趋势预测和短期预测。结果乙类呼吸道传染病的发病情况受季节因素影响较大,呈现循环周期性波动,发病的长期趋势为逐年下降,同时利用Holt-Winter季节指数平滑模型取得了很好的短期预测效果。结论CensusX-12季节调整法和HP滤波法可较好的分析中国乙类呼吸道传染病发病的季节特征和循环周期特征,实现长期趋势预测和短期预测,对疾病防控策略的制定有指导意义。

Objective To analyze the fluctuation law of the incidence of class B respiratory infectious diseases in China,to make long-term trend prediction and short-term prediction,and to provide scientific basis for formulating prevention and control strategies.Methods Based on CensusX-12 seasonal adjustment method and the Hodrick-Prescott(HP)filtering method,the monthly time series data of the incidence of class B respiratory infectious diseases in China was separated.The irregular changes,seasonal factors,trend factors and circulation factors were separated in order to study its fluctuation rules.The long-term trend prediction and short-term prediction were realized by using the regression model and the Holt-Winter seasonal index smoothing model.Results The incidence of class B respiratory infections diseases was greatly affected by seasonal factors,and showing cyclical fluctuations.The long-term trend of the incidence was found to be decreasing year by year.Meanwhile,the Holt-Winter seasonal index smoothing model was used to obtain a good short-term prediction effect.Conclusions Censusx-12 seasonal adjustment method and HP filter method can be effectively used to analyze the seasonal characteristics and circulation cycle characteristics of class B respiratory infectious diseases in China,and achieve long-term trend prediction and short-term prediction.The results have guiding significance for the formulation of disease prevention and control strategies.