时间序列分析法在贵州省乙型病毒性肝炎流行趋势预测中的应用

Application of time series analysis method in predicting the epidemic trend of Hepatitis B in Guizhou Province

目的建立贵州省乙型病毒性肝炎(乙肝)发病的预测模型,预测2017—2019年乙肝发病趋势。方法对2004—2016年贵州省的乙肝月报告发病率建立自回归移动平均(ARIMA)预测模型,对贵州省2017—2019年乙肝发病趋势进行预测。结果 2004—2016年贵州省乙肝发病呈周期性波动,并具有长期趋势,拟合得到ARIMA(0,1,1)(0,1,1)12模型,用模型拟合2004—2016年乙肝月报告发病率,预测值和实际值平均相对误差为7.46%,预测2017—2019年贵州省乙肝月报告发病率在3.27/10万~4.38/10万的范围内波动。结论 ARIMA模型可较好的拟合贵州省乙肝发病在时间序列上的变化趋势,该模型可用于贵州省乙肝发病的短期趋势预测。

[Objective] To establish a predictive model of hepatitis B in Guizhou, and predict the incidence trend of hepatitis B from 2017-2019. [Methods] The ARIMA prediction model of monthly incidence rate of hepatitis B in Guizhou Province from 2004-2016 was established, and the incidence trend of hepatitis B from 2017-2019 was predicted. [Results] The incidence of hepatitis B in Guizhou Province from 2004-2016 fluctuated periodically and had a long-term trend. The model of ARIMA（0,1,1 ） （0,1,1）12was established, the monthly incidence rate of hepatitis B from 2004-2016 was fitted with the model and the average relative error between the predicted value and actual value was 7.46%. It is predicted that the incidence rate of hepatitis B fluctuates in the range of 3.27/lakh to 4.38/lakh from 2017-2019. [Conclusion] The ARIMA model can well fit the time series trend of hepatitis B in Guizhou Province, and be used to predict the short term trend of hepatitis B incidence.