ARIMA模型在流感样病例预测预警中的应用

目的探讨ARIMA模型在流感样病例预测预警方面的应用,建立流感样病例发病预测模型,并证明模型的适用性。方法对北京市顺义区医院、顺义区妇幼老年保健院2家省级流感样病例监测哨点医院报告的2005年9月～2009年3月流感样病例月报告数资料建立ARIMA模型,2009年4～5月数据验证模型,用Q统计量法对模型适应性进行检验。结果对流感样病例月报告数建立季节模型ARIMA(1,0,0)x(0,1,0)12,统计量Q大于Χ2α(m)证实了该模型的适用性。结论ARIMA模型能够较好应用于流感样病例预测预警,为疫情防控提供科学依据。

Application of ARIMA model in forecasting and early warning of influenza-like illness

Objective To explore application of auto-regression integrated moving average （ARIMA） model in predicting and early warning of influenza-like illness （ILI ） and establish a model to predict incidence of ILI and verify its applicability. Methods An ARIMA model was established based on the data of reported cases of ILI collected from surveillance sentinels at Shunyi District Hospital and Shuhn Yi District Maternity and Children＇s Hospital during September 2005 to March 2009. Data of ILI cases in April and May 2009 were used to verify its applicability, with Q-statistic using SPSS software. Results The established ARIMA model （1, 0, 0） × （0,1,0） 12 showed Q-statistic was greater than xα2 （m）, indicating applicability of the model. Conclusion The ARIMA model can be applied to forecasting and early warning of cases of ILI and their dynamic trend, providing scientific evidence for prevention and control of influence epidemic.