论文部分内容阅读
目的探讨CUSUM模型(累积和)和EWMA模型(指数加权移动平均)探测流感流行起始的功效。方法利用北京市顺义区2013年9月1日-2015年4月30日流感样病例监测数据,调整不同参数采用CUSUM和EWMA模型对监测数据进行分析,与流感病原学结果进行比较,分析二种模型在不同参数组合下探测流感流行起始的效果。结果流感样病例监测显示,CUSUM模型当k=0.9和H=2σ时,根据金标准算探测结果恰好是最早流行起始时间,而EWMA模型探测流感流行起始时间当λ=0.7、k=3时与金标准最为接近。结论顺义区利用流感样病例监测结果,CUSUM模型和EWMA模型,均能及时准确地预警流感高峰的到来,但是仍然需要继续对模型参数进行拟合和分析。
Objective To explore the efficacy of CUSUM model (cumulative sum) and EWMA model (exponential weighted moving average) to detect the onset of influenza. Methods The surveillance data of flu-like cases in Shunyi District of Beijing from September 1, 2013 to April 30, 2015 were used to analyze the monitoring data by adjusting the parameters with CUSUM and EWMA models. The results were compared with the results of influenza etiology, The model detects the effect of the initial flu pandemic with different combinations of parameters. Results Influenza-like case surveillance showed that when the k = 0.9 and H = 2σ were used in the CUSUM model, the result of the gold standard calculation was just the earliest popular starting time, while the EWMA model was used to detect the initial time of influenza when λ = 0.7 and k = 3 When the gold standard and the closest. Conclusion The results of influenza-like illness monitoring, CUSUM and EWMA models in Shunyi District can timely and accurately predict the arrival of influenza peak. However, it is still necessary to continue fitting and analyzing the model parameters.