2010-2014年度辽宁省阜新市流行性感冒监测结果分析

Surveillance for influenza in Fuxin,Liaoning,2010-2014

目的了解2010 2014年度辽宁省阜新市的流行性感冒（流感）流行趋势及病原学分布规律,预测流感样病例（influenza-like illness,ILI）短期趋势,为流感的防控提供参考依据。方法对2010 2014年ILI监测哨点医院的ILI和病原学监测结果进行流行病学分析,利用166周ILI%时间序列数据构建自回归移动平均模型ARIMA（1,0,0）。结果 4年间阜新市ILI%峰值均出现在12月至次年2月。4年间ILI%在月份和年龄组分布上不同。4年阜新市共分离到流感毒株116株,季节性H3N2型所占比例较大,4年间分离到的毒株几乎分布于12月至次年2月间,ILI周报告数与阳性检出率呈正相关（rs=0.423,P〈0.05）。ILI%资料构建的ARIMA（1,0,0）模型,残差分析统计量经检验差异无统计学意义（Ljung-Box Q=19.948,P=0.277）,提示残差为白噪声,实际值与预测值的相对误差平均值为0.804%。结论阜新市流感流行呈现冬春季高峰,0～4岁组为流感的高危人群,其次为14岁以下儿童,4个监测年度流感病毒流行型别发生了变化,ILI高峰与阳性毒株检出率联系密切。ARIMA（1,0,0）模型对阜新市ILI的发病趋势模拟效果较好,适宜做短期预测。

Objective To understand the epidemiological and etiological characteristics of influenza in Fuxin, Liaoning province, from 2010 to 2014, predict the short term trend of influenza like illness （ILI） and provide evidence for the prevention and control of influenza. Method Descriptive epidemiological analysis was conducted on the annual surveillance data of ILI and pathogens collected from sentinel hospitals in Fuxin from 2010 to 2014. ARIMA （ 1,0, 0） model was established by using time series data of ILI proportion in the 164 weeks. Results The incidence peak of ILI occurred during December-February during this period, and age specific ILI incidence varied. A total of 116 strains of influenza viruses were isolated, seasonal influenza A （H3N2） virus was predominant during this period. The viruses were mainly isolated during December-February. Weekly reported ILl case number was positively correlated with positive detection rate of influenza viruses （ r, = 0. 423, P 〈 0. 05 ）. The residual difference by ARIMA （ l, 0, 0 ） model had no statistical significance （ Ljung-Box Q = 19. 948, P = 0. 277 ）, indicating the residual was white noise. The average of the relative error between actual value and predicted value was 0. 804%. Conclusion The annual incidence peak of influenza was during winter-spring in Fuxin. Age group 0 -4 years was at high risk, followed by age group 〈 14 years. The predominant pathogens varied with year. The ILI incidence peak was closely correlated with the positive detection of influenza viruses. ARIMA model can well simulate the incidence trend of influenza-like illness in Fuxin, which is suitable for the short-term prediction.