移动流行区间法在北京市朝阳区流行性感冒流行阈值估计及分级中的应用  

Application of the moving epidemic method in epidemic thresholds and warning alert for influenza prevention in Chaoyang district of Beijing,China

作　　者：黄立勇[1] 唐栩莲 李刚[3] 苏微 王海滨[1] 张士尧[1] 仲一 张政[1] 马建新[1] HUANG Li-yong;TANG Xu-lian;LI Gang;SU Wei;WANG Hai-bin;ZHANG Shi-yao;ZHONG Yi;ZHANG Zheng;MA Jian-xin(Cenler for Disease Control and Prevention of Chaoyang District,Beijing 100021,China;不详)

机构地区：[1]北京市朝阳区疾病预防控制中心,北京100021 [2]河北医科大学公共卫生学院,河北石家庄050017 [3]北京市疾病预防控制中心,北京100013

出　　处：《中国病毒病杂志》2022年第3期204-208,共5页Chinese Journal of Viral Diseases

基　　金：首都卫生发展科研专项(首发2021-1G-3013)。

摘　　要：目的 评估移动流行区间法(moving epidemic method,MEM)在北京市朝阳区不同流行性感冒(流感)监测资料分类条件下的流行阈值及强度分级中的应用效果。方法 选择2013—2019年朝阳区流感监测哨点医院上报的流感样病例(influenza-like illness,ILI)监测数据,采用MEM判定各流行季的流行阈值及分级强度阈值,采用交叉验证法对模型的拟合优度进行评价,确定模型的最优参数,并依次预测2019—2020年流行季流行阈值。以2019—2020年3家流感国家级病原学监测哨点医院核酸检测阳性率(positive rate,PR)>峰值40%为金标准,判定2019—2020年流行期与非流行期,与MEM预测的流行开始和结束周次进行比较。结果 以26家哨点医院发热门诊上报的ILI数据纳入建模,参数δ设置为2.4时,模型拟合效果最佳,灵敏度、特异度和约登指数依次为0.82、0.92和0.74。MEM预测显示2019—2020年流行季于第50周进入流感流行期,最高峰达到中等强度流行水平。与利用PR>40%的金标准判定2019—2020年的结果(第50周进入流行期,第6周流行结束)完全吻合。结论 MEM可用来评估北京市朝阳区流感流行强度。Objective To evaluate the moving epidemic method(MEM) in the estimation of influenza epidemic thresholds and intensity grading of different influenza surveillance data in Chaoyang district of Beijing.Methods Influenza-like illness(ILI) surveillance data reported from fever clinics of 26 sentinel hospitals in Chao-yang district from 2013 to 2019 were collected.Weekly influenza surveillance data with different classification were used for cross-validation to determine the optimal parameters of the model as well as the optimal data type by comparing the goodness of fit of the model using cross-validation,including sensitivity,specificity and Youden index.The weekly reported influenza virus nucleic acid test positive rate(PR,hereafter all referred to as PR) at the three surveillance sentinel hospitals in 2019—2020 was used as the baseline,PR>40% was used as the gold standard to determine the epidemic and non-epidemic periods in 2019—2020 comparing with the epidemic start and end weeks predicted by MEM.Results ILI data reported from fever outpatient clinics at the 26 influenza surveillance sentinel hospitals was used for inclusion in the modeling,and the model performed best when parameter δ was set to 2.4.The overall sensitivity,specificity,and Youden index were 0.82,0.92,and 0.74,respectively.The predicted early warning 2019—2020epidemic season entered the influenza epidemic at week 50and peaked at moderate intensity epidemic levels thereafter.The MEM predictions were fully consistent with the results using the gold standard of PR>40%(entering the epidemic at week 50and ending the epidemic at week 6). Conclusions MEM,as an influenza early warning model,can be used as an effective tool to assess the start of influenza epidemic and epidemic intensity.

关 键 词：流行性感冒 移动流行区间法 流行阈值 分级预警 

分 类 号：R181.2[医药卫生—流行病学]