| Objectives:We took Yichang city of Hubei province as a typical medium-sized city in the subtropical region of central China,analyzed the epidemiological characteristics of seasonal influenza in Yichang,observed the surveillance consistency between influenza-like illness(ILI)identified by doctors and acute respiratory infections(ARI)identified by International Classification of Diseases(ICD)diagnosis code,evaluated the accuracy of different methods for identifying ILI cases,support the development of vaccination program,and provide evidence and method for improving influenza surveillance.Methods:Collect the virology test result and medical visit information of influenza-like illness(ILI)surveillance in outpatient emergency department of Yichang Central Hospital and Yichang Second People’s Hospital in 2013-2019,and medical visit information of ARI cases(ICD-10 code range:J00-J22)surveillance at the outpatient emergency department of Yichang Central Hospital from April to October 2019.Based on the virology test result of ILI surveillance in 2013-2019,the laboratory-confirmed influenza positive rate was used as an indicator to assess seasonal characteristics of influenza,including periodicity,duration,peak value,and predominant subtypes/lineages.The epidemiological characteristics of influenza in different age groups were analyzed,and the composition of influenza B virus in the trivalent influenza vaccine recommended by the northern hemisphere was compared with the predominant lineages in Yichang.The consistency between the data of ILI surveillance in outpatient emergency department and the data of ARI surveillance through ICD-10 code were observed and analyzed.To evaluate the computer-assissted surveillance method,we select cases with complete medical record among surveillance data of ARI cases identified through ICD-10 code(range:J00-J22),using the manually classification result of these cases based on medical charts according to the definition of influenza-like illness as the gold standard,draw the receiver operating characteristic(ROC)curve,calculate the sensitivity,specificity,diagnosis consistency rate and other indicators,and compare the accuracy of ILI surveillance by computer automatic identification and doctor report.Results:A total of 8693 ILI cases were sampled in the two sentinel hospitals in the year of 2013-2019,1439(16.55%)were laboratory-confirmed influenza cases,and the influenza A positive rate(10.61%)was higher than the influenza B(5.95%).There were three influenza circulation patterns in Yichang:(1)annual periodicity(in 2013-2014,2015-2016 and 2018-2019),(2)semiannual periodicity(in 2014-2015),and(3)year-round periodicity(in 2016-2017 and 2017-2018).Summer epidemics existed in two of the six years and were dominated by influenza A/H3N2.Winter and spring epidemics occurred in five of the six years,and A/H1N1,A/H3N2,B/Victoria,and B/Yamagata were codominant.During the study period,children who aged during 5-14 years old(26.44%)and individuals over 60 years old(16.87%)had the highest influenza positive rates,the lowest rate was found in children aged less than 5 years(12.07%).For influenza A virus,the positive rate of A/H3N2(7.57%)was much higher than that of A/H1N1(3.96%)among elderly people over 60 years old.For influenza B virus,B/Victoria was rarely detected among people≥40 years old(0.27%).However,the positive rate of B/Yamagata increased in those 30 years old and finally peaked among 70-74year-olds(5.88%).Influenza B was one of the predominant lineages in five of the six surveillance years,the predominant lineages,B/Victoria in 2015-2016 and B/Yamagata in 2017-2018,were both mismatched with the influenza B component of the trivalent vaccine.Analysis based on ICD-10 diagnostic code data showed that there were a total of 9058 ARI cases from April to October 2019 in Yichang Central Hospital,Hubei Province,the median age of ARI was 30(IQR:24-42)years old and the proportion of ARI patients in clinics was highest in April and May(19.88%and 16.29%,respectively).The number of ILI cases shows that there was a summer peak for influenza,while ARI has no significant peak in summer.A total of 3542 ARI cases with complete symptom information were included in the evaluation of the new influenza-like illness surveillance method.Among them,1179 cases(33.29%)met ILI definition criteria,doctors reported 1306 ILI cases(36.87%),and the computer automatically recognized 1150 ILI cases(32.47%).The consistency of kappa values with gold standard judgment of automatic computer analysis and doctor report were 0.97 and 0.66 as well as the area under the ROC curve(AUC)was 0.98 and 0.84,respectively.And the sensitivity and specificity of automatic computer analysis were higher than that of doctor’s report(P values were all<0.001),the sensitivity was 96.95%and 82.27%,and the specificity was99.70%and 85.78%,respectively.Conclusions:(1)There were three epidemic patterns of seasonal influenza in Yichang City,annual periodicity,semi-annual periodicity,and year-round periodicity,which lack regularity.The predominant subtypes/lineages were complex,A/H3N2 dominated in summer,and multiple subtypes/lineages could codominanted in winter and spring.There were differences in the prevalence among different age groups.In some years,the predominant lineages were mismatched with the influenza B component of the trivalent vaccine.Considering adjusting the vaccination time,giving priority to vaccinating school-aged children and the elderly,and promoting the use of quadrivalent influenza vaccines may help.(2)When using influenza-like illness or acute respiratory infection data to warn the influenza epidemic trends should pay attention to the source of the data and be alert to abnormal fluctuations.ILI and ARI have different levels in summer,so it should be prudent to adopt ARI as a surveillance indicator of influenza activity.(3)The automatic computer analysis based on electronic medical records can identified ILI cases with good sensitivity and specificity in ILI case surveillance.For influenza surveillance sentinel hospitals with complete electronic medical records,ILI case surveillance can be carried out by means of automatic computer analysis. |
PDF Full Text Request