Study On The Epidemiological Characteristics Of Hand-foot-mouth Disease In Chongqing And The Construction Of Warning Model

ObjectiveTo analyze the epidemic situation of Hand-foot-mouth disease?HFMD?in Chongqing from 2009 to 2015,to master the epidemic of HFMD in Chongqing,and to provide a scientific basis for formulating disease-related prevention and control strategies.The single and combined warning model of HFMD epidemic in Chongqing City were established to evaluate and screen the optimal warning model for disease outbreaks.To provide a scientific prediction of the incidence of HFMD in Chongqing,an accurate warning of the disease outbreak in the relevant health industry sector,an early warning model for other infectious diseases,and a set of feasible ideas and framework.MethodsThe HFMD epidemic surveillance data and demographic data of Chongqing from 2009 to 2015 were collected and compiled.Descriptive epidemiological methods were used to analyze the HFMD epidemic situation,three distributions,common morbidity cases,severe cases and deaths,and etiological characteristics of Chongqing and other epidemiological features.The seasonal distribution of HFMD were explored by circular distribution method.The Arcgis10.2 and OpenGeoDa software experimental platform were used to analyze the spatio-temporal aggregation of HFMD.Based on the HFMD incidence data from 2009 to2014 in Chongqing?training samples?,three single warning models?SARIMA,BPNN,and Elman neural networks?and three combined warning models?SARIMA-BPNN,SARIMA-Elman,BPNN-Elman?were established.Baesd on the HFMD incidence data for 2015?test sample?,the above model fitting was performed,and the average absolute error?MAE?,average error rate?MPAE?,prediction accuracy?P?,nonlinear correlation coefficient?RNL?and average relative error?MRE?were used to evaluate and screen the optimal early warning.The model predicts the incidence of HFMD in Chongqing from 2016 to 2017.Results1.From 2009 to 2015,a total of 211,416 cases of HFMD were reported in Chongqing,with an average of 30,202 cases?9559-55338?per year.Theaverageincidenceratewas90.51/100000?29.18/100000-163.96/100000?.In the past seven years,a total of 128deaths and 546 severe cases have been reported.The duration of severe cases was 0182 days?median 2 days?.The duration of death was 0-22 days?median 3 days?.Among the severe and dead patients,the male ratio is higher.The age of critically ill patients is 1 year old and 2 years old.The death patients are mostly 23 years old.2.From 2009 to 2015,a total of 11,545 laboratory-tested cases were reported in Chongqing,of which 3,540 were positive for EV71,3,604 were positive for Cox A16,and 4,401 were positive for other intestinal viruses.The severe cases and deaths detected by the laboratory were mainly EV71pathogens.The difference in pathogen composition was statistically significant(^?2=510.697,P<0.001).3.From 2009 to 2015,Chongqing HFMD showed a seasonal trend(Z>Z_0.05,12,P<0.001).There was only one popular peak in 2009 and 2010,which was from April to July.There are two popular peaks from 2011 to2015,from April to July and from October to December.4.The HFMD reported cases were mainly children aged 5 and below,accounting for 91.71%of the total cases.There were 175,240 children aged3 and below,which accounted for 82.89%of the total reported cases.Occupation distribution,the incidence of people mainly scattered children,accounting for 63%of the total incidence.5.From 2009 to 2015,39 districts and counties in Chongqing had an annual incidence,and the incidence was spatially positively correlated.The high incidence area is mainly concentrated in the main urban area of Chongqing and its neighboring districts and counties.6.The HFMD single and combined early warning models constructed in this paper are:SARIMA model?1,0,0??0,1,0?₁₂,BPNN model?4-7-1?,Elman model?4-18-1-1?,SARIMA-BPNN?1-5-1?,SARIMA-Elman?1-6-1-1?,BPNN-Elman combination model?1-9-1-1?.The average absolute error?MAE?,average error rate?MPAE?,prediction accuracy?P?,nonlinear correlation coefficient?RNL?,and average relative error?MRE?of the six model fits and predictions are 0.595,0.515,0.485,0.551,0.493;0.265,0.229,0.771,0.673,0.029;0.234,0.202,0.798,0.737,0.142;0.248,0.215,0.785,0.714,0.101;0.328,0.284,0.716,0.620,0.103;0.229,0.198,0.802,0.745,0.071.7.Integrate the evaluation indicators and screen out the optimal early warning model for this study as the BPNN-Elman combination model.This model was used to predict the monthly incidence in 20162017 in Chongqing,and compared with the actual monthly incidence rate,the model predicted MPAE=0.319,MRE=0.328,and the effect was better.The model predicts that the peak period of onset is from April to July and from October to December,which is in agreement with the actual epidemic situation.ConclusionThe HFMD epidemic in Chongqing City in 2015 will be severe.The relevant departments of the health industry should strengthen the prevention and control of HFMD,carry out etiological detection and identification of severe cases early in the disease,and reduce the mortality.The focus of prevention and control will be at the peak of onset?from April to July and from October to December?and rational allocation of health resources.It also focuses on high-risk groups?children aged 5 and below,scattered population,and boys?,and uses the main urban districts of the 9th district and its neighboring districts and counties as major monitoring areas to prevent the spread and spread of diseases.In addition,the combination of early warning model prediction effect is better than a single warning model.The BPNN-Elman combination model has the best prediction effect among the six prediction models,and can better predict the incidence of HFMD in Chongqing and provide scientific basis for the prevention and control of HFMD.