| Background Hand,foot and mouth disease(HFMD)is an emerging enterovirus-induced infectious disease,which is also a typical self-limited disease,primarily affecting children aged within 5 years.More than twenty enterovirus which can cause HFMD,EV71 and Cox A16 are mainly included.Common symptoms are mouth pain,anorexia,low thermal,small sores or ulcers in hand,foot and mouth,which appear on the majority of patients with mild,a small part of children can appear with myocarditis,neurogenic pulmonary edema,aseptic meningoencephalitis and complications,a few would become severe and even lead to death.HFMD has been frequently reported worldwide with high incidence,since its original identification in New Zealand in 1957.The disease has become one public health issues of common concern to the international society.HFMD had been listed in China as class C notifiable disease since 2008.It’s showed that 11,762,589 cases of HFMD were reported in China till the end of 2015,in which 3,361 cases were fatal,the annual incidence is around 117 cases per 100,000 persons according to a recent report from National Health and Family Planning Commission.Moreover,195 dead cases of HFMD were reported in 2016,incidence in 2016 was 178/100,000,HFMD kept the top among the C class infectious diseases by reported incidence from 2009 to 2017.HFMDwas affected by natural and social environment,therefore,the epidemic features of HFMD in our country is varied among different regions.Some studies suggested that HFMD was associated with temperature,rainfall,wind speed and other natural factors,meanwhile,with the number of factories,population density,ratio of students,and other social factors such as health and economic conditions.As the provincial capital,the average annualreported cases of Xi’anfrom 2008-2015 accounted for more than 35% of the entire province,kept the top in citiesamong Shaanxi Province.Average incidence of HFMD in Xi’an is higher than the national average,which means that Xi’an is the key of the HFMD prevention and control in Shaanxi province.Thus,this study aimed to determine the spatial-temporal distribution and epidemic characteristics of HFMD in Xi’an andprovide the scientific basis for local HFMD prevention and control.Objective 1.To describe the temporal,spatial and demographic distribution characteristics of HFMD and illustrate the change rule byspatial-temporal sequence of HFMD in Xi’an.2.To explore the spatiotemporal patterns and high risk clusters in national levels,based on zoning and matching the clinical data and geographic data.3.To evaluate quantitativelythe relationship between meteorological variables and spread-prevalent regularity of HFMD.4.To establishan adjusted prediction model which accord with the characteristics of Xi ’an area and provide the scientific basis for local HFMD prevention and control.Methods 1.Data collection.Xi’an is the provincial capital of Shaanxi,has a population exceeding 8.46 million residents,and consists of 13 counties and districts divided into 176 townships.The demographic data for each township were obtained from the Shaanxi Provincial Bureau of Statistics.The Reported cases and meteorological data were collected and sortedfrom 2009 to 2015.Average minimum temperature(TEML,0.1°C),average temperature(TEMA,0.1°C),average maximum temperature(TEMH,0.1°C),relative humidity(RHU,1%),evaporation(EVP,0.1mm),ground temperature(GST,0.1°C),precipitation(PRE,0.1mm),pressure(PRS,0.1HPa),sunshine duration(SSD,0.1h),wind speed(WIN,0.1m/s),PM2.5 and ENSO were included.2.Descriptive epidemiological analysis methods were used to analyze the temporal,spatial anddemographic distribution characteristics of HFMD in Xi’anfrom 2009 to 2015.HFMD incidences werecompared with chi-square test.3.Illustrate the spatial distribution and regularity of the tendency changeon the township-level digital map by using Arc GIS.Explore the spatiotemporal patterns and space-time clusters of HFMD with high rate,the spatial-time cluster analysis was performed using Sat Scan10.0.Evaluate the relationship between meteorological variables and weekly incidence of HFMD by Spearman and cross-correlation analysis.4.Establish Nonlinear Neural Network Model with the HFMD and meteorological data of 2009-2014 to predict the weekly incidence of 2015,assessthe forecasting effect by comparing the predicted data and the observed data.5.This study is mainly adopted by the software as Microsoft office 2013,Arc GIS 10.2,R Language 3.1.1,Sa TScan10.0,Eviews,SPSS13.0.Results 1.The epidemiological distribution characteristics: A total of 158,257 cases of HFMD were included in this study,most patients were children younger than 4 years old,accounting for the largest proportion of all reported cases each year(ranging from 90.66% to 92.76%).The median age of reported cases was 2.86 years.The ratio of male to female patients was about 1.47:1.The proportion of scattered children occupied the highest.A seasonal pattern was observed in Xi’an,which showed a major peak in Apr-Jul and a minor peak in autumn(Oct–Nov)from 2009 to 2015.2.The spatial distribution characteristics: the townships over 500/100,000 from 2009 to 2015 presented an encirclement which were mainly distributed at the junctions around Xi’an.In addition,Chang’an District,Cao Tang Town and Dong Da Town always being hotspots,which belong to Hu County.3.Cross-correlation analyses showed that the climatic variables had positive correlation with disease incidence,except for PRS,WIN and PM2.5.In this study location,TMEL,TEMA,TEMH,PRS,EVP and GST were significantly correlated with the weekly reports of HFMD with lags of 3–6 weeks.The PRS with a 4-week lag time were found to have the highest correlation with the incidence of HFMD.4.Prediction model: In this study,we incorporated climatic variables with lags into the model,including: TEML,TEMA,TEMH,EVP,GST,PRS.The predictions of forecasting set were shown that each prediction obtained from the model was very close to each observation.The adjusted R2 value for the model was 0.81.It showed that the effect of this model was good.Conclusion 1.A total of 158,257 cases of HFMD were reported during 2009-2015,most patients were children younger than 4 years old.The ratio of male to female patients was about 1.47:1.An obvious seasonal pattern was observed in Xi’an,which showed a major peak in Spring-Summer and a minor peak in autumn.2.The townships over 500/100,000 from 2009 to 2015 presented an encirclement which were mainly distributed at the junctions around Xi ’an.The characteristics of the typical in urban-rural zones were with large population density,poor sanitary conditions,mostly in the villages with low economic conditions.More policy support should focus on public health education and improvement of living environment to the zones.3.Cross-correlation analyses showed that TEMA(0.1°C),PRS(0.1HPa),EVP(0.1mm)and GST(0.1°C)were significantly correlated with the weekly reported HFMD with lags of 3–6 weeks.The PRS with a 4-week lag time were found to have the highest correlation with the incidence of HFMD.The result promptedthat meteorological factors can be considered to be the main influencing factors into the model of applied research.4.The predictions of forecasting set were shown that each prediction obtained from the model was very close to each observation.The adjusted R2 value for the model was 0.81.It showed that Nonlinear neural network model would be used as a tool for disease forecasting,forapplying to HFMD early warning. |
PDF Full Text Request