Epidemiological Characteristics And Ecological Niche Model Of Severe Fever With Thrombocytopenia Syndrome In Jiangsu Province

Objective（1）To screen specific parameters for early and rapid identification of SFTSV infection and SFTS-like diseases in clinical practice,to provide scientific basis for early detection,early diagnosis and early treatment of SFTS.（2）To analyze the epidemiological characteristics,spatial-temporal aggregation and temporal-spatial evolution of SFTS laboratory-confirmed cases in Jiangsu Province from 2010to 2016,and to provide basis for SFTS prevention and control and health resource allocation in Jiangsu Province.（3）To construct the Ecological Niche Model of SFTS natural foci in Jiangsu Province,quantitatively analyze the impact of environmental factors on the distribution of SFTS natural foci,accurately grasp the distribution of SFTS natural foci in Jiangsu Province,and draw SFTS risk map.Methods（1）The suspected cases of SFTS reported by SFTS surveillance sentinel hospitals in Jiangsu Province from 2010 to 2016 were studied.A retrospective study was conducted on the basic information（age,sex,occupation,etc.）,clinical manifestations,results of routine blood tests and epidemiological exposure history.According to the results of RT-PCR detection of SFTSV nucleic acid,it was divided into SFTS confirmed cases and excluded cases.Through single factor analysis and multivariate logistic regression analysis（backward）,specific parameters were screened for early and rapid identification of SFTSV infection and SFTS-like diseases.（2）From 2010 to 2016,the data of laboratory confirmed cases in SFTS of Jiangsu Province came from the subsystem of"China Disease Prevention and Control Information System"-The National Notifiable Infectious Disease surveillance system（NNIDSS）.Population data are obtained from the annual population bulletin published on the website of Jiangsu Statistical Bureau（http://tj.jiangsu.gov.cn/）.The County（district）boundary map of Jiangsu Province is provided by Jiangsu Basic Geographic Information Center.Case data include age,sex,occupation,address,date of onset,date of diagnosis and reporting unit.The epidemiological distribution of SFTS cases（population distribution,time distribution and regional distribution）in Jiangsu Province was described and analyzed.Based on the retrospective spatio-time analysis of discrete Poisson model,SatScan 9.0.1 software was utilized to detect the spatio-temporal agglomeration area in Jiangsu Province from 2010 to 2016.P value and loglikelihood ratio（LLR）were used to determine the clusters and the exact location and risk of the study area.The results were visually displayed by ArcMap 10.2 software.（3）The site of infection was extracted from the"current address"of the laboratory-confirmed STFTS case report card reported in Jiangsu Province from 2010 to 2016.After eliminating the possible human-to-human cases and the cases with unknown infection address,the longitude and latitude coordinates of the infection site were obtained by geolocation according to the current address of each SFTS case.After eliminating the possible human-to-human cases and the cases with unknown infection address,the longitude and latitude coordinates of the infection site were obtained by geolocation according to the current address of each SFTS case.Thirty-two environmental factors were extracted from online databases such as World Meteorological Database,MODIS Satellite Database of Medium Resolution Imaging Spectrometer and Digital Elevation Model Database of Space Shuttle Radar Topographic Mapping Program.Seventeen environmental factors were screened according to the results of correlation analysis and previous research.Based on the maximum entropy modelling method and MaxEnt v3.3.1 modeling toolkit,the niche model of SFTS natural foci in Jiangsu Province was established.We analyze the environmental factors affecting the distribution of natural foci of SFTS and the contribution rate of each factor,and predicted the potential high-risk areas of SFTS in Jiangsu Province..Jackknife was used to test different environmental variables to predict the final outcome.The performance of the model was assessed from using the area under the curve and omission rate.Result（1）From October 2010 to December 2016,247 suspected cases of SFTS in Jiangsu Province were collected,including 99 confirmed cases and 148 excluded cases.Univariate analysis of 38 variables showed that 18 variables,such as age≥50 years old,farmer,Leukopenia,neutropenia,lymphocytopenia,myalgia,nausea,diarrhea,lymphadenopathy,living in hilly or mountainous areas,living in a month before onset,having seen ticks within one month before onset,having ticks attached to animals,being bitten by ticks within two weeks before onset,planting,mowing,grazing and tourism,had statistical significance（P<0.05）.Multivariate logistic regression analysis showed that four factors entered the logistic regression equation:tick biting within two weeks before onset（OR=5.274,95%CI1.831¹5.187）,age over 50 years（OR=4.004,95%CI 1.226¹3.077）,diarrhea（OR=3.042,95%CI 1.143⁸.096）and Leukopenia（OR=1.977,95%CI 1.836⁵.207）.Logistic regression equation was P=1/(1+e^-Z),Z=-2.418+（1.663*bitten by ticks within two weeks beforeonset）+（1.387×*age≥50years old）+（1.113*diarrhea）+（0.682*Leukopenia）.The total accuracy,sensitivity and specificity of the equation were 78.4%,83.3%and 72.1%respectively.（2）Distribution of disease showed that 162 laboratory-confirmed cases of SFTS werereported in Jiangsu Province from 2010 to 2016.The incidence of SFTS increased year by year,ranging from 0.06/1 million to 0.79/1 million.The number of male cases was slightly higher than that of female cases（the sex ratio was 1.46）.The age range of the cases ranged from 18 to84 years.The age distribution was mainly concentrated in the middle-aged and elderly people over 50 years old（74.70%）and the farmer accounted for the highest proportion（71.6%）.Nanjing and Huai’an are high incidence areas of SFTS.The onset time of SFTS in Jiangsu Province is seasonal（M value is 69.63）.Cases occur from March to November every year,and the epidemic peak is bimodal distribution.The first peak occurs in May and the second peak occurs in July.Spatial-temporal scan statistic for the spatial and temporal distributions of SFTS of each year revealed that 13 spatio-temporal clusters were detected,The trend of interannual dynamic distribution shows that the number of SFTS cases clustering counties/districts is increasing year by year,and the area of SFTS cases clustering is gradually spreading from central Jiangsu Province to southern Jiangsu Province.Overall spatial-temporal clustering analysis showed that from 2010 to 2016,three spatial-temporal clusters were detected in SFTS in Jiangsu Province.the most likely cluster existed in Zhenjiang City,Changzhou City and Nanjing City in the southwestern part of Jiangsu Province from May 2016 to July 2016（RR=85.88,LLR=64.63,P<0.01）.Secondary cluster-1 existed in Xuyi County in the southwestern part of Huai’an City.from May 2016 to July 2016（RR=291.00,LLR=60.24,P<0.01）.Secondary cluster-2 were existed in the north of Nanjing City and the south of Yangzhou City（RR=18.14,LLR=15.44,P<0.01）.（3）Ecological niche model based on Maximum entropy modeling showed that 15environmental factors have different effects on SFTS niche model.Slope（suitable range:0.1-4）and max temperature of warmest month（suitable range:32.8-34.2）are the key environmental factors affecting the distribution of SFTS,with contributions of 46.1%and 23.2%,respectively.Secondly,the monthly mean of diurnal temperature difference,Humidity in the driest month and the mean annual humidity contributed 7.5%,6.9%and 5.2%respectively.Other environmental factors contributed less than 5%.The response curve of environmental factors showed that there was a non-linear relationship between the slope and the highest temperature in the hottest month and the probability of SFTS natural foci,both of which were inverted"V"type.Jackknife diagrams of environmental factors show that the slope and the highest temperature in the hottest month can well fit the training data,indicating that they contain valid information not included in other variables.The risk map shows that the southwest of Jiangsu Province,including Nanjing,Yangzhou and Zhenjiang Changzhou,is a high-risk area of SFTS natural foci.The niche model not only identifies the risk areas of SFTS reported by existing cases,but also identifies 10 counties/districts that have not yet been reported by SFTS cases.The model had high accuracy on prediction with the averaged training AUC of 0.926.Conclusion（1）Grazing,being bitten by ticks within two weeks before onse,age≥50 years,diarrhea and Leukopenia can be used as specific indicators for clinical differential diagnosis of SFTS and SFTS-like diseases.Therefore,in the diagnosis of diseases,medical staff should not rely too much on the results of pathogenic test.They should pay attention to inquiring patients’basic information,clinical manifestations and epidemiological exposure history.If patients meet the above specific indicators,it strongly suggests the possibility of SFTSV infection.The results of this study further confirmed the risk factors of SFTSV infection,and then provided a theoretical basis for defining the etiology of SFTS and formulating comprehensive prevention and treatment measures.（2）From 2010 to 2016,the incidence of SFTS in Jiangsu Province increased year by year.Farmers accounted for the highest proportion.Males were slightly more than females.The incidence of SFTS was high in the middle-aged and elderly population.The onset time of SFTS was obviously seasonal,and the epidemic peak was in May and July.The high incidence areas are mainly located in the central and southwest of Jiangsu Province,and the number of cluster counties/districts is increasing year by year.The clusters areas are spreading from the central and western regions to the southwest regions.We should continue to strengthen monitoring in key areas,improve public health education and vector control.（3）Various environmental factors can affect the distribution of natural foci of SFST.Slope and the highest temperature in the hottest month are the main environmental factors affecting the distribution of natural foci of SFTS.The risk map of natural foci of SFTS in Jiangsu Province was drawn up to identify high-risk areas and 10 potential counties/districts with no reported SFTS cases in hilly areas of Southwest China.It provides a model for risk assessment of other insect-borne infectious diseases,and also provides a basis for surveillance of SFTS epidemic situation and key prevention and control in Jiangsu Province.