| Hand,foot and mouth disease(HFMD)is a global childhood infectious disease caused by enterovirus,which mainly threatens the health of children under 5 years old,become one of the biggest public health problems in China.The increasing impact of global climate change on infectious diseases has accelerated their spread and spread,which is a great challenge and threat for the prevention and control of infectious diseases.The direct or indirect effects of climate change on the incidence of HFMD and the propagation of enterovirus have been confirmed in published studies.The HFMD incidence of Shandong Province ranks among the top five in China,with an average annual incidence of 93.7/100,000 while the average annual incidence of HFMD in Zibo City is 127.9/100,000 far higher than the average level in Shandong Province,which also indicates that HFMD is a public health challenge to be solved in Zibo City.Therefore,it is particularly important to understanding the epidemiological characteristics of HFMD in Zibo City,grasp the influence of meteorological factors on HFMD,and accurately predict the incidence trend of HFMD based on weather factors.It can minimize the number of patients and the loss caused,and provide a scientific basis for actively developing HFMD prevention strategies and public health countermeasures to reduce the spread of infectious diseases caused by climate change.Methods:The purpose of the study was to establish a time series database based on the incidence data of HFMD in Zibo city from January 1,2010 to December 31,2019 and to conduct a descriptive analysis of the epidemiological characteristics of HFMD in Zibo for exploring the susceptible population and the incidence peak in Zibo city.Spatiotemporal scanning method was used to explore the high-risk cluster area of HFMD in Zibo City,which strengthen prevention and control in key areas.The effects of daily mean temperature and relative humidity on HFMD were investigated based on DLNM.We set incidence data of HFMD from 2010 to 2018 as the training set,and 2019 as the test set,combined with meteorological factors and to construct SVR,RFR and GAM,to predict the incidence of HFMD of Zibo city in 2019.We compared predictive accuracy of models through RMSE and MAPE and R2 index,and choose the highest accuracy and the most suitable prediction model,providing a scientific basis for related departments to formulate prevention and control measures.Results:1.From January 1,2010 to December 31,2019.there were 52095 cases of HFMD in Zibo city,with average incidence of 127.9/100,000.The top three year of incidence is 2010,2014 and 2016 respectively,and the lowest year was is in 2019.The total incidence trend presents generally declining trend.The incidence of HFMD shows obvious seasonal characteristics,the peak was concentrated from May to September,and a cycle about 2-3 year.HFMD threats children under the age of five,especially age of 3.The susceptible people are concentrated in scattered children,followed by kindergarten children and students.The incidence area was clustered the downtown area of Zibo city and the surrounding county area.The top three areas were Zhang-dian district,Linzi area and Zichuan district.The distribution of cases present trend of south more than north.2.Based on distributed lag nonlinear model,the relationship between average daily temperature,relative humidity and the incidence of HFMD in Zibo City was fitted.The effect curve of average daily temperature and HFMD was roughly inverted "V" shape,which was divided into high temperature section and low temperature section based on the reference value of the median daily temperature of 14.5℃.The risk value RR of HFMD in low temperature section decreased gradually with the increase of temperature,while the risk value of HFMD in high temperature section increased gradually.Similarly,the relative humidity took the median of 64%as the reference value,and the effect curve of incidence and relative humidity showed an inverted "V" shape.3.Support vector regression,Random forest regression and Generalized additive model were used to predict the incidence of HFMD of Zibo City in 2019,and the prediction accuracy of the model in the order was SVR>RFR>GAM.The prediction error of SVR model is the smallest,suggesting that Support vector regression is more suitable for the prediction of HFMD in Zibo city.Conclusion:1.HFMD occurred in all counties of Zibo city,and the main clustering area was distributed in Zhangdian District and nearby counties and districts,showing a trend of more in north than south of Zibo city.There are obvious seasonal characteristics,and the peak was concentrated is from May to September.The susceptible population is under 5 years old,especially 1-3 years old group.It is suggested that the relevant departments should take protective measures prepare the early warning and prevention and control of foot-and-mouth disease before the peak of the disease for the high-risk group and key areas,2.The effect curve of daily average temperature and HFMD in Zibo showed an inverted"V" shape,and the relative risk RR in low temperature section decreased with the increase of temperature,while it gradually increased in high temperature section.The effect curve of relative humidity and HFMD also showed an inverted "V" shape.The relative risk RR of low humidity decreased with the increase of humidity,while the RR of high humidity segment gradually increased with the increase of humidity.3.The prediction accuracy of Support vector regression is high,which is suitable for the application of HFMD prediction for public health department in Zibo... |
PDF Full Text Request