Influence And Prediction Model Of Meteorological Factors And Rat Density On Occurrences Of Hemorrhagic Fever With Renal Syndrome

ObjectiveTo understand the epidemiological characteristics and spatial and temporal distribution of hemorrhagic fever with renal syndrome in Guangzhou,further explore the dynamic response relationship between meteorological factors,rat density and the occurrences of HFRS.Based on meteorological factors and rat density,building a support vector machine model to predict the occurrences of HFRS,so as to provide reference for the prediction and early warning of diseases.Methods1.To collect the data of hemorrhagic fever with renal syndrome in Guangzhou from 2006 to 2021,and understand its time,population and regional characteristics through descriptive analysis methods.Its spatiotemporal distribution characteristics are analyzed by global spatial autocorrelation and local spatial autocorrelation.2.Meteorological factors,rat density data and case data in Guangzhou from 2006 to 2019 were collected,and the data were sorted out monthly to establish a database.Vector autoregressive model was established to investigate the relationship between meteorological factors,rat density and the occurrences of hemorrhagic fever with renal syndrome.3.Data of meteorological factors,rat density,population density and other environmental factors as well as case data in Guangzhou from2006 to 2019 were collected,and the data were sorted out monthly to establish a database and to build a support vector machine model to predict the incidence risk of HFRS.Results1.A total of 1786 cases of hemorrhagic fever with renal syndrome were reported in Guangzhou from 2006 to 2021,and the annual incidence ranged from 0.26 to 1.12/10 million.The occurrences of HFRS was distributed in spring and winter.The ratio between male and female of the disease was 2.82:1.The majority of cases occurred in the age group of20-49 years.High-risk occupation was housework and unemployed people.The spatial and temporal distribution of HFRS showed aggregation characteristics,and the high-high concentration area was mainly in Haizhu District.2.The vector autoregressive model showed that the monthly cases of HFRS was positively associated with the two months lagging effect of rat density and negatively associated with monthly average temperature,monthly maximum temperature,monthly minimum temperature,monthly relative humidity,monthly wind speed,monthly rainfall,monthly air pressure,and monthly sunshine hours.The most influential factors was the average temperature,which was 26.80%.3.The accuracy of the support vector machine model was better than that of the gradient boosting machine model,and the output parameter of RMSE was 4.852.The results of the support vector machine model showed that the predicted monthly cases of HFRS were basically consistent with the actual cases.Conclusions1.The occurrences of hemorrhagic fever with renal syndrome in Guangzhou showed a downward trend as a whole,with a high incidence in spring and winter.There were differences among the population,and the incidence of HFRS in males was higher than that in females.The spatial and temporal distribution of HFRS showed aggregation characteristics,and the high-high concentration area was mainly in Haizhu District.2.Both meteorological factors and rat density have certain influence on the occurrences of HFRS.The vector autoregressive model showed that the monthly cases of HFRS was positively associated with the lagging effect of rat density and negatively associated with monthly average temperature,monthly maximum temperature,monthly minimum temperature,monthly relative humidity,monthly wind speed,monthly rainfall,monthly air pressure and monthly sunshine hours.The contribution rate of meteorological factors to the occurrences of hemorrhagic fever with renal syndrome was 45.52%.3.Meteorological factors,rat density and environmental factors were included as predictors,and the support vector machine model was constructed,and its output RMSE was 4.852.Support vector machine(SVM)model can be helpful for early prediction of the occurrences hemorrhagic fever with renal syndrome.