The Study Of Spatiotemporal Distribution And Time Series Model Of Chinese Mumps

Objective:To analyze the epidemiology and spatiotemporal distribution of mumps in China,to explore the correlation between the incidence of mumps and meteorological factors,and to establish time series model for mumps in China,evaluating its predictive effect,and providing scientific basis for the prevention and control of mumps in China.Methods:（1）We collected data about mumps cases of different provinces or municipalities from 2006 to 2016 through the"China Disease Prevention and Control Information System",and collect meteorological data from the same period through the China Meteorological Data Network;（2）Thematic maps,spatial autocorrelation and trend surface analysis methods in ArcGIS 10.4 software were used to explore the spatial correlation and"hots pots"or"cold spots"of mumps in China;（3）According to climate and economic differences,the mainland was divided into North China,Northeast China,East China,Central China,South China,Southwest China and Northwest China.Spearman correlation was preliminarily used to analyze the correlation between mumps incidence and meteorological factors in each region;（4）ARIMA product season models were established by using the monthly incidence of each region from 2006 to 2015 after the steps of sequence smoothing,model identification and ordering,parameter estimation and model diagnosis,the optimal ARIMA models were determined based on the BIC criterion;（5）The cross-correlation function is used to screen the meteorological factors related to the mumps,and the selected factors were included in the optimal ARIMA models to establish the ARIMAX models;（6）The mumps incidence of 2016 was used for prediction and verification,and we compared the predictive effect of two models.Results:（1）In 2006-2016,a total of 3,278,967 cases of mumps were reported nationwide,and the numbers decreased year by year.The number of children under 10 years old accounted for 60.70%of the total reported cases,and the ratio of male to female was about 1.67:1.The number of cases in spring and winter accounted for 30.00%and 30.06%of the the annual incidence,respectively.（2）The spatial autocorrelation of the incidence of mumps showed that the distribution of mumps in China was spatial clustering with Moran’s I index（Z>1.96,P<0.05）.The local spatial autocorrelation analysis has detected 54"hot spots"and 10"cold spots"in China with Moran’s I and G coefficient.The"hot spots"was mainly concentrated in East China,South China,Central China and some southwestern regions of China,while the"cold spots"was mostly distributed in the northwest,north and northeast regions of China.The trend surface analysis showed that the mumps have a weak inverted U-shaped distribution,and the east was higher than the west,the south was higher than the north.（3）Spearman correlation showed that the relationship of mumps in different regions was mainly related to monthly mean pressure,monthly average temperature,monthly average maximum temperature and monthly average precipitation.（4）Analysis of cross-correlation function found that monthly mean pressure,monthly average relative humidity,monthly average precipitation,monthly average temperature,and monthly average maximum wind speed have lag effect on the incidence of mumps,after progressive screening into the ARIMAX model,we found temperature and precipitation were the main influencing factors.（5）After stepwise inclusion,the optimal models of North China,Northeast China,East China,Central China,South China,Southwest China and Northwest China were obtained as ARIMAX（1,0,0）（1,0,1）₁₂,ARIMAX（0,1,0）（0,1,1）₁₂,ARIMAX（0,1,0）（0,1,1）₁₂,ARIMAX（0,1,1）（0,1,1）₁₂,ARIMAX（0,1,1）（0,1,0）₁₂,ARIMAX（1,0,0）（2,1,0）₁₂,ARIMAX（1,0,0）（2,1,0）₁₂,separately.The relative error of prediction was between 0.00%and 42.86%,which was better than ARIMA models（0.00%⁵0.00%）,and the long-term prediction effect is better.Conclusion:（1）The incidence of mumps in China showed spatial clustering,the"hot spots"were mainly distributed in the southern provinces,while the"cold spots"were mainly concentrated in the northern provinces.And the incidence was cyclical and periodic,mainly concentrated in spring and winter.（2）Temperature and precipitation have an impact on the incidence of mumps in most parts of China.（3）The multi-time series ARIMAX model with meteorological factors has a better goodness of fit than the ARIMA model,and the relative error of prediction is reduced,especially for long-term prediction.