Spatiotemporal Distribution Characteristics And Prediction Of Dengue Fever In Guangdong Province From 2013 To 2019

Background:Dengue fever is an acute infectious disease caused by the dengue virus that affects nearly half of the world’s population and has a significant social impact and economic burden.It is endemic in many places such as Guangdong and Yunnan,and a trend toward expanding the scope and intensity of the dengue fever epidemic has been observed in China in the past decade.The incidence rate in Guangdong Province was significantly higher than the domestic average.However,the factors influencing the dengue fever epidemic remain unclear,and studies on the spatial and temporal distribution characteristics in the last decade are lacking.Objective:To analyze the spatial and temporal distribution characteristics of dengue fever in Guangdong Province from 2013 to 2019,and to make short-term forecasts of dengue fever epidemic data.This study provides decision support for monitoring and predicting the dengue fever epidemic in Guangdong Province,and a scientific reference basis and theoretical guidance for guiding the prevention and control of dengue fever and rational allocation of limited health resources.Data:Information on the dengue fever epidemic was obtained from the Statutory Infectious Disease Reports published by the Public Health Science Data Centre,Guangdong Provincial Centre for Disease Control and Prevention,and Guangdong Provincial Health and Wellness Commission to collect monthly incidence rates of dengue fever in Guangdong Province from 2013 to 2019,with monthly incidence rates by age group and gender in each city.Meteorological information was obtained from the National Meteorological Scientific Data Sharing Service Platform to collect the monthly average temperature,monthly average maximum temperature,monthly average minimum temperature,monthly average relative humidity,and monthly average precipitation data for Guangdong Province from 2013 to 2019.Methods and content:（1）The annual incidence distribution of dengue fever in each city of Guangdong Province from 2013 to 2018 was mapped.（2）The global Moran’s I spatial autocorrelation analysis in the Arc GIS 10.8 software was used to identify a spatial autocorrelation in the incidence of dengue fever in the Guangdong Province region.Two local spatial aggregation study methods,local Moran’s I spatial autocorrelation analysis in the Arc GIS 10.8 software and simple spatial scan statistical method,were used to identify the high-risk areas of the dengue fever epidemic in Guangdong Province from 2013 to 2018.The temporal and spatiotemporal distribution characteristics of dengue fever incidence were studied using a simple temporal scan.High-risk periods and areas of dengue fever incidence were identified.（3）The correlation between dengue incidence and meteorological factors in Guangdong Province was analyzed using Spearman’s rank correlation.A generalized additive model combined with a Poisson distribution was used to screen for meteorological factors influencing the incidence of dengue fever,and meteorological factors were used to fit the change in the incidence of dengue fever by regression.（4）An autoregressive integrated moving average（ARIMA）model was constructed using the incidence of dengue fever in Guangdong Province from 2013 to 2018.Incidence data from January to December 2019 were used to test the model and evaluate its fitting and prediction effects.Results:（1）The incidence of dengue fever in Guangdong Province indicated an overall significant decreasing trend from 2013 to 2018.The global spatial autocorrelation analysis results suggest a specific spatial aggregation of dengue fever in Guangdong Province.From 2013 to 2018,local Moran’s I spatial autocorrelation identified six cities with a high risk of dengue fever.The median incidence of dengue fever in high-risk cities decreased from 29.61/100,000 in 2014 to 6.48/100,000 in2018.The results of the single spatial scan analysis identified 25 aggregation areas in Guangdong Province in the year,including 13 primary aggregation areas and eight secondary aggregation areas.The results of the spatiotemporal aggregation analysis revealed primary aggregation areas,including the immediate aggregation area for Guangzhou,and the aggregation period from 2014.（2）The cyclical variation in dengue incidence in Guangdong Province from 2013 to 2019 was significantly correlated with the cyclical variation in the mid-year climate.The correlation between temperature-related indicators and rainfall was more significant,with correlation coefficients of mostly>0.7.The 93.4%variation in dengue incidence each month in Guangdong Province can be attributed to the average maximum temperature,relative humidity,and precipitation in the first two months.（3）Using the monthly dengue incidence rates in Guangdong Province from 2013 to 2018,the results indicated that the ARIMA（0,1,0）（2,1,0）₁₂ model fit was relatively optimal,and the predicted incidence trends were entirely consistent with the actual incidence trends,except for the actual value in October,which was slightly higher than the upper limit of the95%CI.The predicted values in all months were within the CI of the predicted values,indicating that the predicted values of the model are consistent with the actual situation and that the fit is good.Conclusions:（1）The Cochran-Armitage trend test revealed a significant decreasing trend in the incidence of dengue fever in Guangdong Province from 2013to 2018.The incidence of dengue fever in Guangdong Province indicated some spatial aggregation,and high-risk areas demonstrated a significant decreasing trend.（2）We achieved better results using meteorological factors to fit the incidence of dengue fever in the province.（3）The ARIMA model constructed to fit the incidence of dengue fever in Guangdong Province demonstrated satisfactory and good prediction results,and can be used to forecast the trend of dengue fever in Guangdong Province.