| Background:Dengue fever is an acute mosquito-borne infectious disease caused by dengue virus.With the impact of climate change,population movement,urbanization,the incidence of dengue fever has increased markedly worldwide in recent decades.In 2015,Chaozhou City experienced the largest dengue fever epidemic since the record.To explore the epidemiological features,transmission and driven factors of dengue fever at different spatial-temporal scale,thereby can provide reference for its prevention and control.Objective:1.To describe the epidemiological characteristics of dengue fever in Chaozhou City in 2015,then explored its transmission at different spatial-temporal scale.2.Constructing spatial-temporal multicomponent model based on multidimensional data,and analyzed the driven factors of dengue fever.Methods:The information about indigenous dengue fever cases,meteorological,social,economic and demographic data of Chaozhou in 2015 were collected.Descriptive analysis was used to analyze the meteorological,social demography,economic data and the time,spatial and population distribution of dengue fever cases.The global spatial autocorrelation Moran’s I index of dengue fever cases were calculated.Knox test was used to identify possible spatial-temporal clusters by given corresponding critical thresholds.The spatial distances were set varying from 0 to 1000 m(step: 100 m)and from 5 to 100 m(step: 5 m)to test the spatial clusters of cases at large and fine scale distances,respectively.The temporal distances were set by varying from 1 to 15 days by 1-day step.The strength of clustering(S-value)and relative risk(RR)of each critical threshold was calculated.Power-law method on spatial-temporal multicomponent models was used to analyze the epidemic characteristics of dengue fever cases,which were further decomposed into autoregressive components,spatiotemporal components and endemic components.Results:In 2015,a total of 1380 indigenous dengue fever cases were reported in Chaozhou City,while the incidence rate of dengue was 50.6 per 100000.Male to female ratio of 0.91: 1,the incidence of women than men(x2=5.25,P=0.022).Incidence increased with age(x2=270.85,P<0.01),the incidence of 65 years of age and above the group was 107.4 / 10 million,occupation to domestic and unemployed,retired staff,workers mainly.Male to female ratio was 0.91:1,with the incidence rate of female was greater than male(x2=5.25,P=0.022).The incidence rate showed an increase tendency with age(tendency x2 = 270.85,P<0.01),while the group of beyond 65 years old owned the highest(107.4 per 100000).Household workers,the unemployed and workers represented the major groups.The temporal distribution of cases throughout August to December,while the peak of the epidemic was observed in September 13th(90 cases reported).All district-level regions(included 4 districts/counties and 27 streets/towns)in Chaozhou city reported dengue cases,in which Xiangqiao District(included 13 streets)reported a total of 1249 cases(accounting for 90.5%).The global spatial autocorrelation analysis showed that there were positive spatial correlations of each street epidemic,which was nonrandomly distributed(Moran’s I index was 0.406,Z score was 6.20,P <0.01).The average time distance for all cases was 15.8 days,with an average spatial distance of 3.1 km,which density distribution of case pairs was like the power-law distribution.Large-scale analysis showed spatial-temporal clusters under different time and space combinations(P<0.05),while S-value reached a high level in short-distance(≤100m)within 10 days of critical thresholds.Moreover,the fine-scale analysis found that the higher S-value was displayed within ≤20 m and 3 days.The relative risk of dengue fever(RR)decreased rapidly with the increase of spatial distance.Within 1 day,we found the RR value was 2.3 in short spatial distances.As the distance increased to 0.3 km,the RR value dropped rapidly to 1.4,then dropped to 1.2(0.8 km),while the RR value was still greater than 1 until 3km.In gender aspect,spatial-temporal clustering features of dengue fever cases is similar.The total strength of clustering(∑S)of cases pairs occurred male-male > male-female > female-female.With different ages,varying spatial-temporal clustering patterns of dengue spread in people were found.The S-value of children-children cases pairs was obvious lower than that other combinations,and there were four spatial-temporal clusters in approximate distances at 2,4-5,7-9,11-13 days.In addition,we found a spatial-temporal cluster with high risks of dengue fever in the pairs between children and the elderly cases with 1 day as the temporary boundary,and 15-30 meters as the spatial boundary.Higher population density and per capita GDP in the region,the local risk level is greater,while the daily maximum temperature on opposite.The decay parameter estimate of Power-law method is d = 5.808(95%CI: 4.076~7.539),which represents a strong decay of spatial interaction for higher-order neighbors.The follow-up impact from the previous dengue fever epidemic was strong to the Xixin and Chengxi street.Impact of the epidemic from nearby districts(Xixin and Chengxi street)was obvious in Taiping and Xihu street.Local risk of dengue fever seemed increasing in Xiangqiao and Fengxi street before early October.Conclusion:Dengue incidence showed a rising trend in Chaozhou city.The factors of population,economic and temperature were closely related to the local risk level of dengue fever.The spatial-temporal clustering features of different characteristics people were heterogeneous.By using spatial-temporal multicomponent models,different components of each study area were identified to propose targeted measures. |
PDF Full Text Request