Analysis Of Multi-scale Spatial-temporal Characteristics And Risk Factors Of Infectious Diseases

Analysis of spatial-temporal characteristics and risk factors of infectious diseases is crucial for prevention and control of infectious diseases.However,due to the scarcity of data and methodologies,existing researches are still in an early exploratory stage.Limited studies consider the multiple temporal scales,different disease types and investigation of risk factors in depth.Based on the data about gonorrhea,bacillary dysentery and mumps collected in Hangzhou from 2006 to 2013,this thesis,firstly,analyzes the overall characteristics of the three types of infectious diseases.Secondarily,this thesis proposes a multi-scale spatial-temporal analytical framework for systematically analyzing spatial-temporal distribution patterns of infectious diseases at multiple temporal scales(year,season and month),by integrating the models of self-organizing map,kernel density estimation and space-time scan statistic.Finally,the geographical detector is applied to detect the main risk factors of infectious diseases in order to reveal the determinant powers differences and interactive effects between them.This thesis quantitatively analyzes the correlations between factors and diseases and identifies the high-risk areas of infectious diseases.The results show that:(1)The characteristics of demographic distribution vary over different types of infectious diseases.Regarding gonorrhea,the occupation groups with relatively higher incidence are farmers and workers and the majority of cases are young adults.For bacillary dysentery,most of patients are children,students and farmers.The average age of female patients is older than males.The cases of mumps also mainly occur in children and students with the average age about 10.(2)The incidences of the three types of infectious diseases exhibit remarkable inter-annual variations.Generally,the incidences decrease gradually in the districts with high population density,but remain stable with a smaller amount in the sparsely populated areas over the years.Evidently,different types of diseases present spatial-temporal clustering mainly in main urban districts of Hangzhou for different seasons.There is no remarkable seasonal variation in the density of gonorrhea.Bacillary dysentery shows a relatively higher density in summer and autumn but a lower occurrence rate in winter and spring.The seasonal distribution characteristics of mumps are contrary to those of bacillary dysentery.The months that the spatial-temporal clusters of gonorrhea happen in are scattered,whereas those of bacillary dysentery and mumps mainly occur during June to November and December to May,respectively.(3)The number of haze days has greatest determinant power for the spatial heterogeneity of gonorrhea and bacillary dysentery,while the strongest explanatory power for the epidemic of mumps is the proportion of residential area.Each factor has different levels of affection to these three infectious diseases,but the relationships between the factor and the three types of diseases show a similar trend.Population density,average surface temperature,the number of haze days,the proportion of residential area and road density exhibit positive correlations with the incidence of these infectious diseases.On the other hand,the proportion of agricultural population,elevation and the proportion of forest land show negative ones.The interactive effects between factors are stronger than their individual effects.The explanatory powers of interactions between some factors are higher than 0.5,with a few extreme cases of higher than 0.7.This thesis systematically analyzes the spatial-temporal patterns of infectious diseases from macro-to micro-level and explores their risk factors in depth.The conclusion yielded in this research will provide more scientific and sound references for effective prevention and control of infectious diseases and optimal allocation of resources.Furthermore,the proposed spatial-temporal data analytical methodology will provide a new perspective for exploration of spatial-temporal patterns of infectious diseases and investigation of risk factors.