| ObjectiveThis study aims to describe the epidemiological distribution and spatiotemporal characteristics of typhoid ad paratyphoid fever, to find out spatiotemporal hotspots, to clarify the spatial and temporal effect quantificationally, to explore the correlation between urbanization and enteric fever, to provide reference for further research and offer scientific basis for making monitoring and prevention and control tactics of typhoid and paratyphoid fever.MethodReport cards of typhoid and paratyphoid fever from 2006 to 2014 were collected from national notifiable disease reported system, and subtotal was made by time interval(year and month), spatial interval(cities and districts), then incidence classification maps were drawn after incidences were calculated. Spatial autocorrelation and heterogeneity was analyzed by global and local spatial autocorrelation analysis. Spatial and temporal hotspots was detected by space-time permutation scanning method. Spatial, temporal and spatiotemporal effect were quantitatively described with Bayesian hierarchical models. Besides, the urbanization of each city and district in Zhejiang was decribed with classification maps and its influence on enteric fever was illustrated by Bayesian models.ResultsA total of 7073 typhoid cases(3823 male and 3250 female)and 4709 paratyphoid(2667 male and 2042 female)cases were included in this analysis. Almost the incidence of men is higher than women each year. Most cases concentrated in 30 to 50 year old group. Generally speaking, the incidence of each age group declined gradually, while that of under 18 group had the slowest pace. The main populations affected were farmers, workers, migrant workers and students. The enteric fever had obvious and similar seasonality, onset most frequently in summer, generally increased in April, highest in July and then decreased. Usually presenting a unimodal distribution except for bimodal distribution in 2006 and 2007, a small peak in spring before summer.The average typhoid incidence per year of Ningbo, Wenzhou, Shaoxing were respectively 2.446,2.475,2.159/105, among the top three. Except for Lishui and Hangzhou in 2007, the incidences of other cities were over 1/105, and declined gradually gradually afterward. The incidence had declined to 0.5/105 except Ningbo and Wenzhou. Especially, Shaoxing, Jinhua, Taizhou were at a quick decreasing pace, while Lishui always had low typhoid incidences. As to paratyphoid fever, Hangzhou and Jinhua had the highest yearly incidences, respectively 1.683,1.991/105. The incidences of Hangzhou and Taizhou in 2007 were specially high, respectively 8.416,8.162/105. Generally, the paratyphoid incidences declined after 2007, fluctuating during those years, and had decreased to below to 0.3/105 by 2014. Jiaxing, Huzhou, Quzhou, Lishui always had low typhoid incidences.Moran’s I index about typhoid incidence was statistically significant each year besides 2012,indicating that typhoid was positive spatial autocorrelation rather than random. While paratyphoid incidence did not have this significant result as typhoid. Local autocorrelation analysis of typhoid indicated two spatial cluster areas, one is Yuecheng, Kecheng, Shangyu of Shaoxing and Yuyao, Cixi, Zhenhai, Jiangdong, Yinzhou of Ningbo, the other is Lucheng, Longwan, Ouhai of Wenzhou and Yongjia of Taizhou, presenting high-high cluster. Local autocorrelation analysis of paratyphoid indicated spatial cluster mainly located in Yuhuan of Taizhou and Dongtou of Wenzhou, presenting high-high cluster, and Lin’an presented high-low cluster, while its neighbor Tonglu presented low-high cluster.Scanned for typhoid cases within district and day as basic unit, the study obtained four spatiotemporal clusters with statistically significance. The four clusters were all high aggregations. The most likely cluster was Jiaojiang of Taizhou from October 10, 2006 to December 4,2006(RR=21.53,P<0.001). The second likely cluster was Cixi of Ningbo from February 3,2014 to May 15,2014(RR= 8.00, P<0.001). The third likely area contained 23 districts, from June 11,2012 to December 11,2014(RR=0.50, P<0.001), and the districts were Xiaoshan, Fuyang, Tonglu, Jiande of Hangzhou, Yuecheng, Kecheng, Xinchang, Zhuji, Shengzhou of Shaoxing, Wucheng, Jindong, Wuyi, Pujiang, Panan, Lanxi, Yiwu, Dongyang, Yongkang of Jinhua, Tiantai, Xianju, Linhai of Taizhou, Liandu, Jinyun of Lishui. The forth likely cluster was Pingyang of Wenzhou, from June 28,2007 to August 27,2007(RR=3.98, P<0.001). Scanned for paratyphoid cases, the study obtained three spatiotemporal clusters with statistically significance. The first two were high aggregation and the third was low aggregation area. The most likely cluster was Linan of Hangzhou, from June 28,2007 to August 27, 2007(RR=3.52, P<0.001). The second likely cluster was Wenling of Taizhou, from September 26,2007 to October 19,2007(RR=6.70, P<0.001). The third likely area contained 48 districts, from June 12,2007 to November 8,2007(RR=0.34, P<0.001). The scanning result within county as basic unit verified the result and stated the detailed towns.The urbanization in north of Zhejiang was superior to the south. Hangzhou had the highest urbanization rate, which had increased to 56.5% in 2014. The urbanization rates of Jiaxing, Huzhou, Shaoxing, Ningbo and Zhoushan were over 30%, and Jiaxing’s rate increased most rapidly. Jinhua, Wenzhou, Quzhou, Taizhou and Lishui had low urbanization rates, less than 30%. In each city, these districts located in the urban had higher urbanization rates than rural areas.The result of Spearman correlation analysis illuatrated per capita gross national product and the male ratio had statistical negative correlation with typhoid and paratyphoid fever incidence, but urbanization rate had not. However, the result of Bayesian model illustrated that typhoid risk ratio had negative correlation with urbanization rate and positive correlation with the male ratio, paratyphoid risk ratio had negative correlation with per capita gross national product and the male ratio.Conclusion1. The typhoid and paratyphoid fever had similar epidemiological characteristics. Firstly, the male were more liable to be infected. Secondly, most cases concentrated in 30 to 50 year age group. Thirdly, the incidence of each age group declined gradually, while that of under 18 years group had the slowest pace. Fourthly, the main populations affected were farmers, workers, migrant workers and students. Fifthly, the enteric fever had obvious seasonality, onsetting most frequently in summer, while in 2006 and 2007, a small peak time appeared in spring before summer.2. The high-incidence areas of typhoid and paratyphoid were partial coincided. The high incidence of typhoid mainly located in coastal region in the east of Zhejiang, especially in Ningbo, Taizhou, Wenzhou. The high incidence of paratyphoid mainly located in Linan of Hangzhou, Taizhou, and the middle and north of Zhejiang.3. The distribution of typhoid had statistical spatial autocorrelation rather than randomized, presenting obvious spatiotemporal cluster. Although the distribution of paratyphoid had not statistical global spatial autocorrelation, the scanning result indicated high aggregation spatiotemporal spot and low spot.4. The urbanization in north of Zhejiang was superior to the south. Hangzhou had the highest urbanization rate. In each city, these districts located in the urban had higher urbanization rates.5. The Bayesian hierarchical model indicated independent spatial and temporal effect, and time correlation wasn’t statistically significant.6. The result of Bayesian spatiotemporal hierarchial model illustrated that typhoid risk ratio had negative correlation with urbanization rate and positive correlation with the male ratio, paratyphoid risk ratio had negative correlation with per capita gross national product and the male ratio. |
PDF Full Text Request