The authors have declared that no competing interests exist
Hemorrhagic Fever with Renal Syndrome (HFRS) is considered as a globally distributed infectious disease, which results in many deaths annually in Hubei Province, China. The outbreak of HFRS is usually characterized with spatio-temporal heterogeneity and is seasonally distributed. Further, it might also be impacted by the influencing factors such as socio-economic and geographical environment. To better understand and predict the outbreak of HFRS in the Hubei Province, the spatio-temporal pattern and influencing factors were investigated in this study. Moran’s
In China, HFRS was mainly caused by two types of Hantaviruses, named Hantaan virus (HTNV) and Seoul virus (SEOV), each associated with a unique rodent host [
Generally, human activities and natural factors were related to the occurrence and epidemic of
Although a variety of methods have been implemented to reduce the occurrence of HFRS, the HFRS cases were still more than 20,000 annually in China from 1980 to 2009 according to the report [
Liu et al. found that temporal and geographic factors both played important roles on the outbreaks of HFRS in central-south China between 2000 and 2009 [
The highlighted block in the above map indicates its location. The underneath map indicates the county boundary of Hubei Province.
Four data sources were used in this study. First, the number of HFRS cases of 76 counties in Hubei Province was provided by Hubei Provincial Centre for Disease Control and Prevention (CDC) between 2005 and 2014. Secondly, human population data was provided by Hubei Provincial Bureau of Statistics based on the 2009 censuses. Thirdly, the map at a scale of 1:4000, 000 was obtained by National Geomatics Centre of China. Fourthly, climatic data from 2005 to 2014 was provided partly by Hubei Provincial Meteorological Bureau and partly from online open sources. These four sources of data were stored in the geographical information database and attribute database. Geographical information database was linked to administrative region by using ArcGIS 10.2. The extent contour and central points for each county were also stored in the geographical information database. Data fields included name, centre point coordinates, county code and etc. Attribute database consists of HFRS cases data per month and per year for each county, demographics and climate data. Data fields for attributes data included county code, average temperature value, average humidity value, average rainfall value, cases number for each county and death number for each county.
Spatial autocorrelation analysis is applied to describe the similarity of geographically proximate units. In this study, spatial autocorrelation analysis was used to clarify the overall autocorrelation pattern of HFRS incidence (clustered, dispersed, or random). Global Moran’s
Moran’s
Spatial autocorrelation analysis measures the globe spatial autocorrelation characteristics based on both county locations and number of HFRS cases simultaneously. It evaluates the spatial distribution pattern by a series number of county locations and the associated number values of HFRS cases.
It is helpful to identify the moving tendencies of HFRS incidence in Hubei Province by investigating its spatial and temporal clustering areas. After the first step of spatial autocorrelation analysis, if HFRS occurred in a clustered way (which can be treated as an epidemic), the spatio-temporal scan statistical analysis could be used to identify the spatial and temporal characteristics of this epidemic
Using this method, the radius of the cylindrical window varies continuously from zero to a specified maximum size. The maximum-size specified the percentage of the maximum total population at risk within the scanning window [
Previous studies have showed that the transmission of HFRS in Hubei/China was influenced by environmental factors such as temperature, humidity and rainfall
First, the ratio of HFRS cases in each month over the whole year, we define it as the HFRS outbreak value.
Each point represents the HFRS incidence rate in a specific year. All of the points are lined to indicate the trend of the HFRS incidence in Hubei Province. Trend line is simulated according to the incidence rates.
Each point represents the Moran’s
year | Moran's Index | Expected Index | Variance | z-score | p-value |
---|---|---|---|---|---|
2005 | 0.0523 | -0.0133 | 0.0055 | 0.8844 | 0.3765 |
2006 | 0.1630 | -0.0133 | 0.0050 | 2.4834 | 0.0130 |
2007 | 0.1242 | 0.1242 | 0.1242 | 2.2474 | 0.0246 |
2008 | 0.1413 | -0.0133 | 0.0026 | 3.0412 | 0.0024 |
2009 | 0.1722 | -0.0133 | 0.0045 | 2.7571 | 0.0058 |
2010 | 0.2374 | -0.0133 | 0.0037 | 4.1336 | 0.0001 |
2011 | 0.0550 | -0.0133 | 0.0030 | 1.2442 | 0.2134 |
2012 | 0.1526 | -0.0133 | 0.0038 | 2.6845 | 0.0073 |
2013 | 0.1642 | -0.0133 | 0.0027 | 3.4013 | 0.0007 |
2014 | 0.2604 | -0.0133 | 0.0038 | 4.4143 | 0.0001 |
The spatio-temporal scan statistical analysis showed that HFRS was not randomly distributed in space during these ten years, which achieved the same conclusion as our previous studies did [
Time | Cluster Number | Total Cases |
---|---|---|
2005 | 5 | 824 |
2006 | 5 | 830 |
2007 | 3 | 605 |
2008 | 2 | 365 |
2009 | 2 | 217 |
2010 | 2 | 302 |
2011 | 3 | 232 |
2012 | 2 | 167 |
2013 | 2 | 253 |
2014 | 1 | 214 |
Spatial and temporal clusters were identified each year from 2005 to 2014. Different colors represent different levels of clusters. The cluster one for each year was regarded as the Most Likely Cluster.
Year | StartDate | EndDate | Counties | P-value |
---|---|---|---|---|
2005 | 2005/6/1 | 2005/11/30 | Huanggang | <0.01 |
2006 | 2006/6/1 | 2006/11/30 | Macheng, Luotian,Yingshan, Huanggang, Zuoshui | <0.01 |
2007 | 2007/7/1 | 2007/12/31 | Yicheng | <0.01 |
2008 | 2008/1/1 | 2008/6/30 | Yicheng | <0.01 |
2009 | 2009/7/1 | 2009/12/31 | Yicheng,Xiangfan,Zhongxiang | <0.01 |
2010 | 2010/4/1 | 2010/8/31 | Yicheng,Xiangfan,Zhongxiang, | <0.01 |
2011 | 2011/1/1 | 2011/6/30 | Yicheng | <0.01 |
2012 | 2012/10/1 | 2012/12/31 | Yicheng, Xiangfan, Zhongxiang, Nanzhang | <0.01 |
2013 | 2013/7/1 | 2013/12/31 | Yicheng | <0.01 |
2014 | 2014/7/1 | 2014/12/31 | Yicheng,Zhongxiang, Jingshan, Dangyang, Jingmen, Tianmen | <0.01 |
Monthly HFRS outbreaks cases are presented in
(A) Average monthly HRFS cases from 2005 to 2009. (B) Average monthly HRFS cases from 2009 to 2014.
Related studies have demonstrated that climate factors had a great contribution to the outbreak of HFRS [
Average monthly humidity values for Hubei Province and Yicheng County are depicted by dotted line and solid line, respectively.
Factors | Spearman Correlation | Sig. (2-tailed) |
---|---|---|
average temperature | 0.011 | 0.417 |
average humidity | 0.041 |
0.001 |
total rainfall | 0.015 | 0.234 |
human population density | 0.397 |
0.000 |
**. Correlation is significant at the 0.001 level (2-tailed).
HFRS is an infectious disease. The related studies have indicated that the density of the virus carriers, human immunity and immigration played significant roles in the transmission of HFRS [
In this study, the relationships between the HFRS cases, spatio-temporal distribution and clustering patterns have been studied. We also explored some influencing factors for the HFRS cases in Hubei Province.
Our previous study of the past three decades of HFRS fatalities data revealed a general trend of decreased in Hubei Province [
First,
Secondly, human migration (from rural areas to cities) and urbanization have both reduced the risk of exposing to rodent excreta. As shown in
Each point represents the ratio in a specific year. All of the points are lined to indicate the trend of the rural population ratio in Hubei Province from 2005 to 2014.
Thirdly,
Each point represents the GDP value in a specific year.
In the study, we also found that the clustering degree was increasing and gathering to the central area of Hubei Province. In addition, from geographic characteristics perspective, HFRS has not been endemic near abundant water resources until 2005. Combining with the topography, geomorphology and vegetation covered environments, this finding might explain by the following.
The previous studies have demonstrated that HFRS in Hubei Province was mainly affected by SEOV virus which was transmitted by Rattus norvegicus during 1980–2012 [
The industrial development often means high-speed economy development and the concentration of population. As listed in
Peng et al. found that HFRS cases were distributed along large water systems (wetland) between 1983 and 1995, such as Yangtze river and Huai river in Anhui Province [
The colored areas indicate the water boundaries in Hubei Province.
The color gradient presents the elevation value of Hubei Province in thematic maps. It is grouped in 10 categories based on the elevation value for each location.
As displayed in
Pervious study has concluded that there are many environmental factors that impact the spatio-temporal dynamics of HFRS, such as the changes of precipitation, temperature, land use and vegetation community dynamics[
Li et al. found that the increase of human population density could make the transmission of the virus much easier [
Climate factors have been long considered as important impact factors for the spreading of HFRS [
The findings in our research presented some new features regarding the trend and influencing factors for the HFRS outbreaks in Hubei Province. First, SEOV and HTNV are both associated the HFRS outbreaks. HFRS seasonal distribution in Hubei Province was characterized by a bimodal pattern (March to May, September to November). This bimodal pattern was corresponding to the peaks of SEOV and HTNV virus respectively. Compared with the peaks happened in each year, we could infer that if peaks mostly emerged in spring time during a year, it could be presumed that SEOV-type HFRS have more influences vice versa. Secondly, the average humidity and human population density were associated with the HFRS epidemic. It can be inferred that the congregation of population may provide more opportunities for the transmission of HFRS virus. And humid climate may contribute to the survival and reproduction for the virus. Thirdly, HFRS outbreaks were more in plains than in other areas of Hubei Province. HFRS case clusters mostly gathered in the central and southern part, the central and southern regions are mainly plain areas or small hilly areas. We did not find that whether the terrain of the wetland (water system) plays a significant role in the outbreak of HFRS incidence. We believe these findings can better understand the spatio-temporal pattern and influencing factors, and can further help to improve the reducing cases of HFRS in the future.
(RAR)
(RAR)
This work was supported by the National Natural Science Foundation of China, No. 71503068; Fundamental Research Funds in Key Research Areas for the Central Universities, No. 2015B09614; Fundamental Research Funds for the Central Universities, No. 2014B15114.