Abstract
Viral hepatitis and tuberculosis affect social productivity seriously. Because they are two typical incidences of morbidity and mortality of infectious diseases, analysis of tuberculosis and hepatitis, and grasping their temporal and spatial differentiation, is an important task. The incidence and mortality of viral hepatitis and tuberculosis in China were revealed by ESDA. It was found that: (1) the morbidity and mortality of tuberculosis kept increasing before 2005, then began declining. The incidence of viral hepatitis volatility peaked in 1988, then declined slowly, with mortality declining; (2) morbidity of tuberculosis and viral hepatitis acted in a highly clustered manner. The incidence of tuberculosis in the west was the highest, decreasing gradually in the east; that in the north was higher than in the south with a lower gradient; the mortality in the middle was the lowest, with the east lower than the west. The incidence of viral hepatitis in the east was lower than in the west; the south was lower than the north, and mortality in total was low; (3) the incidence of tuberculosis morbidity was related closely to SO2 emission, and the incidence of hepatitis morbidity was related closely to GDP; the mortality rate was related closely to GDP.
Introduction
In the past 20 years, hepatitis has been a viral, highly prevalent disease in China (Puoti, Guarisco, & Bellis, et al., 2009); the mortality rate has ranked at the forefront of infectious diseases. Similarly, China’s tuberculosis infection rate was second highest in the world, with only India outranking it (WHO, 2005). These two typical infectious diseases affect people’s production capacity seriously. Viral hepatitis or tuberculosis patients will lose their ability to work an average of three months each year. That is a huge loss and economic burden for a country or society, and consequently the prevention and treatment of infectious diseases is the focus project of relevant government concerns (Dombrovskiy, Martin, & Sunderram, et al., 2007). Therefore, the analysis of spasticity of tuberculosis and other infectious diseases is an important task. In the past, the research and analysis of the regularity of infectious diseases mainly focused on the qualitative narrative of the text, and lacked detailed exploration of its spatial laws and development.
In this study, the spatial and temporal distribution of viral hepatitis and tuberculosis infectious diseases was studied by spatial analysis (Roza, Caccia-Bava, & Martinez, 2012). The temporal and spatial characteristics of infectious diseases in the whole country as well as the various provinces and cities are analyzed from the aspects of autocorrelation, aggregation, trend and hotspots.
Research area and data
The data for researching mainly includes three aspects: (1) the annual incidence of viral hepatitis and tuberculosis in the provinces and municipalities and its mortality (the data come from China’s data network reporting system in the relevant statistical statements); (2) the provinces and municipalities nationwide average population; (3) the total national sulfur dioxide emissions and nearly two decades of average temperature statistics (the data come from the China Statistical Bureau Yearbook).
Methodology
Pushing mapping analysis
Using descriptive epidemiological methods, we analyzed developing trends through the data concerning viral hepatitis and tuberculosis at the scale of provinces (or municipalities) from 1996 to 2014.
Moran’s I autocorrelation analysis
Autocorrelation analysis is an analytical method of inferring whether the observed value of a point in space is related to the observed value of its adjacent point. It is divided into two indices as global autocorrelation and local autocorrelation (Ord & Getis, 1995). Global autocorrelation is used to describe the average degree of association, spatial distribution patterns, and the significance of all objects throughout the study area. The local autocorrelation statistic variables can identify different spatial association patterns (or spatial agglomeration patterns) that may exist in different spatial positions, so as to observe the non-stationary space in the local space and find the spatial heterogeneity between the data.
Moran’s I is interpreted as a correlation coefficient (Jong, Sprenger, & Veen, 1984), which is generally valued -1 to 1. The positive value indicates that the attribute value distribution of the space has a positive correlation, that is, the high observation tends to gather together with the high observed value. In addition, the distribution of Moran’s I value is close to 1 – the closer the relationship between the spatial units, the more the similarity. The negative value indicates that the attribute value distribution of the space has negative correlation, that is, the high observation value tends to gather together with the low observed value; the closer to -1, the greater the difference between the spatial units or the less concentrated the distribution, representing the overall discrete type distribution; 0 value means there is no spatial correlation, i.e. a randomly spatial distribution (Li, Calder, & Cressie, 2007).
The global Moran’s I statistical method assumes firstly that there is no spatial correlation between the subjects, and then the z-score test is used to verify that the hypothesis is true. The z-score statistic consists of Moran’s I, Expected, and Variance 3. As the number of samples n increases, the expected value will gradually approach 0. Z statistic corresponds to the value of P; if it is less than the set level of inspection, you can assume the overall existence of spatial autocorrelation. To clarify the properties of each spatial object in “local” conditions, we use local Moran’s I to find out whether there is spatial autocorrelation in local space. Local Moran’s I method is to decompose the global Moran’s I method into local space for each distribution object.
Getis-Ord Gi* statistic
Getis-Ord Gi* is a measure of a statistically significant test. It is obvious whether the spatial autocorrelation is significant relative to the whole research range. If the spatial unit is the spatial aggregation of the spatial unit area, there is a large significant value. The Gi* statistic for each element returned in the dataset is the z-score. For a positive score with significant statistical significance (e.g. z > 1.96), the higher the z-score, the closer the clustering of the high value (hotspot). For a statistically significant negative score, the lower the z-score, the closer the clustering of the lower value (cold spot). Z values close to 0 indicate that the incidence of tuberculosis in the region around the absence of the situation shows a random distribution. Therefore, according to the Getis-Ord Gi* statistic, the study area is divided into primary hotspots of Z > 2.58, secondary hotspots of 2.58> Z > 1.96, and tertiary hotspots of 1.96 > Z > 1.64; in addition; primary cold spots of Z < -2.58, secondary cold spots of -2.58 < Z <-1.96, tertiary cold spots of -1.96 <Z < 1.64 and other different strengths of the hot/cold area.
Results
Trends maps
The trends of morbidity and mortality of tuberculosis and hepatitis for several years are shown in Figures 1 through 4. Figure 1 shows that the incidence of tuberculosis in 1998–2005 acted as a clear upward trend, reaching its highest point in 2005, when 96 per 100,000 people were infected with tuberculosis; the overall showing is of a downward trend after 2005, falling to more than a ten-thousandth of the ratio in 2006 than in the previous year, although after a small rebound, but the overall decline is steady. It can be seen that the incidence of tuberculosis in China is effectively controlled after 2005. It can also be seen that the overall mortality rate of pulmonary tuberculosis in the whole country from 1998 to 2009 showed an upward trend. Although there was a drop in 2004, it remained at more than 0.2%. After 2009, the incidence of tuberculosis did not rise, indicating that the level of tuberculosis treatment in China is on the rise (Guo, Xiao, & Sun, et al., 2014).
Tuberculosis morbidity (incidence, 1/10 million) and mortality (fatality, %) in 1998–2013.
Hepatitis morbidity (incidence, 1/10 million) and mortality (fatality, %) in 1975–2011.
Analysis of tuberculosis (left) and hepatitis (right) incidence in 2012.
Analysis of national tuberculosis incidence trend in 2012 / 2009 / 2008 / 2007.
We can see from Figure 2 that, from 1998 to 2012, the incidence of hepatitis in China did not show a fixed rise and fall trend. In the 38 years of statistics, there were three small peaks in the incidence of hepatitis, respectively, in 1980, 1988 and 2007. The incidence of hepatitis between 1995 and 2003 was low, at 70 per 100,000 people. But after 2003 it continued to rise, and hepatitis morbidity rebounded. It can be seen that after a small peak of mortality in 1985, the overall mortality rate of viral hepatitis in the country showed a downward trend; in 2009 the downward trend is obvious. Although the number of patients with hepatitis in China has risen in recent years, the incidence of hepatitis has been controlled and the mortality has been reduced due to the increase in medical level of care. Control of the death rate from viral hepatitis has achieved remarkable results.
Spatial autocorrelation clustering
Tables 1 and 2 show global clustering indices of two diseases. The national statistical data for 2003, 2006, 2009 and 2012 were extracted by spatial autocorrelation analysis. The obtained P value is less than 0.01; Z value is higher than 2.58, indicating that the data has a high degree of statistical significance. The Moran’s indices of the provinces and municipalities were positive (> 0.36), indicating a positive correlation between the incidence of tuberculosis in the provinces, where the provinces with high incidence of disease come together.
Partial year tuberculosis global index clustering indices.
Partial year hepatitis index cluster indices.
Hepatitis mortality was analyzed by spatial autocorrelation analysis. P values were less than 0.05 and Z values were lower than 1.64, indicating that the data were statistically significant. The Moran’s I indices of the provinces in the data are negative, indicating that there is a negative correlation between the incidences of hepatitis in the provinces and municipalities, that is, there is no aggregation effect.
Hotspots and cold spots
Figure 3 shows a high concentration and low concentration of the distribution pattern. It can be seen from Figure 3 that the provinces and municipalities with a high incidence of tuberculosis in the country are clustered together, and the provinces with low morbidity are adjacent to each other. There were two significant syndromes in the high incidence of Xinjiang, and the incidence rates in the central and northern regions were lower. These were in Henan, Hebei and Inner Mongolia. The incidence of tuberculosis in various regions basically showed the characteristic western high distribution and eastern low distribution. From the incidence of hepatitis from the eastern coastal to the western inland hepatitis incidence, from the low value of low neighbors to the high value of high neighbors, spatial clustering is obvious. Among the more significant high-value high neighbors are Qinghai and Ningxia; significant low-value neighborhoods are Hainan, Anhui and Hubei.
Trend analysis results
The trend analysis of tuberculosis incidence, mortality and mortality in the whole country from 2003 to 2012 was analyzed. The results of the trend analysis in four years were as follows (rotation angle: position 0: level 120, vertical -11.5).
From Figure 4, we can see that, from the west to the east, the incidence of tuberculosis showed a gradual decline trend, and in the north–south direction the incidence of tuberculosis showed a high north, with the eastern part of the decline to the south continuing to rise. On the whole, the high incidence of tuberculosis is in the western, northeastern and southeastern regions of China, while the incidence in the eastern part of the region remained at a low level.
It can be seen from Figure 5 that the mortality rate of tuberculosis is higher in the west, lower in the central region and increasing in the east. From the results it can be seen that in the incidence of tuberculosis and mortality distribution and geographical location there is a clear correlation. The mortality rate of tuberculosis in the north and the south is basically the same. But overall, the eastern provinces and municipalities of the mortality rate are lower than in the midwest.
Analysis of tuberculosis mortality trend in 2012 / 2009 / 2008 / 2007 (China’s data network reporting system).
It can be seen from Figure 6 that, from west to east, the incidence of hepatitis showed a trend of gradual decline trend, and in the north and south the incidence of hepatitis is relatively flat; before 2010, the north and the south showed a low–high–low trend; after 2010 in the north and the south, the incidence of tuberculosis was essentially flat. On the whole, the incidence of hepatitis is high in the western and middle areas of China while the incidence of hepatitis in the eastern region has been kept at a low and uniform trend.
Analysis of national hepatitis incidence trend in 2012 / 2010 / 2008 / 2006 in order (China’s data network reporting system).
It can be seen from Figure 7 that the mortality rate of hepatitis and tuberculosis is relatively low. From west to east, the mortality rate increased slowly; in the direction of north to south, the hepatitis mortality rate decreased slowly. However, according to detailed statistics, the incidence of hepatitis in Shanghai is higher than in other provinces and cities nationwide. It can be seen from the figure that the mortality rate of hepatitis gradually decreased.
Analysis of national mortality trend of hepatitis 2006 / 2008 /2010 / 2012 (China’s data network reporting system).
Factor analysis
In a crowd, contaminated gas can spread the tuberculosis virus through the respiratory tract, making the incidence of tuberculosis higher (Lawley & Maxwell, 1962). The SO2 emissions from 1998 to 2013 were selected, and the per capita GDP and population were the influencing factors, i.e. the independent variables for tuberculosis. The test of the model shows that for this multivariate linear regression model (Walker, 2008) the coefficient of determination (R Square) is 0.862, and the adjusted R is 0.81, which indicates that the degree of fit is high. The results of the model test–variance analysis table show that the regression model of SIG is 0, indicating that the model meets the F test, a significant statistical significance, while the per capita GDP corresponds to the Sig 0.163, greater than 0.05, so the regression equation of the independent variables does not include per capita GDP. The analysis of the incidence includes SO2 emissions and population. We can get the function as follows: Tuberculosis incidence = -403.647 + (6.849E-6) SO2 + 0.002 population Hepatitis incidence = 5.863E-5GDP + 3.376E-6 population
Conclusion
China’s tuberculosis morbidity and mortality rate has an upward trend, with the west of China being a high incidence zone due to climate and air quality, and the east of China a high incidence zone due to air quality. China’s incidence of hepatitis showed ups and downs to a greater degree, but the overall mortality rate is reducing significantly. This shows that the control of China’s viral hepatitis death rate has achieved remarkable results. The incidence and mortality of hepatitis is closely related to per capita GDP, and tuberculosis morbidity is related closely to SO2 emissions, though the morbidity and mortality rate of the two diseases also relies on population.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Central University of Basic Research Business Project of China (Grant / Award Number: 2009B11714); National Natural Science Key Fund (Grant / Award Number: 41130639).