Abdrakhmanov, Mukhanbetkaliyev, Korennoy, Karatayev, Mukhanbetkaliyeva, and Abdrakhmanova: Spatio-temporal analysis and visualisation of the anthrax epidemic situation in livestock in Kazakhstan over the period 1933-2016

Spatio-temporal analysis and visualisation of the anthrax epidemic situation in livestock in Kazakhstan over the period 1933-2016

Abstract

An analysis of the anthrax epidemic situation among livestock animals in the Republic of Kazakhstan over the period 1933-2016 is presented. During this time, 4,064 anthrax outbreaks (mainly in cattle, small ruminants, pigs and horses) were recorded. They fall into five historical periods of increase and decrease in the annual anthrax incidence (1933-1953; 1954-1968; 1969-1983; 1984-2001; and 2002-2016), which has been associated with changes in economic activity and veterinary surveillance. To evaluate the temporal trends of incidence variation for each of these time periods, the following methods were applied: i) spatio-temporal analysis using a space-time cube to assess the presence of hotspots (i.e., areas of outbreak clustering) and the trends of their emergence over time; and ii) a linear regression model that was used to evaluate the annual numbers of outbreaks as a function of time. The results show increasing trends during the first two periods followed by a decreasing trend up to now. The peak years of anthrax outbreaks occurred in 1965-1968 but outbreaks still continue with an average annual number of outbreaks of 1.2 (95% confidence interval: 0.6-1.8). The space-time analysis approach enabled visualisation of areas with statistically significant increasing or decreasing trends of outbreak clustering providing a practical opportunity to inform decision-makers and allowing the veterinary services to concentrate their efforts on monitoring the possible risk factors in the identified locations.





Introduction

The zoonotic disease anthrax, caused by the spore-forming bacterium Bacillus anthracis, affects most mammals, especially herbivores but also humans. The spores are highly resistant and can survive in the environment for decades. Although vaccination and several effective antibiotics exist, treatment is generally sporadic in the less developed parts of the world where anthrax is common. The course of the disease is usually rapid and almost always fatal.

The first reliable information of anthrax in the Republic of Kazakhstan (RK), formerly part of the Union of Soviet Socialist Republics (USSR), refers to the end of the 19th century, while official recording of cases did not start until the early 1930s. The sapronotic existence of the infectious agent determines its natural focality (Lukhnova et al., 2004). Anthrax soil foci have been accumulating over a long period of time in the RK. Throughout the 20th century, burial of animals that had died from anthrax was carried out at the site or in its immediate vicinity until the Ministry of Agriculture, USSR issued a decree in 1951 requiring the mandatory burning of all animal corpses without prior action in an effort to end further soil contamination (Cherkassky, 2002). However, a significant number of anthrax cases in animals and humans were still recorded during the period from the mid-1950s to the 1980s. It should be mentioned that in the 1930s until the 1950s a lot of anthrax outbreaks remained unaccounted for due to a lack of staff, lack of proper accounting and illiteracy of the population leading to thousands of anthrax-infected corpses of farm animals being buried in the ground. Due to the long-term survival of B. anthracis spores in the soil, the sites of these spontaneous burials contributed to an epidemic instability of the area.

From the early 1950s, vaccination of agricultural animals against anthrax was introduced in the RK, but mass vaccination was only launched in 1961. Nevertheless, even during this period the vaccination coverage was far from complete due to intensive growth of livestock populations and inadequate local provision of vaccines, lack of staff and poor accountability. In parallel, a farreaching campaign to develop virgin and fallow lands was started in the USSR that largely affected the territory of the RK, accounting for 61% of developed land (McCauley, 1976; Rowe, 2011). The development of virgin and fallow land was accompanied by a massive influx of people into the Republic from other regions of the USSR (more than 2 million people), as well as a concomitant increase in the number of livestock (the number of small ruminants and pigs increased four times during the period 1941 to 1961). This resulted in a significant growth of incidence in both humans and animals in 1950s. Although there was a downward trend in the annual incidence of anthrax after 1969, the situation remained tense and more 50 anthrax outbreaks per year were registered in the Republic until the early 1980s. Following the introduction of follow-up vaccination of susceptible animals in 1981 and strict monitoring of compliance with veterinary legislation, the situation was stabilised at an average of 20 outbreaks per year.

After the collapse of the USSR, the number of livestock decreased by more than half between 1991 and 1998 leading to the disintegration of the existing management system and the large collective and state farms and significant migration of the rural population to cities producing a decrease in annual anthrax mortality. The gradual improvement of the economic situation in the country since the late 1990s followed by large investments in livestock and upgraded veterinary services also contributed to an improvement of the epizootic situation. The current epizootic situation can be characterised as stable with an average annual number of anthrax outbreaks of 1.2 (95% confidence interval: 0.6-1.8). Despite the success achieved, animals and humans still succumb to this disease today in various parts of the RK. As a result, the risk of death from anthrax infection in humans, and the huge financial costs associated with eliminating outbreaks in animals require a study of the epidemic situation in the country, where old foci continue to be identified and new ones registered. A specific feature of the causative agent is that the reactivation of conserved soil foci usually occurs because of human economic activity, such as wasteland reclamation or construction work that brings B. anthracis back to the surface from deeper soil layers (Knop, 1981). The formation of new foci due to natural factors, e.g., the relocation of soil with spores by wind, rain or flooding, is also likely (Hugh-Jones and Blackburn, 2009).

Access to records of anthrax outbreak in the country since the early 20th century enables in-depth studies of patterns of emergence and re-emergence of the disease. Available datasets can be analysed using special research methods, in particular those including space-time analysis that can visualise patterns based on geographical information systems (GIS). However, only few studies employing GIS-based methods to search for patterns of spatial and spatio-temporal clustering of anthrax exist, e.g., Kracalik et al. (2011, 2012) who studied outbreaks using a dataset covering the period 1960 to 2006.

The purpose of the present research was to identify and visualise spatio-temporal anthrax hotspots and their temporal trends taking into account data for time period as broad as possible (1933-2016). We applied the user-friendly space-time cube technique supplemented by linear regression as this would contribute to the development of preventive and contra-epizootic measures as well as rational use of material, labour and financial resources. The technical approaches used made it possible to comprehensively analyse epidemiologically significant information related to the geographical spread of anthrax. They also allowed the compilation of an accurate review and, potentially, a forecast of the epidemic situation by revealing regularities in the spatio-temporal distribution and the activity of stationary unfavourable anthrax, significantly contributing to the assessment of the current epidemic situation.

Materials and Methods

The anthrax epidemic situation in the RK was investigated for the longest period for which official records exist, i.e., 1933-2016.

Study area

RK is a central Asian country with an area of more than 2.7 million km2 and a current population of 17.8 million. The average population density is about 6.5 people/km2. The country has extensive natural resources with agriculture, in particular livestock breeding, being one of the main branches of the national economy. Administratively, the country is divided into 14 regions or oblasts, which in turn are divided into 179 districts or rayons.

Data

The data were obtained from the RK register of areas with a high risk of anthrax (Cadastral, 2002), supplemented with material from statistical veterinary reports posted on the official websites of regional and republican veterinary authorities, as well as with official data from the World Organization for Animal Health (OIE) and material from our own expeditionary research.

In total, there were 4,064 anthrax outbreaks in the RK during the study period (Figure 1). In this research, we defined an outbreak as a recorded occurrence of anthrax cases in a geographically localised population of animals (a herd on pasture, on a farm, in a village, etc.), followed by the burial of dead animals in the immediate vicinity (with or without cremating the remains). Since some outbreaks were repeatedly recorded in the same locations (villages, pastures) at different years, the total number of unique locations was 1,798.

Methods of analysis

In order to facilitate the space-time analysis, we divided the entire historical period from 1933 to 2016 into five time intervals corresponding to the main phases of changes in economic development and veterinary surveillance of the RK: 1) 1933-1953; 2) 1954-1968; 3) 1969-1983; 4) 1984-2001; 5) 2001-2016. For each of these periods we 1) estimated the general temporal trend of annual incidence; 2) identified statistically significant hotspots and 3) indicated territories with increasing/decreasing tendencies in hotspot emergence. We carried out the analysis through the construction of a set of space-time cubes for hotspot analysis as used by individual researchers (Andrienko et al., 2003; Harris et al., 2017), which are available from the Space Time Pattern Mining toolbox within ArcGIS geographical information software package (ESRI, 2017). The space-time cube can be defined as a regular three-dimensional structure based on the geographic space along the x an y coordinates, with the vertical dimension representing time. The cube consists of space-time bins aggregating studied data points. For each bin, space and time dimensions were set. The further analysis relied on counting the number of data points within individual bins and application of various statistical approaches to reveal patterns of data distribution through both space and time looking for statistically significant cluster of data points or associated values. Epidemiologically, a hotspot is the clustering of outbreaks in a particular location over particular time periods. For example, the presence of a large number of outbreaks within a particular bin indicates the presence of a statistically significant hotspot in a given space-time cell provided that it is surrounded by bins (both in space and time dimensions) that also have a large number of outbreaks. We selected 50 km for the spatial bin dimension and one year as time period, which would reveal the patterns of annual changes in the epidemic situation in a 50 x 50 km space (2,500 km2). The size of 50 km allows accounting for possible reporting bias when known outbreaks could be linked to the nearest villages.

When doing this type of analysis, it is useful to bear in mind that hotspots come in various guises, e.g., new, consistent, increasing, sporadic, decreasing, fluctuating and historical (ESRI, 2017) depending on the patterns of their emergence over time. The identification of a specific type of hotspot at a given location demonstrates an increased clustering of outbreaks of the disease at the location during different time periods.

Software

The analysis using the space-time cube was performed using a Space Time Patterns Mining Tools software tool package, embedded in the geographical information system (GIS) ArcGIS 10.4.1 by ESRI (Redlands, CA, USA). The systematisation of the anthrax database and fitting the distributions were carried out using a standard Microsoft Excel package with @Risk add-in (Palisade Inc., Ithaca, NY, USA).

Statistical approach

We applied Getis-Ord Gi* spatial statistics in its temporal interpretation to identify statistically significant hotspots of anthrax outbreaks (Ord and Gettis, 1995), where a conclusion on the presence/absence of a hotspot in a space-time bin is reached on the basis of z-values. We also used Mann-Kendall statistics (Mann, 1945; Kendall and Gibbons, 1990) to identify statistically significant trends of hotspot emergence. Mann-Kendall statistics, based on a comparison of the values in consequent space-time bins, are used to estimate trends in the evolution of the z-statistics over time. The value delivered by the Mann-Kendall statistics allows a conclusion to be reached about increasing or decreasing tendencies in the emergence of hotspots at a given location, which tendencies can be treated as indicators of improvement or deterioration of an epidemic situation in this particular location. In addition to the spatiotemporal analysis, we used a linear regression model to estimate the intensity of increases/decreases in the annual incidence during each time period. In this case, the annual number of anthrax outbreaks (Ni) was the dependent variable, with the year (Yi) as the independent variable. Thus, the regression equation was Ni ~ β+αY, where β is the intercept and α the slope. Goodness of a model fit were assessed by the coefficient of determination R2, which indicates how much of the data variability is explained by the model (range: 0 ... 1). The statistical software environment R was applied to calculate the temporal correlation of the annual incidence (R Core Team, 2014).

Results

Table 1 summarises the results of the analysis of the annual anthrax incidence for the five chosen periods using both spacetime analysis and linear regression.

1933 to 1953

The annual number of outbreaks during this period ranged from 3 to 78 with an average value of 29±4. The maximum number of outbreaks was registered in 1947. The annual incidence increased with an average rate of 2.587±0.446 outbreaks per year. New hotspots in this period emerged in East Kazakhstan region, and consecutive and sporadic hotspots in the Kostanay and South Kazakhstan regions. Statistically significant increasing trends of hotspots with wide distributions were registered in the Akmola and Karaganda regions (Figure 2).

1954 to 1968

There was a sharp increase in the annual incidence from 40 to 220 outbreaks with an average rate of 12.754±1.193 outbreaks per year in this period. The consecutive hotspots were registered in the West Kazakhstan, South Kazakhstan as well as in the Zhambyl and East Kazakhstan regions. In this case, statistically significant upward trends of hotspot emergence were recorded on most of the inhabited and cultivated territory of the RK (Figure 3).

1969 to 1983

There was a decline in the annual incidence from 220 to 20-30 annual outbreaks with an average rate of 6.632±1.124 outbreaks per year. No hotspots were detected during this period and statistically significant decreasing trends of hotspots were present in the vast territories of the northern, eastern and southern regions of the RK (Figure 4).

1984 to 2001

In this period, the anthrax incidence showed a tendency to slowly decline at a rate of 1.154±0.285 outbreaks per year. The general trend was negative. During this period, sporadic hotspots were recorded in the South Kazakhstan and Zhambyl regions, but the overall trend of hotspot emergence decreased (Figure 5).

2002 to 2016

From 0 to 5 outbreaks per year were registered in the latest period providing an average number of 1.2 (0.6-1.8). Given the relative constancy of the annual number of outbreaks, this value can be accepted as the average predicted incidence rate per year for the next time period. As the space-time analysis method requires at least 60 events, it could not be applied due to the low total number of outbreaks for the given time. The outbreak map for this period is shown in Figure 6.

Discussion

Our analysis of the epizootic situation for anthrax in the RK over the whole study period shows that the main temporary stages of economic and agricultural growth in the Republic satisfactory explain the variant periods of anthrax incidence. Thus, there was a significant increase in annual incidence throughout the period 1933 to 1953, likely to be associated with the development of the national economy leading to an increase in the number of livestock without any measures for disinfection and disposure of dead animals. According to the national statistics agency, the cattle stock in the Republic consisted of 3.3 million heads in 1941 and reached 4.5 million heads by 1951. It should also be noted that the absence of widespread registration of anthrax cases along with the population’s low literacy rate with respect to pathogenesis and disease control measures probably led to a significant underestimation of the incidence rate during this time period.

The campaign to develop virgin and fallow lands that started in 1954 led to a dramatic growth of anthrax incidence. The increase of the susceptible populations (both of humans and animals) together with expansion of activities into virgin lands led to the removal of the causative agent from the soil surface at old (unaccounted) anthrax burial sites and to intensive wind erosion of the soil, which contributed to the wind-blown distribution of anthrax spores over considerable distances. All this significantly increased exposure to anthrax spores, as confirmed by studies by Soviet scientists (Adamovich and Nikonov, 1970). In contrast, improvements in the veterinary surveillance along with introduction of mass vaccination led to a considerable reduction in the number of foci and brought anthrax away from its previous status of largescale epizooty in the period 1961-2010. The spatio-temporal analysis of the epizootic situation demonstrates the absence of hotspots and a confident downward trend in their formation in nearly all the affected territories in the Republic starting from 1969 (Figure 4). Still, there was a number of sporadic hotspots in the southern part of the Republic after the 1980s, though combined with a downward trend in their emergence in previously affected territories (Figure 5).

The low number of outbreaks in the most recent period (2010-2016) did not allow the application of the chosen method of spacetime analysis, but a significant spatial fragmentation of outbreaks can be observed during this period as well as the absence of repeated outbreaks in particular locations. Currently, as a result of the systematic implementation of comprehensive preventive and antiepizootic measures, the area of anthrax registration has gradually diminished. Compared to the 1950-1970s, large territories have become disease-free, including the North Kazakhstan, Aqtobe, Atyrau, Kostanay and Mangystau regions, where neither animal deaths due to anthrax, nor human cases have been recorded in the last 20 to 30 years (in Mangystau for 50 years). Nevertheless, disease outbreaks are being registered in regions that previously experienced intensive epidemics, which indicates a continuing threat of the removal of the anthrax disease agent from old (unrecorded) burials as a result of agriculture, construction and other human activities, as well as the impact of various natural factors related to old soil foci (spring floods, inundations, earthquakes, etc.).

Our analysis shows a clearly expressed non-uniformity in the territorial distribution of anthrax foci and differences in their epidemiological activity throughout the country. Currently, new data on the epidemiology and ecology of anthrax are being accumulated, which are associated with the introduction of new forms of economic relations in the country, changes in people’s social conditions and the formation of private property in agriculture. Today, anthrax in the RK is recorded in areas where developed livestock breeding exists and where the soil is neutral or slightly alkaline with a high humus content together with high temperatures and sufficient humidity (Aikembayev et al., 2010; Abdrakhmanov et al., 2017). In this connection, there is a need for an in-depth causal analysis of the incidence of anthrax in relation to natural (soil types, precipitation, temperature etc.) and anthropogenic (livestock, construction, land improvement, etc.) factors.

We found analysis using the space-time cube useful for studying retrospective data on disease incidence as it allowed us to draw conclusions regarding both trend variations within the target territory over time and localisation of sites with the most intensive incidence. From the epidemiological point of view, it is especially useful to assess the extent of the manifestation of the epidemiological process in the target territory of the country by visualising spatial cells where a steady increase in the incidence exists (i.e., repeated outbreaks). In the case of anthrax as a natural focal disease, this tendency may indicate the presence of foci in the given territory, such as spontaneous, unregistered and insufficiently disinfected anthrax burial sites. As for the shortcomings of this method, the need to set temporal and spatial dimensions of the cube cells was noted, a fact which is not always obvious and requires additional study. The choice of spatial and temporal resolutions of the space-time cube has a significant effect on the returned result. In our case, since the dates of the outbreaks could not be known more accurately than years, the minimum possible time step had to be one year. If more detailed time data were available, and if necessary to analyse the situation for shorter time periods, time steps such as months or weeks could be chosen. For our analysis, we accepted that the spatial size of the bin automatically calculated by ArcGIS was based on the spatial distribution of the data analysed. When studying a disease situation with a strongly pronounced local transmission, the spatial step can be estimated as the maximum clustering distance, for example, using the K-function (Dixon, 2012).

In general, the use of studies with information and communication technologies allows critical information to be obtained in advance on the various scenarios for the emergence, development and counteraction of natural focal epidemics. The most urgent problem is the development of an epidemic or outbreak as a result of the accidental removal of the causative agent of anthrax from conserved soil foci. Therefore, more complete and detailed information permitting a more detailed analysis of the spatio-temporal patterns of the disease’s spread could serve a basis for the development of effective preventive and anti-epizootic measures that would ensure epidemiological well-being throughout the Republic.

Conclusions

We show here an example of the application of the space-time analysis technique using a space-time cube to identify trends in increasing and decreasing anthrax incidence on the territory of the RK over a long historical period. This type of analysis is well suited for working with historical data on disease incidence and allows a visual assessment of the development of the epidemic situation during certain time periods that are associated with the main stages in the development of the country’s economy. The visualisation of hotspots and their trends enabled us to reveal the territories most at risk for an epidemic during the period under consideration. The obtained data should facilitate veterinary surveillance leading to the development of plans for preventive and anti-epizootic measures against anthrax.

Acknowledgements

This work has been accomplished under the National Budgetary Program #249 providing conditions for the development of manufacturing, processing and sale of livestock production. The authors are particularly grateful to the Editor-in-Chief Prof. Robert Bergquist for his invaluable assistance in formatting the manuscript.

References

1 

SK Abdrakhmanov, YY Mukhanbetkaliyev, FI Korennoy, AA Sultanov, AS Kadyrov, DB Kushubaev, TG Bakishev, 2017. Maximum entropy modelling risk of anthrax in the Republic of Kazakhstan. Prev Vet Med 144:149-57.

2 

VL Adamovich, NN Nikonov, 1970. The importance of landscapeecological factors in the epizootology of anthrax. Message 2: Comparative method of assessment of epizootic tension of a territory. J Microbiol 8:113-7.

3 

AM Aikembayev, L Lukhnova, G Temiraliyeva, T Meka-Mechenko, Y Pazylov, S Zakaryan, G Denissov, Easterday W Ryan, MN Van Ert, P Keim, SC Francesconi, JK Blackburn, M Hugh-Jones, T Hadfield, 2010. Historical distribution and molecular diversity of Bacillus anthracis, Kazakhstan. Emerg Infect Dis 16:789-96.

4 

N Andrienko, G Andrienko, P Gatalsky, 2003. Visual data exploration using space-time cube. Proceedings of the 21st International Cartographic Conference (ICC), Durban, South Africa.

5 

BL Cherkassky, 2002. Epidemiology and prevention of Anthrax. INTERSEN, Moscow, Russia.

6 

PM Dixon, 2012. Ripley ’s K function. In: AH El-Shaarawi, WW Piegorsch (eds.) Encyclopedia of Environmetrics. John Wiley & Sons, Chichester, UK, pp. 1796-803.

7 

ESRI, 2017. How emerging hot spot analysis works. Available from: http://desktop.arcgis.com/en/arcmap/latest/tools/spacetime- pattern-mining-toolbox/learnmoreemerging.htm#GUID- 09587AFC-F5EC-4AEB-BE8F-0E0A26AB9230

8 

NL Harris, E Goldman, C Gabris, J Nordling, S Minnemeyer, S Ansari, M Lippmann, L Bennett, M Raad, M Hansen, P Potapov, 2017. Using spatial statistics to identify emerging hot spots of forest loss. Environ Res Lett 12:024012.

9 

M Hugh-Jones, J Blackburn, 2009. The ecology of Bacillus anthracis. Mol Aspects Med 30:356-67.

10 

MG Kendall, JD Gibbons, 1990. Rank correlation methods. Griffin, London, UK.

11 

AG Knop, 1981. The influence of anthropogenic transformation of environment on Anthrax soil foci. Actual problems of zoonotic infections. pp. 25-27.

12 

IT Kracalik, JK Blackburn, L Lukhnova, Y Pazilov, Jones ME Hugh, A Aikimbayev, 2012. Analysing the spatial patterns of livestock anthrax in Kazakhstan in relation to environmental factors: A comparison of local (Gi*) and morphology cluster statistics. Geospat Health 7:111-26.

13 

I Kracalik, L Lukhnova, A Aikimbayev, Y Pazilov, G Temiralyeva, JK Blackburn, 2011. Incorporating retrospective clustering into a prospective cusum methodology for anthrax: Evaluating the effects of disease expectation. Spat Spatiotemp Epidemiol 2:11-21.

14 

LY Lukhnova, AM Ajkimbaev, EK Pazylov, 2004. Epidemic process of Anthrax in Kazakhstan. Bull Agric Sci Kazakhstan Bastau 7:44.

15 

HB Mann, 1945. Nonparametric tests against trend. Econometrica 13:245-59.

16 

M McCauley, 1976. Khrushchev and the development of Soviet agriculture. The Virgin Land Programme 1953-1964. Palgrave Macmillan, UK.

17 

K Ord, A Getis, 1995. Local spatial autocorrelation statistics: distributional issues and an application. Geogr Anal 27:286-306.

18 

Core Team R, 2014. R foundation for Statistical Computing, Vienna. Available from: http://R-project.org/

19 

WC Rowe, 2011. Turning the Soviet Union into Iowa: The Virgin Lands Program in the Soviet Union. In: S. Brunn (ed.) Engineering earth. Springer, Dordrecht, Germany. pp. 237-56.

Figure 1.

Annual dynamics of the registration of anthrax outbreaks in the Republic of Kazakhstan, 1933-2016.

gh-12-2-589-g001.jpg
Figure 2.

Spatio-temporal analysis of anthrax outbreaks in the Republic of Kazakhstan 1933-1953. Statistically significant hot/cold spots (A); statistically significant trends of hot/cold spots emergence (B).

gh-12-2-589-g002.jpg
Figure 3.

Spatio-temporal analysis of anthrax outbreaks in the Republic of Kazakhstan, 1954-1968. Statistically significant hot/cold spots (A); statistically significant trends of hot/cold spots emergence (B).

gh-12-2-589-g003.jpg
Figure 4.

Spatio-temporal analysis of anthrax outbreaks in the Republic of Kazakhstan 1969-1983. Statistically significant hot/cold spots (A); statistically significant trends of hot/cold spots emergence (B).

gh-12-2-589-g004.jpg
Figure 5.

Spatio-temporal analysis of anthrax outbreaks in the Republic of Kazakhstan, 1984-2001. Statistically significant hot/cold spots (A); statistically significant trends of hot/cold spots emergence (B).

gh-12-2-589-g005.jpg
Figure 6.

Anthrax outbreaks in the Republic of Kazakhstan over the period 2002-2016.

gh-12-2-589-g006.jpg
Table 1.

Summary of annual anthrax incidence in the Republic of Kazakhstan for the five historical periods 1993-2016.

Period Number of outbreaks Spatio-temporal analysis Linear regression
Total Cattle Horse Pig Small ruminants Ann min° Ann max# Trend and Z-value P Intercept Intercept SD Slope Slope SD P R2
1933-1953 615 415 34 22 144 3 78 Increasing Z=4.3543 <0.001 0.8286 5.5952 2.5870 0.4456 <0.001 0.6395
1954-1968 1,850 1,051 90 95 614 39 221 Increasing Z=4.4098 <0.001 21.305 10.851 12.754 1.193 <0.001 0.8978
1969-1983 1,236 747 73 30 386 22 155 Decreasing Z=-3.5631 <0.001 135.457 10.216 -6.632 1.124 <0.001 0.7283
1954-2001 346 218 34 5 89 3 31 Decreasing Z=-3.1863 0.001 30.1830 3.0813 -1.1538 0.2847 <0.001 0.5066
2002-1916 17 15 1 1 0 0 5 NA§ NA 0.14286 0.66566 0.13214 0.07321 0.0943 0.2004

[i] SD, standard deviation; NA, not available. °Minimum annual number of outbreaks for the corresponding time period

[ii] #maximum annual number of outbreaks for the corresponding time period

[iii] §corresponding analysis method could not be applied due to insufficient number of outbreaks in this time period.

Abstract views:
948

Views:
PDF
68
HTML
321

Article Metrics

Metrics Loading ...

Metrics powered by PLOS ALM


Copyright (c) 2017 Sarsenbay K Abdrakhmanov, Ersyn E Mukhambetkalyev, Fedor I Korennoy, Bolat Sh Karatayev, Aizada A Mukhambetkalyeva, Aruzhan S Abdrakhmanova

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
 
© PAGEPress 2008-2018     -     PAGEPress is a registered trademark property of PAGEPress srl, Italy.     -     VAT: IT02125780185