|Year : 2019 | Volume
| Issue : 3 | Page : 150-155
High risk disease mapping and spatial effect of pulmonary tuberculosis in Kerbala, Iraq
Suhad Hadi Mohammed1, Mohanad Mohsin Ahmed2, Zuhair Hasan Mohammed3, Azeez Adeboye4
1 Department of Clinical Laboratories, College of Applied Medical Sciences, Kerbala University, Kerbala, Iraq
2 Department of Microbiology, College of Medicine, Kerbala University, Kerbala, Iraq
3 Department of Architecture, College of Engineering, Kerbala University, Kerbala, Iraq
4 Department of Statistics, University of Fort Hare, Alice, South Africa
|Date of Submission||22-Jun-2019|
|Date of Decision||26-Jul-2019|
|Date of Acceptance||16-Aug-2019|
|Date of Web Publication||10-Sep-2019|
Department of Statistics, University of Fort Hare, Alice
Dr. Suhad Hadi Mohammed
Department of Clinical Laboratories, College of Applied Medical Sciences, Kerbala University, University Street 11252, Kerbala
Source of Support: None, Conflict of Interest: None
Introduction: Pulmonary tuberculosis (PTB) remains one of the top ten causes of death globally. A crucial component of controlling this infection is by limiting the spread of the disease. This study aimed to identify hot spot geographical areas with PTB incidence and to evaluate spatial global autocorrelation using geographical information science (GIS) technology. Methods: Using tuberculosis register software, cases of PTB were recorded by Chest and Respiratory illnesses center in Kerbala governorate. Seven years' data records were used to investigate hot spots and spatial distribution of PTB cases in Kerbala. Patients whom reside outside Kerbala were excluded from the current analysis. Loess smoothing Seasonal decomposition trend was applied to analyze temporal patterns and clusters mapping of tuberculosis, using R statistical software version 3.5.1. Correlation analysis (pairwise Spearman) was used to observe the association among the factors contributing to PTB prevalence. Standard morbidity ratio (SMRs) was used to find the morbidity rate Bayesian conditional auto-regressive model was used in the analysis and estimation of the parameters was done through Markov Chain Monte Carlo methods to estimate the space mapping variability cluster of disease risk and covariates effect. Results: A significant correlation was found between PTB prevalence and the age (r = 0.731) and was also found in gender (r = 0.822). Most of the cases were distributed within the age of 17–50 years (68.4%). Tuberculosis cases were not randomly distributed in which the variables occur with asymptotic probabilities with unpredictable spacing and that there was the presence of high global autocorrelation among PTB cases in the City of Kerbala. Approximately, 59% of all PTB cases were seen in six quarters. Conclusion: Spatial analysis using GIS reveals useful information about epidemiological situation of PTB cases in Kerbala Province, Iraq. Additionally, this study predicted possible places for PTB transmission based on hot spot analysis and the continuous presence of infection during the studied period.
Keywords: Autoregressive effects, Bayesian approach, cluster mapping, loess smoothing, nonspatial effects
|How to cite this article:|
Mohammed SH, Ahmed MM, Mohammed ZH, Adeboye A. High risk disease mapping and spatial effect of pulmonary tuberculosis in Kerbala, Iraq. Biomed Biotechnol Res J 2019;3:150-5
|How to cite this URL:|
Mohammed SH, Ahmed MM, Mohammed ZH, Adeboye A. High risk disease mapping and spatial effect of pulmonary tuberculosis in Kerbala, Iraq. Biomed Biotechnol Res J [serial online] 2019 [cited 2019 Nov 20];3:150-5. Available from: http://www.bmbtrj.org/text.asp?2019/3/3/150/266569
| Introduction|| |
Pulmonary tuberculosis (PTB) is still considered one of the most communicable diseases worldwide in spite of great efforts made to minimize the incidence of infection. Since 1990 WHO ranked TB as the seventh most morbidity-causing disease in the world, and expected it to stay in the same position up to 2020. In 2015, It's estimated that there are 10.4 million new cases and 1.5 million deaths worldwide. Persistence and dissemination of this disease might be attributed to the existence of high-risk population whom they live in crowded area (especially with poor health care access),, and still transmit the bacteria (Mycobacterium tuberculosis) to noninfecting persons. That's why the Center for Disease Control and Prevention recommend testing high-risk population.,
In Iraq and according to the WHO global tuberculosis report in 2017, the estimated incidence of TB cases is 16000 with an estimated incidence rate of about 43/100,000 populations. The mortality rates due to TB disease is about 1200 deaths. The total number of notified TB in 2016 is 7317 case, 59% of these are PTB (~4317 PTB notified PTB cases).
Kerbala represents one of holiest governorates in Iraq after Al-Najef as it is the place of Imam Hussayn (The grandson of Prophet Mohammed) and Abbas shrines (Imam Hussayn's Brother). The number of its populations about 1,012,356 million persons. Millions of pilgrims are visiting Kerbala all over the year and especially in the months of Muharram and Safar.
According to internal-agency information and analysis unit in 2010, The number of registered and notified PTB cases is lower than the real estimated cases, and there are many hidden, especially for women cases. There are many reasons behind hidden cases; poverty, stigma, and unavailability of qualified diagnostic tests, are the most important ones. Approximately 11.4% of Kerbala population are living under poverty line. This makes Kerbala required great efforts to eliminate TB incidence.
Recently, geographical information science (GIS) technology become one of the most important tools in public health researches that are involved in disease mapping and identifying high-risk locations., The availability of this technology might improve understanding of PTB dynamics of incidence in the area with high burden and consequently recognizing factors that are responsible for disease persistence. Several previous studies were made aiming to identify whether the existing of tuberculosis is in clustered or random.,,, This idea came from the first law in geography in which the geographer Tobler (1930) stated that “Everything is related to everything else, but near things are more related than distant things.” therefore, several geographical surveillance studies and spatial analysis have been conducted to identify high prevalence or hot spot areas (i.e., area with elevated cluster of an event) in different countries such as India, Portugal, Japan, South Africa. In all these studies, significant high incidence rates and clusters of PTB infection in the studied areas were found.,,,,,,,,,
To the best of our knowledge, no previous research has been designed to investigate the distribution of confirmed PTB cases and define the area with high disease burden in Holy Kerbala, Iraq, using GIS technology. If so, this will facilitate the determination of high-risk populations and consequently, enhance TB control programs. This study aimed to identify hot spot geographical areas with PTB incidence and to evaluate spatial global autocorrelation by using GIS technology.
| Patients and Methods|| |
This is a retrospective study in which seven years data of PTB patients were recorded and analyzed for spatial distribution of PTB diseas in Kerbala.
Kerbala governorate is located in South-central Iraq (about 100 Km2 far from Baghdad, Capital of Iraq). The governorate size is about 5034 km2. Estimated number of population is approximately 1,012,356 million people. Urban-Rural Density of population is 66.5%–33.5%. Kerbala is divided into three districts; Kerbala, Ain Al-Tamur and Al-Hindiya, [Figure 1]. Kerbala is classified into five Sectors, each sector divided into several Quarters. The population density in each quarter is very heterogeneous.
The governorate has 32 main primary health care centers and 23 subcenters. The climate of Kerbala is very hot and dry during Summer and very cold during Winter.
Using tuberculosis register software, cases of PTB were recorded by Chest and Respiratory illnesses center in Kerbala governorate. Seven years' data records were used to investigate hot spots and spatial distribution of PTB cases in Kerbala. Patients whom reside outside Kerbala were excluded from the current analysis.
| Methods|| |
Loess smoothing Seasonal decomposition trend was applied to analyze temporal patterns and clusters mapping of Tuberculosis, using R statistical software, version 3.5.1 (R Foundation for Statistical Co., Vienna, Austria). Correlation analysis (pairwise Spearman) was used to observe the association among the factors contributing to PTB prevalence. However, only variables with a coefficient of correlation >0.7 are included in the model. Standard morbidity ratio (SMRs) was used to find the morbidity rate for each geographical area (i.e., the number of observed PTB cases divide by the number of expected PTB cases). The expected cases were derived from the multiplying the overall incidence and population average of each of the PTB area considered in this study. Bayesian conditional autoregressive (CAR) model was used in the analysis and estimation of the parameters was done through Markov Chain Monte Carlo methods to estimate the space mapping variability cluster of disease risk and covariates effect. With Bayesian approach applied in the study, Poisson model was used to assess the spatiotemporal variation of PTB incidence and factors associated with PTB. Nonspatial and spatial Bayesian models were constructed. The models assumed that the observed PTB cases followed a Poisson distribution as:
Where Yij is the number of observed PTB cases in area i and month j, eij is the number of expected PTB cases in area i and month j, and rij is the Standard morbidity ratio risk in area i and month j.
We define the non-spatial model as:
The spatial model is defined as:
where ϕ is the intercept from the regression model, βp are the coefficients of regression, Xpij are the covariates, Ψιrepresents the unstructured random effects, µi represents the spatial structured of random effects,ωΨ represents the autoregressive effects and t represents the coefficients of the area average temporal trend.
| Results|| |
Temporal trend of tuberculosis
There were 525 cases of TB used in this study from January 1, 2010, to December 31, 2016. Among them (345, 65.8%) were smear positive, Others (180, 34.3%) were smear negative and confirmed PTB cases either by culturing, GeneXpert, or physical examination. The number of male patients was 275 (52.4%), and female patients was 250 (47.6%) with male to female ratio was (1.1). The mean and median age of the TB patients were 40.73 ± 17.52 and 38, respectively [Table 1]. Most of the cases were distributed within the age of 17–50 years (68.4%). The maximum days for the TB laboratory confirmatory was 31 days and the minimum was 1 day.
[Table 2] shows the correlation among the biographic factors of PTB prevalence. It is observed that both the age and the gender were having correlation coefficients >|0.7| and were included in the models.
A significant correlation was found between PTB prevalence and the age (r = 0.731) and was also found in gender (r = 0.822). Both the age and gender were included in the subsequent model.
The monthly prevalence of PTB cases is shown in [Figure 2] and indicates that the major peak of PTB cases appeared between March and May while the minor peak were observed between October and December.
We summarized the deviance information criterion (DIC)values in [Table 3] for the two models approaches. The spatiotemporal model was found to with lowest DIC, indicating that it performed better. We, therefore, select the spatiotemporal model and biographic factors as covariates in the model.
The risk association between the covariate factors and the PTB prevalence is summarized in [Table 4], which indicated that they are all positively significantly associated with PTB risk. The effect of age and gender were significant with RRs of 3.41 (95% confidence interval [CI]: 2.37, 4.60) and 1.23 (95% CI: 1.88, 3.06) per unit increase, respectively.
|Table 4: Bayesian poisson model for the association between covariate factors and tuberculosis|
Click here to view
Case detection rates estimation
[Table 5] represents the estimated and case detection rate of PTB incidence in Kerbala, Iraq. The median estimated PTB incidence with 95% credible interval was 397 (95% Crl: 255, 492) per 100000 in 2010 and 354 (95% Crl: 201, 438) per 100000 in 2016. The estimated incidence is widely differing from the case detection rate, with an estimated 61.3 in 2016–41.7 in 2010 related to the notification rates and case detection rate.
Spatial autocorrelation analysis
We performed Global Moran's I statistics to evaluate spatial global autocorrelation in the data, which were summarized in [Table 6]. It obviously indicates that the Tuberculosis cases were not randomly distributed in which the variables occur with asymptotic probabilities with unpredictable spacing and that there was the presence of high global autocorrelation among PTB cases in the City of Kerbala. The results of local autocorrelation analysis could not be mapped becuase there was no special information for it.
|Table 6: Spatial autocorrelation test results on Tuberculosis incidence in Kerbala City|
Click here to view
Hot spots mapping
[Figure 3]a, [Figure 3]b, [Figure 3]c, [Figure 3]d, [Figure 3]e, [Figure 3]f, [Figure 3]g is a satellite image of Kerbala province. The Global Positioning System coordinates used throughout the period from 2010 to 2016 were monitored and added as a layer to the maps. Adding a further layer of physical structures allowed these clusters to be matched to their corresponding area for PTB incidence and identified by those familiar with the province. All the points were subjected to a hot spot analysis using an embedded analytical tool in ArcGIS and the result revealed that several locations were corresponded to PTB areas and GIS data points for all the patients. This generated an image of clusters of high PTB spot from surrounding space. From the public buildings with high PTB spot identified in [Figure 3]a, [Figure 3]b, [Figure 3]c, [Figure 3]d, [Figure 3]e, [Figure 3]f, [Figure 3]g, those areas were found to be a hot spot with a high degree of confidence.
|Figure 3: (a-g) Satellite mapping cluster in Kerbala province from 2010 to 2016|
Click here to view
| Discussion|| |
Tuberculosis is considered as a global emergency as it ranked by the WHO, the seventh most morbidity causing disease and expected it to continue in the same position up to 2020. Unless properly treated, PTB patient can infect 10–15 people in a year. Identification of areas with a high prevalence of a disease represents the basic problem in geographical surveillance. The detection of these areas may be highly useful in the surveillance of this disease. This study aimed to determine the spatial distribution of PTB cases.
A significant positive correlation was found between PTB prevalence and the age and was also found with gender. It has been reported that PTB incidence increases with age. Unfortunately, most of the cases in the current study were distributed within the age of 17–50 years (68.4%). The presence of high percentage of disease among active younger individuals (reproductive age) and living in crowded low socioeconomic regions, smoking, and malnutrition might reflect higher rates of transmission of bacteria rather than reactivation of the disease. In addition, the monthly prevalence of PTB cases [Figure 2] indicated that the major peak of PTB cases appeared between March and May. Several previous studies, including our previous study, had been documented the presence of spring peaks in PTB incidence.,, Cold weather and indoor condition in addition to little sunlight exposure and Vitamin D deficiency may affect the immune system and enhance both transmissibility and reactivation of the disease.
PTB incidence is thought to be an indicator for both evaluations of the epidemiological situation and calculation of case detection rates which considered target for the WHO Stop TB program., Unfortunately, in many settings, case detection rates cannot be considered complete. Therefore, the number of notified PTB cases is lower than that of real incident cases within the population. In the current study, although calculated case detection rates seem to increase, the estimated PTB incidence is found to be different from case detection rates during the studied period [Table 5]. The presence of smear false negative cases, low access of the patients to the health-care centers and/or registration of these patients in other health centers belong to other governorates of Iraq might be responsible for the difference between the estimated incidence and case detection rates in Kerbala, Iraq.
A crucial factor in PTB control program is the determination of hot spot areas and therefore blocks the transmission of the disease. To assess the presence of spatial global autocorrelation, Global Moran's I statistics were applied, which gives a score ranging from −1 to 1. The resulted score of zero indicates the acceptance of the null hypothesis. Whereas, the positive scores represent the presence of clustering, and the negative score represents no clustering of PTB cases. The current study showed that tuberculosis cases were not randomly distributed in which the variables occur with asymptotic probabilities with unpredictable spacing and that there was the presence of high global autocorrelation among PTB cases in the City of Kerbala. IN addition, The Global Positioning System coordinates used throughout the studied period were monitored and added as a layer to the Satellite image of Kerbala province. This allowed PTB clusters to be matched to their corresponding area. Hot post analysis using an embedded analytical tool in ArcGIS revealed the presence of hot spot area with high degree of confidence. This means that there were more PTB cases observed in these areas. Moreover, most of PTB cases (58.5%) was seen to be aggregated in seven quarters within Kerbala Province, Al-Amil (60, 11.4%), Al-Gadeer (56, 10.7%), Al-Ascary (54, 10.3%), Al-Hindia (39, 7.4%), Al-Husseiniya (34, 6.5%), Al-Abbasia (33, 6.3%), Al-Hur Discret (31, 5.9%), (data not shown). These results might possibly reflect the presence of unknown environmental and socio-economic factors responsible for the continuous presence of PTB cases during the studied period, and therefore, these areas require great attention in the future.
The current study has some limitation including: the absence of official census during the studied period for each quarter. Second; because this is a retrospective study, the residential addresses of each PTB patients were not geocoded. However, only the resident main quarters in the province were geocoded.
| Conclusion|| |
GIS usage in studying of distribution of PTB cases reveals useful information about the epidemiological situation of PTB cases in Kerbala Province, Iraq. In addition, this study predicted possible places for PTB transmission based on the hot spot analysis and the continuous presence of infection during the studied period, and at the same time, Moran's I statistics revealed the presence of positive spatial global autocorrelation in PTB cases distribution. This may help future strategies in PTB control program.
We are grateful to the Advisory Clinic for Chest and Respiratory Illnesses-Holy Kerbala Province for their help in data collection.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Murray CJ, Lopez AD. Alternative projections of mortality and disability by cause 1990-2020: Global burden of disease study. Lancet 1997 349:1498-504.
Zumla A, Oliver M, Sharma V, Masham S, Herbert N. World TB Day 2016--advancing global tuberculosis control efforts. Lancet Infect Dis 2016;16:396-8.
Gallagher TC, Andersen RM, Koegel P, Gelberg L. Determinants of regular source of care among homeless adults in Los Angeles. Med Care 1997;35:814-30.
Shi L, Stevens GD. Vulnerability and unmet health care needs. The influence of multiple risk factors. J Gen Intern Med 2005;20:148-54.
Uphold CR, Mkanta WN. Review: Use of health care services among persons living with HIV infection: State of the science and future directions. AIDS Patient Care STDS 2005;19:473-85.
Centers for Disease Control and Prevention. From the centers for disease control and prevention. Tuberculosis morbidity among U.S.-Born and foreign-born populations – United States, 2000. MMWR Morb Mortal Wkly Rep 2002;51:101-4.
Tuberculosis elimination revisited: Obstacles, opportunities, and a renewed commitment. Advisory council for the elimination of tuberculosis (ACET). MMWR Recomm Rep 1999;48:1-3.
Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus statement. Global burden of tuberculosis: Estimated incidence, prevalence, and mortality by country. WHO global surveillance and monitoring project. JAMA 1999;282:677-86.
Ibrahim SA. Comparing alternative methods of measuring geographic access to health services: An assessment of people's access to specialist hospital in Kebbi state. Acad J Interdiscip Stud 2013;2:109.
Zulu LC, Kalipeni E, Johannes E. Analyzing spatial clustering and the spatiotemporal nature and trends of HIV/AIDS prevalence using GIS: The case of Malawi, 1994-2010. BMC Infect Dis 2014;14:285.
Kulldorff M, Nagarwalla N. Spatial disease clusters: Detection and inference. Stat Med 1995;14:799-810.
Ibrahim SA, Hamisu I, Lawal U. Spatial pattern of tuberculosis prevalence in Nigeria: A comparative analysis of spatial autocorrelation indices. Am J Geogr Inf Syst 2015;4:87-94.
Moonan PK, Bayona M, Quitugua TN, Oppong J, Dunbar D, Jost KC Jr., et al.
Using GIS technology to identify areas of tuberculosis transmission and incidence. Int J Health Geogr 2004;3:23.
Tiwari N, Adhikari CM, Tewari A, Kandpal V. Investigation of geo-spatial hotspots for the occurrence of tuberculosis in Almora district, India, using GIS and spatial scan statistic. Int J Health Geogr 2006;5:33.
Ratovonirina NH, Rakotosamimanana N, Razafimahatratra SL, Raherison MS, Refrégier G, Sola C, et al.
Assessment of tuberculosis spatial hotspot areas in Antananarivo, Madagascar, by combining spatial analysis and genotyping. BMC Infect Dis 2017;17:562.
Tiwari N, Kandpal V, Tewari A, Rao KR, Tolia V. Investigation of tuberculosis clusters in Dehradun city of India. Asian Pac J Trop Med 2010;3:486-90.
Couceiro L, Santana P, Nunes C. Pulmonary tuberculosis and risk factors in Portugal: A spatial analysis. Int J Tuberc Lung Dis 2011;15:1445-54, i.
Onozuka D, Hagihara A. Geographic prediction of tuberculosis clusters in Fukuoka, Japan, using the space-time scan statistic. BMC Infect Dis 2007;7:26.
Randremanana RV, Sabatier P, Rakotomanana F, Randriamanantena A, Richard V. Spatial clustering of pulmonary tuberculosis and impact of the care factors in Antananarivo city. Trop Med Int Health 2009;14:429-37.
Nunes C. Tuberculosis incidence in Portugal: Spatiotemporal clustering. Int J Health Geogr 2007;6:30.
Munch Z, Van Lill SW, Booysen CN, Zietsman HL, Enarson DA, Beyers N. Tuberculosis transmission patterns in a high-incidence area: A spatial analysis. Int J Tuberc Lung Dis 2003;7:271-7.
Touray K, Adetifa IM, Jallow A, Rigby J, Jeffries D, Cheung YB, et al.
Spatial analysis of tuberculosis in an urban West African setting: Is there evidence of clustering? Trop Med Int Health 2010;15:664-72.
Maciel EL, Pan W, Dietze R, Peres RL, Vinhas SA, Ribeiro FK, et al.
Spatial patterns of pulmonary tuberculosis incidence and their relationship to socio-economic status in Vitoria, Brazil. Int J Tuberc Lung Dis 2010;14:1395-402.
Jia ZW, Jia XW, Liu YX, Dye C, Chen F, Chen CS, et al.
Spatial analysis of tuberculosis cases in migrants and permanent residents, Beijing, 2000-2006. Emerg Infect Dis 2008;14:1413-9.
Al Mousawi A, Alwash H. Tuberculosis program health care workers knowledge about tuberculosis in Kerbala governorate in 2017. Iraqi J Public Health 2017;1:61-4.
India T. RNTCP Status Report. Central TB Division, Directorate of General of Health Services. New Delhi: Ministry of Health and family Welfare; 2005.
Comstock GW, Livesay VT, Woolpert SF. The prognosis of a positive tuberculin reaction in childhood and adolescence. Am J Epidemiol 1974;99:131-8.
Mohammed SH, Ahmed MM, Al-Mousawi AM, Azeez A. Seasonal behavior and forecasting trends of tuberculosis incidence in Holy Kerbala, Iraq. Int J Mycobacteriol 2018;7:361-7.
] [Full text]
Thorpe LE, Frieden TR, Laserson KF, Wells C, Khatri GR. Seasonality of tuberculosis in India: Is it real and what does it tell us? Lancet 2004;364:1613-4.
Schaaf HS, Nel ED, Beyers N, Gie RP, Scott F, Donald PR, et al.
Adecade of experience with mycobacterium tuberculosis culture from children: A seasonal influence on incidence of childhood tuberculosis. Tuber Lung Dis 1996;77:43-6.
Dye C, Maher D, Weil D, Espinal M, Raviglione M. Targets for global tuberculosis control. Int J Tuberc Lung Dis 2006;10:460-2.
Avilov KK, Romanyukha AA, Borisov SE, Belilovsky EM, Nechaeva OB, Karkach AS. An approach to estimating tuberculosis incidence and case detection rate from routine notification data. Int J Tuberc Lung Dis 2015;19:288-94, i-x.
Lai PC, So FM, Chan KW. Spatial Epidemiological Approaches in Disease Mapping and Analysis. Boca Raton: CRC Press; 2008.
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]