J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 310 http://jad.tums.ac.ir Published Online: September 30, 2019 Original Article Prone Regions of Zoonotic Cutaneous Leishmaniasis in Southwest of Iran: Combination of Hierarchical Decision Model (AHP) and GIS Elham Jahanifard1,2; Ahmad Ali Hanafi-Bojd1; Hossein Nasiri3; Hamid Reza Matinfar4; Zabihollah Charrahy5; Mohammad Reza Abai1; *Mohammad Reza Yaghoobi-Ershadi1; *Amir Ahmad Akhavan1 1Department of Medical Entomology and Vector Control, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran 2Department of Medical Entomology and Vector Control, School of Public Health, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran 3Faculty of Geography, University of Tehran, Tehran, Iran 4Department of Soil Science, Collage of Agriculture, Lorestan University, Khoramabad, Iran 5Open Training Center, School of Geography, Tehran University, Tehran, Iran (Received 3 Mar 2018; accepted 10 Apr 2019) Abstract Background: Cutaneous leishmaniasis due to Leishmania major is an important public health problem in the world. Khuzestan Province is one of the main foci of zoonotic cutaneous leishmaniasis (ZCL) in the southwest of Iran. We aimed to predict the spatial distribution of the vector and reservoir(s) of ZCL using decision-making tool and to pre- pare risk map of the disease using integrative GIS, RS and AHP methods in endemic foci in Shush (plain area) and Khorramshahr (coastal area) counties of Khuzestan Province, southern Iran from Mar 2012 to Jan 2013. Methods: Thirteen criteria including temperature, relative humidity, rainfall, soil texture, soil organic matter, soil pH, soil moisture, altitude, land cover, land use, underground water depth, distance from river, slope and distance from human dwelling with the highest chance of the presence of the main vector and reservoir of the disease were chosen for this study. Weights of the criteria classes were determined using the Expert choice 11 software. The presence proba- bility maps of the vector and reservoir of the disease were prepared with the combination of AHP method and Arc GIS 9.3. Results: Based on the maps derived from the AHP model, in Khorramshahr study area, the highest probability of ZCL is predicted in Gharb Karoon rural district. The presence probability of ZCL was high in Hossein Abad and Benmoala rural districts in the northeast of Shush. Conclusion: Prediction maps of ZCL distribution pattern provide valuable information which can guide policy makers and health authorities to be precise in making appropriate decisions before occurrence of a possible disease outbreak. Keywords: Decision model; Cutaneous leishmaniasis; Risk map; Iran Introduction Cutaneous leishmaniasis (CL), a neglected vector-borne disease, is an important public health problem in various parts of Iran. Two endemic forms of CL have been identified in the country: zoonotic cutaneous leishmaniasis (ZCL) due to Leishmania major and anthro- ponotic cutaneous leishmaniasis (ACL) due to L. tropica (1, 2). Iran is a high-risk country for leishmaniasis with approximately 20,000 re- ported cases annually, out of which about 80% are ZCL cases (3). Khuzestan Province is one of the main foci of ZCL in the southwest of Iran (4). In 2012, annual incidence of cutane- ous leishmaniasis was 9.5 to 21.7 per 100000 inhabitants in the province (5). This variation may be due to migration of non-immune per- sons to endemic areas, living people near ro- dent colonies, migration of rodent and increas- *Corresponding authors: Prof Mohammad Reza Yaghoobi-Ershadi, E-mail: yaghoobia@tums.ac.ir, Dr Amir Ahmad Akhavan, E-mail: aaakhavan@tums.ac.ir http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 311 http://jad.tums.ac.ir Published Online: September 30, 2019 ing synanthropic index for the reservoir and al- so lack of knowledge and attitude toward zo- onotic cutaneous leishmaniasis (6). Comprehensive epidemiological and mo- lecular studies in Iran have indicated that Phlebotomus (Phlebotomus) papatasi is the main vector of the disease (2, 7-9). The pres- ence of this sand fly has been reported in dif- ferent parts of Khuzestan Province (10-13). A high synanthropic index in this species in their natural habitats in Shush and Khorramshahr counties is reported (14). Natural promastigote infection of Ph. papatasi has been reported in Khuzestan (12). Moreover, L. major has been detected and isolated from this species in Rof- fyeh district of the province (15). Based on the distribution of rodent reser- voirs of the zoonotic disease, 4 foci of ZCL have been identified in the country. Leishma- nia major was detected in Tatera indica (Ro- dentia: Cricetidae) and Nesokia indica (Ro- dentia: Muridae) using PCR technique in the west and southwest of Khuzestan (16, 17). Risk factors such as development of agri- cultural projects, travelling to the endemic ar- eas, road construction, land cover, groundwater level, sand fly and rodent distribution, soil type, humidity, composting animal manures around homes, presence of secondary reservoirs such as dogs, socio-economic status, sleeping out- side, precipitation, altitude, temperature, and the species composition of sand flies have all contributed to the increase in the prevalence of zoonotic cutaneous leishmaniasis (4, 18, 19). Cross-sectional studies on the identification and detection of the Leishmania parasite in sand flies showed the distribution of the vector and the disease in different geographical areas of the world, however, the incidence of the dis- ease could not be predicted (20). Geo-envi- ronmental factors, which play important role in the spread of ZCL, are often neglected (21). Passive and active surveillances of the disease can be useful for finding and monitoring pa- tients in endemic areas, particularly in the in itial stages of an outbreak (22). The spatial distribution of the main vector of ZCL is mainly affected by relative humid- ity and temperature (23). Leishmaniasis occur- rence is correlated with environmental varia- bles such as the El Nino phenomena, which is the effect of climate change on sand fly and reservoir population, parasites and emergence and transmission of leishmaniasis (24). Based on the effects of environmental factors on cu- taneous leishmaniasis, risk maps of the disease have been prepared in various regions of Iran (21, 25). Remote sensing (RS) and Geographical In- formation Systems (GIS) are computer based programs used for determining the presence and abundance of vectors and predicting risk map of vector-borne diseases over the last 25 years (26). RS technique provided outstand- ing results when used to evaluate the risk of arthropod-borne diseases such as malaria, rift valley fever, West Nile, Lyme, rocky moun- tain spotted, leishmaniasis and onchocercia- sis. Satellite images may be used to determine environmental variables such as land use/land cover and temperature where data are other- wise not available (22). GIS combines software and hardware sys- tems to analyze, manage and display all spa- tial distributions in a map framework. GIS tech- nique has been used in some studies for the preparation of risk map of diseases and dis- playing hazardous zones (26-28). The GIS was used for health policy purposes due to the po- tential of estimating hospital requirement and health facilities (29). Moreover, GIS is able to determine the correlation between some fac- tors like land use, climatic variables, distance to health center, and malaria transmission (30). This technique has also been used for the study of biodiversity, the presence, and abundance of vector and vector-borne diseases with respect to time (26). Today, it plays an important role in the health sciences, and it is useful in un- derstanding and visualizing disease distribu- tion and epidemiological data (27). Application of GIS in spatiotemporal (space and time) ep- http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 312 http://jad.tums.ac.ir Published Online: September 30, 2019 idemiology research is useful in predicting dis- ease distribution in regions at risk (31, 32). Multiple linear regressions with GIS were used to predict malaria susceptibility zone (33). Analytic hierarchy process (AHP) tech- nique has been applied in conjunction with GIS and RS to determine at-risk regions in healthcare over the past years (34-36). GIS- based multicriteria decision analysis (MCDA) is the combination of geographical data and valuable judgments of experts that concentrates on solving spatial decision problems (37). The aims of this study were to determine the spatial distribution of vector(s) and res- ervoir(s) of ZCL using decision-making tools and to prepare risk map of the disease using integrative GIS, RS and AHP methods in two foci in Khuzestan Province (Shush and Khorramshahr counties). The results of this study can be helpful to health authorities in making precise decisions before disease out- breaks. Materials and Methods Study areas Khuzestan Province (29° 57ˈ and 33° 00ˈ N, 47° 40ˈ and 50° 33ˈ E) is one of the 31 provinces in Iran. It is located in the south- west of the country with 27 counties, 76 cit- ies, and 67 rural districts. This province shares border with Iraq and the Persian Gulf. The weather condition is generally warm, but some parts of the northeast have a temperate cli- mate. Three climate zones are identified in the province including mountain, desert, and semi- desert zones. Shush County (32° 11` 21” N, 48° 15ˈ 28” E) has a land area of 3577km2 and it is situated in the northwest of Khuzestan Province with hot and arid weather. The av- erage maximum and minimum temperatures are 46.9 °C and 9.5 °C, respectively. The av- erage annual rainfall in this county is about 180mm. Khorramshahr County (30° 26ˈ 21̎ N, 48° 10ˈ 45̎ E) is located in the southwest of the province and shares border with Iraq and the Persian Gulf. It is situated at a height of 3 meters above sea level. The weather is hot in summer but mild in winter. The mean max- imum and minimum temperatures were 47.5 °C and 9 °C respectively, and mean annual pre- cipitation of 150.6mm in 2013 (Fig. 1). How- ever, Khorramshahr and Shush Counties were regarded as plain and coastal areas in these en- demic study areas. Sampling sites were cho- sen based on villages had the most cases of cutaneous leishmaniasis in the last five years. Data collection and preparation Reservoir sampling The sampling of reservoirs was carried out during four seasons from Mar 2012 to Jan 2013. Rodents were captured mostly dur- ing the months of May, July, Sep, Dec, Jan and Feb. Thirty Sherman traps with cucumber, nut, date, bread and butter, Puff and tomato baits were set in natural, agricultural, semi ur- ban and urban ecotypes. Animals were pre- served by taxidermy technique and were iden- tified based on morphological characteristics (38). Locations of each of the sampling sites were obtained using GPS technique (Global Positioning System). Only T. indica and N. in- dica were included as layers in the Arc Map. These data were used for determining the ac- curacy of the proposed reservoir map. Vector sampling Vector sampling carried out monthly from 2012 to 2013. Sand flies were collected by 90 sticky traps (30 in human dwelling, 30 in sta- bles and 30 in rodent burrows or outdoor) from various ecotypes (Natural, agricultural, semi- urban and urban) in Shush and Khorramshahr. Trapping was conducted before sunset till sun- rise in the next morning and location coordi- nates were determined using GPS technique. Captured sand flies were preserved in 70% ethanol. Permanent microscopy slides of sand flies were prepared using Puri’s medium (39). All samples were identified using reliable keys http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 313 http://jad.tums.ac.ir Published Online: September 30, 2019 (40, 41). Only the collection points of the main vector (Ph. papatasi) were included in ArcMap as a layer to determine the accuracy of the model used for the estimation of presence prob- ability of the vector. Human infection and eco-environmental data Data on cases of human leishmaniasis due to L. major infection were collected from the Health Centers at Khorramshahr and Shush counties from 2004 to 2011. Indian Remote Sensing (IRS) satellite images, topographical, land use, river and soil maps, and point shape- file of the rural areas were obtained from the Iranian National Geographical Organization of Armed Forces. Distance from human dwell- ing and river layers were estimated using the buffer operation in ArcGIS. Climate variables were collected from the Meteorological Organ- ization of Khuzestan Province. The annual av- erages of precipitation, humidity and temper- ature raster layers were obtained by IDW (In- verse Distance Weighted model) surface anal- ysis of the data. These layers were then clipped based on the boundaries of the study area (Shush and Khorramshahr counties). Water in- formation was obtained from the Water Re- search Center of Iran. Databank in excel for- mat included water and disease cases used as layers in ArcGIS 10.2 software. ENVI (Environment for Visualizing Im- ages) software was employed for analyzing the IRS images. Normalized difference vegeta- tion index (NDVI) was calculated and used as land cover map in the modeling process. AHP model and processing The analytic hierarchy process (AHP) is a decision-making tool based on mathematics, pioneered by Thomas L. Saaty in the1970s (42). It has been applied in several decision- making scenarios such as choosing the best alternative, prioritizing or determining the rel- ative merit among a set of alternative, re- source allocation, bench markings and qual- ity management (43). This Multi-Criteria Decision Making (MCDM) technique consists of an ultimate goal, series of alternatives for reaching the goal and a group of criteria that evaluates the alternatives for reaching the goal. It helps deci- sion makers to evaluate factors by pairwise comparisons using standard scales (Table 1) (44). The numerical scale (1-9) reflects the im- portance of one factor relative to others. Com- parison with the numerical scale also shows which elements are more dominant over the others. Analytic hierarchy process can be summa- rized in three main steps: a. Arranging the elements in a hierarchy such that the goal of the decision making oc- cupies the top level, with criteria and alterna- tives occupying intermediate and lower lev- els, respectively. b. Establishing the matrices of pairwise comparison (judgment matrices) and determin- ing priorities among the elements. c. Checking judgment consistency assigned by the consistency ratio (CR). Saaty (42) sug- gested that CR less than 0.1 is acceptable whilst CR more than 0.1 needs judgment revision. Thirteen criteria including temperature, rel- ative humidity, rainfall, soil texture, soil or- ganic matter, soil pH, soil moisture, altitude, land cover, land use, distance from human dwellings, underground water depth, and dis- tance from rivers with a higher probability of presence and abundance of the main vector (Ph. papatasi) of ZCL were chosen, but slope element was used instead of distance from hu- man dwellings for the study conducted on the presence and abundance of the main reservoirs (T. indica and N. indica) in the study areas. Pairwise comparison matrix was designed based on thirteen criteria by Iranian leishman- iasis experts who have expertise in the ecolo- gy and biology of ZCL vector and reservoirs; climatic and environmental factors affecting in the spatial distribution of Ph. papatasi, T. indica and N. indica. The experts’ answers were based on Saaty’s pairwise comparison method http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 314 http://jad.tums.ac.ir Published Online: September 30, 2019 (Table 1). The weights of factors affecting the spatial distribution of vector and reservoirs of ZCL were determined by Expert Choice 11 software. The judgment matrixes were con- structed regarding sub-criterion (divisions of criteria) and completed using fundamental scale to pairwise comparison by Iranian leishman- iasis experts. Then, the matrix tables were an- alyzed by Expert Choice 11 software. Every layer was reclassified according to sub-crite- ria matrix analysis. Final maps were obtained by multiplying standard weight derived from experts’ ideas by weighted maps of the crite- ria. Hazard map of cutaneous leishmaniasis due to L. major was obtained by overlaying the probability maps of vector and reservoirs pres- ence in the study areas. The presence probability maps of vector and reservoirs e were divided into five classes using the natural breaks method in ArcMap (very low, low, moderate, high and very high), and only areas with high and very high risk were mapped as hazard zones. The high and very high strata were considered as the hot spots for ZCL transmission. Accuracy assessment The accuracy of the presence probability maps of the vector and reservoir were calcu- lated by overlaying the shape files of Ph. pa- patasi, T. indica and N. indica to the proposed maps in Arc Map. Only vector and reservoir sampling sites with high and very high classes, according to the prepared maps, were consid- ered and the results were reported as a per- centage. Results Comparison matrices of the thirteen criteria affecting the presence probability of the vec- tor and reservoirs in Shush and Khorramshahr counties are presented in Tables 2 and 3. The weight of each criterion is showed in Table 4. CR was less than 0.1 in all pairwise com- parison matrices. The experts’ opinion indicat- ed that the most effective factor affecting the spatial distribution of the ZCL vector is tem- perature followed by humidity and precipita- tion. On the other hand, the most important fac- tor affecting the presence of the reservoir was soil texture followed by land cover and land use. The weighted maps of layers like temper- ature, relative humidity, rainfall, soil texture, soil organic matter, soil pH, soil moisture, al- titude, land cover, land use, distance from hu- man dwelling, underground water depth, dis- tance from river and slope were prepared for both counties. The presence probability maps of the vector and reservoirs of each county were calculated in ArcMap by multiplying weight of the criteria in their weighted maps (Figs. 2 and 3). Through the combination of the pres- ence probability maps for Ph. papatasi, N. indica and T. indica, the ZCL risk maps in Shush and Khorramshahr Counties were de- rived (Fig. 4). Based on the maps derived from the AHP model, in Khorramshahr study area, the highest probability of vector existence is predicted in Gharbe Karoon and small part of Homeh Ghar- bi rural districts and the highest probability of reservoir presence and ZCL risk was found in Gharbe Karoon rural district. There was no risk of disease transmission in Homeh-Ghar- bi rural district. The presence probability of the vector was high in Sorkheh rural district in the north of Shush, but the highest probability of the ex- istence of this species was found in Hossein Abad and Benmoala rural districts in the north- east of Shush County. The presence proba- bility maps of the reservoirs showed that the probability of reservoirs presence was high in Benmoala. However, the risk of ZCL was found to be higher in Hossein Abad and Benmoala rural districts. The accuracy of the models used for the estimation of presence probability of the vector and reservoirs were 90% and 75% respectively in the study areas of Shush (Fig. 5) and 80% http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 315 http://jad.tums.ac.ir Published Online: September 30, 2019 and 83.33%, respectively, in Khorramshahr County (Fig. 6). Comparison between the ZCL case dis- tribution and spatial distribution of the vector (Ph. papatasi) and reservoirs (T. indica and N. indica) in the two counties showed that ma- jority of the ZCL cases were reported from the high and very high strata of the presence probability of vector and reservoirs. In other words, the spread of cutaneous leishmaniasis due to L. major was directly related to vector and reservoirs’ distribution probabilities. Table 1. Fundamental scale to pairwise comparison (44) Intensity of importance Definition and explanation 1 Equal importance (two activities contribute to the objective) 3 Moderate importance(experience and judgment slightly favor one activity over another) 5 Strong importance (Experience and judgment strongly favor one activity over another) 7 Very strong importance (an activity is favored very strongly over another) 9 Extreme importance (the evidence favoring one activity over another is of the highest possible order of affirmation) 2, 4, 6, 8 Intermediate values between the two adjacent judgments (when comparison is needed) Fig. 1. Map of study areas in Khuzestan Province, southern Iran http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 316 http://jad.tums.ac.ir Published Online: September 30, 2019 Table 2. Comparison matrix of the criteria affecting the presence probability of the main vector (Phlebotomus papatasi) Criteria T e m p e r a tu r e R e la tiv e h u m id ity R a in fa ll S o il te x tu r e S o il o r g a n ic m a tte r S o il p H S o il m o istu r e A ltitu d e L a n d u se L a n d c o v e r D ista n c e fr o m h u m a n d w e llin g U n d e r g r o u n d w a te r d e p th D ista n c e fr o m r iv e r Temperature 1 2.605 3.728 5.194 5.194 4.384 2.702 5.674 4.02 2.477 1.431 3.347 3.126 Relative humidity - 1 2.724 2.569 3.987 4.555 2.569 4.241 3.758 2.239 1.037 3.936 4.02 Rainfall - - 1 3.728 3.987 4.416 1.393 4.241 1.644 1.246 0.591 3.882 3.707 Soil texture - - - 1 2.537 3.017 0.903 4.004 1.719 1.673 0.684 3.022 3.898 Soil organic matter - - - - 1 2.993 1.059 3.005 1.719 1.149 0.591 2.237 3.594 Soil pH - - - - - 1 0.506 1.516 0.422 0.384 0.476 0.974 1.105 Soil moisture - - - - - - 1 3.692 2.016 2.477 0.728 3.63 3.845 Altitude - - - - - - - 1 0.833 1.099 0.459 2.112 2.091 Land use - - - - - - - - 1 1.262 0.59 2.605 2.713 Land cover - - - - - - - - - 1 2 2.569 3.594 Distance from human dwelling - - - - - - - - - - 1 2.085 2.876 Underground water depth - - - - - - - - - - - 1 0.994 Distance from river - - - - - - - - - - - - 1 Table 3. Comparison matrix of the criteria affecting the presence probability of main the reservoirs (Tatera indica and Nesokia indica) Criteria T e m p e r a tu r e R e la tiv e h u m id ity R a in fa ll S o il te x tu r e S o il m o istu r e S o il o r g a n ic m a tte r S o il p H S lo p e A ltitu d e L a n d c o v e r L a n d u se U n d e r g r o u n d w a te r d e p th D ista n c e fr o m r iv e r Temperature 1 2.569 3.201 0.762 1.695 2.424 2.268 1.516 3.245 0.718 0.693 2.713 0.944 Relative hu- midity 1 1.585 0.703 0.725 1.461 1.398 1.431 1.741 0.578 0.509 1.719 0.616 Rainfall - - 1 0.520 1.059 0.922 1.563 1.246 1.888 0.328 0.509 1.046 0.631 Soil texture - - - 1 3.273 3.277 4.258 5.357 5.123 1.431 2.667 4.324 2.954 Soil moisture - - - - 1 1.974 1.585 3.438 3.227 0.544 0.474 1.572 0.668 Soil organic matter - - - - - 1 1.657 2.512 2.268 0.564 0.552 1.552 0.725 Soil pH - - - - - - 1 0.833 1.217 0.343 0.328 1 0.484 Slope - - - - - - - 1 1.38 0.333 0.340 1.585 0.525 Altitude - - - - - - - - 1 0.476 0.527 1.32 0.668 Land cover - - - - - - - - - 1 2.29 4.891 3.129 Land use - - - - - - - - - 1 5.619 3.064 Underground water depth - - - - - - - - - - - 1 0.53 Distance from river - - - - - - - - - - - - 1 http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 317 http://jad.tums.ac.ir Published Online: September 30, 2019 Table 4. The weight of each criterion affecting the presence of reservoirs and vector derived from AHP technique Criteria AHP Weight T e m p e r a tu r e R e la tiv e h u m id ity R a in fa ll S o il te x tu r e S o il o r g a n ic m a tte r S o il p H S o il m o istu r e A ltitu d e L a n d u se L a n d c o v e r D ista n c e fr o m h u m a n d w e llin g U n d e r g r o u n d w a te r d e p th D ista n c e fr o m r iv e r S lo p e T o ta l Vector 0.207 0.148 0.11 0.077 0.06 0.03 0.08 0.034 0.051 0.055 0.095 0.027 0.026 - 1 Reservoir 0.105 0.063 0.052 0.174 0.058 0.038 0.072 0.036 0.117 0.139 - 0.034 0.072 0.04 1 Fig. 2. Presence probability maps of (a) reservoir and (b) vector in Shush County, Khuzestan Province of Iran, 2014 Fig. 3. Presence probability maps of (a) reservoir and (b) vector in Khorramshahr County, Khuzestan Province of Iran, 2014 a b a b http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 318 http://jad.tums.ac.ir Published Online: September 30, 2019 Fig. 4. Risk map of zoonotic cutaneous leishmaniasis in (a) Khorramshahr and (b) Shush Counties, Khuzestan Prov- ince of Iran, 2014 Fig. 5. Accuracy of distribution probability maps of (a) reservoir and (b) vector based on analytic hierarchy process model in Shush County, Khuzestan Province of Iran, 2014 Fig. 6. Accuracy of distribution probability maps of (a) reservoir and (b) vector based on analytic hierarchy process model in Khorramshahr County, Khuzestan Province of Iran, 2014 a b a b a b http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 319 http://jad.tums.ac.ir Published Online: September 30, 2019 Discussion This study provides spatial distribution of the main vector and reservoirs and risk map of ZCL using decision-making tools and integrated GIS and RS in Shush and Khorramshahr counties in the southwest of Iran. Data on human leishmaniasis cases col- lected from the two counties show that most of the ZCL cases occurred in Benmoala, Hos- sein Abad, Shavoor and Sayedabbas rural dis- tricts of Shush and Homeh-Sharghi rural dis- trict of Khorramshahr, where there are many agricultural fields. The hazard maps of the dis- ease revealed that most of the cases were rec- orded in areas where there are agricultural lands. Human factors such as population pres- sure, urbanization and new agricultural pro- jects affect the distribution of leishmaniasis. Land use including agricultural lands, rivers, canals and irrigation water are suitable circum- stances for the growth of vectors and provide nesting habitats for reservoirs (45, 46). Sand flies prefer humid habitats for the laying of eggs, survival, and development of immature stages (47). Inland waters and streams affect the abundance and distribution of vectors. In- crease in sand fly abundance can be facilitat- ed by shortening the reproductive cycle due to an increase in surface moisture in areas such as inland waters and river banks. In the cen- tral part of Bihar in eastern India, inland wa- ter bodies increased the transmission risk of visceral leishmaniasis (VL) by providing suit- able breeding places for the vector Ph. argen- tipes (48). In the present research study, disease cases and spatial distribution maps of Ph. pa- patasi derived from a combination of AHP and GIS revealed that most of the cases due to L. major occurred in areas with high and very high presence probability of sand flies. A decrease in the depth of underground water provides suitable condition such as moisture, critical fac- tor that increases the population of sand flies. Sand fly activities are strongly associated with two environmental parameters; precipi- tation and temperature (49). Precipitation, mean temperature and slope are the main environ- mental variables that affect the distribution of the main ZCL vector in Iran, using the MaxEnt model (50). Increased vector abundance is due to the presence of infected rodent and increased soil moisture in the region. The epidemiology of CL is mostly related to the spatiotemporal distribution of the vector and reservoir (49). There is a correlation between the geograph- ical distribution of zoonotic cutaneous leish- maniasis and its main vector, and L. major has been frequently isolated from Ph. papatasi in various foci in the country (2). Several studies have been conducted on the effects of environmental factors on the dis- tribution and incidence of CL (51, 52). The risk map of visceral leishmaniasis was prepared in two districts in East Azarbaijan Province (Ahar and Kalaybar) using group decision-mak- ing tools and GIS (34). Fuzzy AHP model is widely used as a predictor of the prevalence of the disease with an accuracy of more than 80%. In a research study, among eight param- eters only dogs, nomads and altitude were the most effective factors affecting the incidence of VL. Two studies have been conducted on mapping hazard areas of ZCL in Esfahan and Golestan Provinces using the fuzzy AHP meth- od (21, 25). Altitude and land cover have a neg- ative influence on disease incidence. Temper- ature and relative humidity are the two main factors that directly correlate with the incidence of ZCL (21). Environmental factors that af- fect vector and reservoir population influence the spread and incidence of leishmaniasis. Moreover, distribution of cutaneous and vis- ceral leishmaniasis was found to be affected by temperature fluctuations (53). In the present research, variables such as soil texture, land cover, land use and temper- ature, in order of descending influence, were found to have the highest impact on the pres- ence probability of the reservoirs. The most ef- http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 320 http://jad.tums.ac.ir Published Online: September 30, 2019 fective criteria affecting the presence proba- bility of Ph. papatasi, based on experts’ judg- ment, were temperature, relative humidity and precipitation. There were some differences be- tween our results and that of some previous studies on leishmaniasis. The difference may be due to the different forms of the disease studied and the various criteria selected for the study. In some previous study, criteria such as average altitude, health center distance, pop- ulation and evaporation were chosen whilst ig- noring the two most important factors of ZCL cycle like vector and reservoir in the prepara- tion of hazard maps of the disease. Environ- mental elements like slope and seasonal pre- cipitation contribute to the spreading predic- tion of N. indica and T. indica (ZCL reservoirs), respectively, in Iran (17). The risk map of Kala-azar was prepared using geo-environmental factors such as land cover/land use, condition of vegetation, surface dampness, climate, illiteracy and unemployment rate of the inhabitants in the Vaishali District of Bihar in India (48). In the present investigation, the hazard map of ZCL was prepared using climate and envi- ronmental elements in Shush and Khorramshahr counties. According to the multi- criteria decision-making technique, the highest probability of the disease presence is predicted in Gharb-e- Karoon rural district of Khorramshahr whilst Sorkheh, Hoseinabad and Benmoala rural districts where the high-risk areas of ZCL in Shush County. Different multi- criteria decision analysis techniques for disease susceptibility mapping were compared (54). AHP model, with high accuracy, was the best among the tools used for the prediction of VL in the northwest of Iran and dengue virus dis- ease in Ecuador. Conclusion Using a combination of GIS, RS and AHP decision-making techniques, valuable risk maps of cutaneous leishmaniasis due to L. major were prepared while GIS and RS are judged only on the base of evidence and environmental factors, and there is no possibility of intervention by experts. Our study confirms the usefulness of AHP and GIS techniques in preparing the spa- tial distribution of Ph. papatasi, N. indica and, T. indica and the risk map of ZCL in areas where similar vector and reservoirs exist. Fur- thermore, it is possible to update the maps by entering new data and satellite images. In the present study, eco-environmental fac- tors were found to have a greater influence on the presence of vector and reservoirs, which can affect the incidence of ZCL. We recommend the use of Multi-Criteria Decision Making tech- nique for the preparation of risk maps of ZCL in other counties of the province. The proposed maps are visual tools that highlight areas with the potential for the disease transmission and/ or outbreak and the population at risk. Educat- ing and increasing the knowledge of the people on reservoir, vector, transmission and control of the disease is deemed necessary. In addition, hazard maps are valuable data for health author- ities in allocating budget for effective disease control programs and health care for the peo- ple living in high-risk zones. Acknowledgements The authors are grateful to responsible per- sons and staffs of health centers in Shush and Khorramshahr Counties. This survey was fi- nancially supported by Tehran University of Medical Sciences (TUMS); project No.90- 04-27-16311. The authors declare that there is no con- flict of interests. References 1. Yaghoobi-Ershadi MR, Akhavan AA, Zahraei-Ramazani AR, Javadian E, Mo- tavalli-Emami M (2000) Field trial for http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 321 http://jad.tums.ac.ir Published Online: September 30, 2019 the control of zoonotic cutaneous leish- maniosis in Badrood, Iran. Ann Saudi Med. 20: 386–389. 2. Yaghoobi-Ershadi MR (2012) Phlebotom- ine sand flies (Diptera: Psychodidae) in Iran and their role on Leishmania trans- mission. J Arthropod Borne Dis. 6: 1–17. 3. Yaghoobi-Ershadi MR, Hakimiparizi M, Zahraei-Ramazani AR, Abdoli H, Akha- van AA, Aghasi M, Arandian MH, Ranjbar AA (2010) Sand fly surveillance within an emerging epidemic focus of cutaneous leishmaniasis in southeastern Iran. Iran J Arthropod-Borne Dis. 4: 17– 23. 4. Nadim A, Javadian E, Mohebali M, Momeni AZ (2009) Leishmania Parasite and Leishmaniasis. Third ed. Markaz Nashr Daneshgahi Press, Tehran, Iran (In Per- sian). 5. Shirzadi MR, Esfahania SB, Mohebali M, Yaghoobi-Ershadi MR, Gharachorlo F, Razavia MR, Postigo JA (2015) Epide- miological status of leishmaniasis in the Islamic Republic of Iran, 1983–2012. East Mediterr Health J. 21(10): 736–742. 6. Shirzadi MR (2012) Cutaneous leishman- iasis control guideline in Iran. Zoonoses Control Department, Center for Com- municable Diseases Control, Ministry of Health and Medical Education, Tehran, Iran. Available at: http://vch.iums.ac.ir/uploads/ketab_sal ak.pdf. 7. Yaghoobi-Ershadi MR, Javadian E (1996) Seasonal variation of Leishmania major infection rates in sandflies from rodent burrows in Isfahan Province, Iran. Med Vet Entomol. 10: 181–184. 8. Yaghoobi-Ershadi MR, Akhavan AA (1999) Entomological survey of sand flies (Dip- tera: Psychodidae) in a new focus of zo- onotic cutaneous leishmaniasis in Iran. Acta Trop. 73: 321–326. 9. Karimi A, Hanafi-Bojd AA, Yaghoobi-Er- shadi MR, Akhavan AA, Ghezelbash Z (2014) Spatial and temporal distribution of phlebotomine sand flies (Diptera: Psy- chodidae), vectors of leishmaniasis, in Iran. Acta Trop. 132: 131–139. 10. Nadim A, Mesghali A, Javadian E (1974) Cutaneous Leishmaniasis in southern Iran. Colloq Int CNRS. 239: 215–218. 11. Jahanifard E, Navidpour S, Vazirianzadeh B (2009) Study on Phlebotominae of two big marshlands of Khuzestan Province, Iran. J Exp Zool India. 12: 407–708. 12. Javadian E, Messghali A (1974) Studies on CL in Khuzestan, Iran. Part I: The lep- tomonad infection of sand flies. Bull Soc Path Exot. 67: 513–519. 13. Kavarizadeh F, Vazirianzadeh B, Rassi Y, Jalali-Glusang A, Moravvej A (2013) A faunistic study of sand flies of Musian District, Southwestern of Iran. Pak J Zo- ol. 45: 549–554. 14. Jahanifard E, Yaghoobi-Ershadi MR, Akha- van AA, Akbarzadeh K, Hanafi-Bojd AA, Rassi Y, Sedaghat MM, Shirzadi MR, Karimi A (2014) Diversity of sand flies (Diptera, Psychodidae) in southwest Iran with emphasis on synanthropy of Phlebotomus papatasi and Phlebotomus alexandri. Acta Trop. 140: 173–180. 15. Vazirianzadeh B, Saki J, Jahanifard E, Zarean M, Amraee K, Navidpour SH (2013) Isolation and identification of Leishmania species from sand fly and rodents collected from Roffaye District, Khuzestan Province, southwest of Iran. Jundishapur J Microb. 6: e10025. 16. Sedaghat MM, Salahi Moghadam A (2010) Mapping the distribution of the important rodents reservoir in Iran. J Army Univ. 8: 210–223 (In Persian). 17. Gholamrezaei M, Mohebali M, Hanafi- Bojd AA, Sedaghat MM, Shirzadi MR (2016) Ecological niche modeling of main reservoir hosts of zoonotic cuta- neous leishmaniasis in Iran. Acta Trop. 160: 44–52. 18. Shirzadi MR, Mollalo A, Yaghoobi-Ershadi http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 322 http://jad.tums.ac.ir Published Online: September 30, 2019 MR (2015) Dynamic relations between incidence of zoonotic cutaneous leish- maniasis and climatic factors in Goles- tan Province, Iran. J Arthropod Borne Dis. 9: 148–160. 19. Oryan A, Alidadi S, Akbari M (2014) Risk factors associated with leishmaniasis. Trop Med Surg. 2: e118. 20. Gebre-Michael T, Mslone JB, Balkew M, Ali A, Berhe N, Hailu A, Herzi AA (2004) Mapping the potential distribu- tion of Phlebotomus martini and P. ori- entalis (Diptera: Psychodidae), vectors of Kala-azar in East Africa by use of geo- graphic information systems. Acta Trop. 90: 73–86. 21. Mollalo A, Khodabandehloo E (2016) Zo- onotic cutaneous leishmaniasis in north- eastern Iran: a GIS-based spatio-temporal multi-criteria decision-making approach. Epidemiol Infect. 144(10): 2217–2229. 22. Kalluri S, Gilruth P, Rogers D, Szczur M (2007) Surveillance of arthropod vector- borne infectious diseases using remote sensing techniques: a review. PLOS Pathog. 3(10): e116. 23. Cross ER, Newcomb WW, Tucker CJ (1996) Use of weather data and remote sensing to predict the geographic and seasonal distribution of Phlebotomous papatasi in southwest Asia. Am J Trop Med Hyg. 54: 530–536. 24. Salmon OS, Quintana MG, Mastrangelo AV, Fernandez MS (2012) Leishmani- asis and climate change-case study: Ar- gentina. J Trop Med. 2012: 601242. 25. Mesghati Amoli G (2011) GIS-based risk map analysis of leishmaniasis disease in Isfahan, Iran. In: IEEE symposium on business, engineering and industrial ap- plications (ISBEIA), 2011Sept 25–28, Bayview Hotel, Langkawi, Kedah, Ma- laysia, pp. 275–280. 26. Palaniyandi M (2012) The role of Re- mote Sensing and GIS for spatial predic- tion of vector-borne disease transmission: A systematic review. J Vector Borne Dis. 49(4): 197–204. 27. Fradelous EC, Papathanasia IV, Mitis D, Tsaras K, Kleisiaris CF, Kourkouta L (2014) Health based Geographic Infor- mation Systems (GIS) and their applica- tions. Acta Inform Med. 22: 402–405. 28. Rai PK, Nathawat MS, Mishra A, Singh SB, Onagh M (2011) Role of GIS and GPS in VBD mapping: A case study. J GIS Trens. 2(1): 20–27. 29. Rai PK, Nathawat MS (2013) GIS in Healthcare Planning: A Case Study of Varanasi, India. Forum Geografic. 12(2): 153–163. 30. Rai PK, Nathawat MS, Rai S (2014) Us- ing the information value method in a geographic information system and re- mote sensing for malaria mapping: a case study from India. Journal of Innovation in Health Informatics. 21(1): 43–52. 31. Abedi-Astaneh F, Hajjaran H, Yaghoobi- Ershadi MR, Hanafi-Bojd AA, Mohe- bali M, Shirzadi MR, Rassi Y, Akhavan AA, Mahmoudi B (2016) Risk mapping and situational analysis of cutaneous leishmaniasis in an endemic area of Cen- tral Iran: A GIS-based survey. PloS One. 11: e0161317. 32. Jacquez GM (2000) Spatial analysis in epi- demiology: Nascent science or a failure of GIS? J Geogr Syst. 2: 91–97. 33. Rai PK, Nathawat MS (2013) Application of GIS and statistical methods to select optimum model for malaria susceptibil- ity zonation: a case study. Analele sti- intifice ale Universitatii" Alexandru Ioan Cuza" din Iasi-seria Geografie. 59(2): 73– 94. 34. Rajabi M, Mansourian A, Bazmani A (2012) Susceptibility mapping of Vis- ceral Leishmaniasis based on fuzzy mod- eling and group decision-making meth- ods. Geospat Health. 7: 37–50. 35. Chikodzi D (2013) Spatial modeling of malaria risk zones using environmental, http://jad.tums.ac.ir/ J Arthropod-Borne Dis, September 2019, 13(3): 310–323 E Jahanifard et al.: Prone Regions of … 323 http://jad.tums.ac.ir Published Online: September 30, 2019 anthropogenic variables and geographical information system techniques. J Geosc Geom. 1: 8–14. 36. Bhatt B, Joshi JP (2014) Analytical hier- archy process modeling for malaria risk zone in Vadodara District, Gujarat. The 8th ISPRS Technical Commission Sym- posium, 2014 December 09–12, Hyder- abad, India, pp. 171–176. 37. Malczewski J (2006) GIS-based multicrite- ria decision analysis: a survey of the lit- erature. Int J Geogr INF Sci. 20: 703–726. 38. Etemad E (1978) Mammals of Iran, Vol. 1. Rodents and their Identification Keys. National Society of Guardianship of Nat- ural Resources and Human Environment, Tehran, Iran (In Persian). 39. Smart J, Jordan K, Whittick RJ (1965) In- sect of Medical Importance. Fourth ed. British Museum, Natural History, Ad- len Press, Oxford. 40. Theodor O, Mesghali A (1964) On the phlebotominae of Iran. J Med Entomol. 1: 285–300. 41. Seyed-Rashti MA, Nadim A (1992) The genus of Phlebotomus (Diptera: Psycho- didae: Phlebotominae) of the countries the Eastern Mediterranean region. Iran J Public Health. 21: 11–50. 42. Saaty TL (2008) Decision making with the analytic hierarchy process. Int J Services Sciences. 1: 83–98. 43. Bhushan N, Rai K (2004) Strategic Deci- sion Making: Applying the Analytic Hi- erarchy Process. Springer. 44. Saaty TL (1990) How to make a decision: The analytic hierarchy process. Eur J Oper Res. 48: 9–26. 45. Desjeux P (2001) The increase in risk fac- tors for leishmaniasis worldwide. Trans R Soc Trop Med Hyg. 95: 239–243. 46. Colwell DD, Dantas-Torres F, Otranto D (2011) Vector-borne parasitic zoonoses: emerging scenarios and new perspectives. Vet Parasitol. 182: 14–21. 47. Sharma U, Singh S (2008) Insect vectors of Leishmania: distribution, physiology and their control. J Vect Borne Dis. 45: 255–272. 48. Bhunia GS, Kesari S, Chatterjee N, Pal DK, Kumar V, Ranjan A, Das P (2011) Incidence of visceral leishmaniasis in the Vaishali district of Bihar, India: Spa- tial patterns and role of inland water bod- ies. Geospat Health. 5: 205–215. 49. Aytekin S, Ertem M, Yagdiran O, Ayte- kin N (2006) Clinico-epidemiologic study of cutaneous leishmaniasis in Diyarbkir, Turkey. Dermatol Online J. 12(3): 14. 50. Hanafi-Bojd AA, Yaghoobi-Ershadi MR, Haghdoost AA, Akhavan AA, Rassi Y, Karimi A, Charrahy Z (2015) Model- ing the distribution of cutaneous leish- maniasis vectors (Psychodidae: Phlebotominae) in Iran: a potential trans- mission in disease prone, reservoir are- as. J Med Entomol. 52: 557–565. 51. Ali-Akbarpour M, Mohammadbeigi A, Tabatabee SHR, Hatam GH (2012) Spa- tial analysis of eco-environmental risk factors of cutaneous leishmaniasis in southern Iran. J Cutan Aesthe Surg. 5: 30–35. 52. Florez MP, Ocampo CB, Ardila CV, Al- exander N (2016) Spatial modeling of cutaneous leishmaniasis in the Andean region of Colombia. Mem Inst Oswal- do Cruz. 111: 433–442. 53. Ozbel Y, Balcioglu C, Kirami Olgen M, Simsek FM, Ozensoy Toz S, Ertabak- lar H, Demir S, Alkan Z (2011) Spatial distribution of phlebotomine sand flies in the Aydin Mountains and surround- ings: the main focus of cutaneous leish- maniasis in western Turkey. J Vect Ecol. 36: S99–S105. 54. Dong SH (2016) Comparisons between different multi-criteria decision analysis techniques for disease. [Master thesis]. Department of Physical Geography and Ecosystem Science, Lund University, Sweden. http://jad.tums.ac.ir/