Cutaneous leishmaniasis in north-central Ethiopia: trend, clinical forms, geographic distribution, and determinants

Background Cutaneous leishmaniasis (CL), being among the neglected tropical diseases, catches little attention despite its considerable influence. This study aimed at estimating the prevalence and associated factors of CL in Boru Meda Hospital, Dessie town, north-central Ethiopia. Methods Medical records of patients who attended the Dermatology Department of the Hospital in 2012–May 2018 were assessed. In addition, dermatological patients who were visiting the hospital during the data collection period (November 2017–May 2018) were interviewed to capture socio-demographic, environmental variables, and related factors. The source population was individuals who visited the hospital for skin problems in the stated years and CL positives were the targets. The association between CL and its determinants was tested by logistic regression. Results CL prevalence was 1.5% showing increasing trend with the year of examination. Localized, diffused, and mucosal CL was evident across the years. Dessie town had the highest prevalence, 291 (32.8%) patients out of 888 cases. The number of examined (29,701) and positives (543, 1.8%) for males was comparable with females, 28,459 and 345 (1.2%), respectively, increasing with age but without significant difference. Dessie town residence (adjusted odds ratio (AOR) 12.2, 95% confidence interval (CI) 2.2–18.6, p = 0.01), no bed net (AOR 9.9, 95% CI 2.7–16.7, p < 0.01), nearby irrigation (AOR 8.1, 95% CI 1.9–12.4, p < 0.01), and travel to CL endemic areas (AOR 13.9, 95% CI 4.4–14.3, p < 0.01) were significantly associated with CL. Conclusion CL is a growing health problem in Dessie and its surroundings. Known risk factors prevail. Comprehensive parasitological, entomological, and social studies are warranted to better manage the disease.


Introduction
Leishmaniasis is a collective name of diseases caused by different species of the intracellular protozoa of the genus Leishmania [1] and transmitted by the bite of female phlebotomine sandflies of the genera Phlebotomus or Lutzomyia [2]. Human leishmaniasis is manifested in distinct clinical forms ranging from mild self-healing cutaneous lesions or nodules termed cutaneous leishmaniasis (CL) to a fatal systemic illness called visceral leishmaniasis (VL) or more commonly kala-azar. In between, there is mucocutaneous leishmaniasis (MCL) or "espundia" and post-kala-azar dermal leishmaniasis-PKADL [3]. MCL, which is primarily caused by Leishmania braziliensis braziliensis and most commonly existing in Bolivia, Brazil, and Peru, affects the mucosa membrane and if untreated may be fatal invariably [4]. VL caused by L. donovani complex is a febrile malariaresembling illness characterized by swelling of the lymph nodes, the liver, and spleen (hepatosplenomegaly) and aethiopica [35] though there were some cases by L. major and L. tropica [36].
Several people from Dessie town, north-central Ethiopia, and the surrounding areas visit Boru Meda Hospital (BMH) for skin diseases of multiple etiologies including CL. Furthermore, people with cured scars or active CL lesions are visible in schools, on the street, markets, and other places in the town. However, there is no recent information on the magnitude of CL in and around Dessie, except an attempt on the treatment efficacy of anti-leishmaniasis agents on a limited number of patients (n = 97) in BMH [34]. This study was aimed at assessing the prevalence of CL among the patients attending BMH and the associated risk factors. Medical reports, combined with future field works and molecular analyses as well as entomological surveys will help identify the status of local transmission of CL and its spatial and temporal distribution. Such an investigation will provide better insights into the problem and contribute towards designing sustainable CL management strategies in Dessie and its environs.

The study area
The study area is Dessie town in north-central Ethiopia some 388 km from Addis Ababa with latitude and longitude of 11°8′ N 39°38′ E and 2470-2550 m above sea level. Dessie was established by King Michael in 1893 currently covering an area of 16000 ha. It is divided into 10 sub-cities and 6 rural kebeles (the lowest administrative unit). The study was conducted in BMH. According to the information obtained from the Hospital, BMH was established in 1955 by the Sudan Interior Mission (SIM). During its establishment, it was intended that the hospital would provide ophthalmological and dermatological services. After serving for about 40 years as per its initial objective, the hospital was upgraded to deliver comprehensive medical services now. Around 2.5 million people living in South and North Wollo Zones, Oromia Special Zone, North Shewa, South Tigray, and parts of Afar Region are served by the hospital. The hospital remained the only well-recognized center for skin disease management in the eastern part of the Amhara Region and beyond.

Study design, population, and data collection
Patients who presented to BMH Dermatology Department between October 2017 and May 2018 were recruited and interviewed to gather important socio-demographic and other related information, and their CL status was determined clinically and/or parasitologically (Giemsa stains of skin slit smear or fine-needle aspiration cytology (FNAC)/ skin biopsy) as well as culture. The questionnaire was adopted from the literature and used with slight modification. The original English version of the questionnaire was translated first into Amharic and back into English to assure consistency. Then, it was administered to patients who came to the hospital for CL and other skin disease treatment. The collected data were checked for completeness, accuracy, clarity, and consistency. Additionally, the medical records of individuals who visited the hospital in the preceding years (January 2012-September 2017) for the same problem were considered and CLpositive cases (clinical/parasitological/culture) identified and socio-demographic data extracted.

Data analysis
While the dependent variable was CL positivity, the independent variables constituted socio-demographics such as age, sex, knowledge, residence area, and related household and environmental factors. The differences between age and gender categories as well as residence, which were the only available variables for the retrospective data, with regard to CL have been tested by the chi-square test. Univariable and multivariable logistic regression tests at 95% confidence intervals (CI) were used to assess the association between the dependent and independent variables for the interviewed patients only.
Statistical significance was at p < 0.05. The statistical analysis was carried out using the Statistical Package for Social Sciences (SPSS IBM Statistics) version 20.

CL cases by age and gender
In the past six and half year (January 2012-May 2018), totally 58,163 people were examined in the Dermatology Department of BMH. Of these, 29,704 were males and 28,459 females (male to female ratio nearly 1:1). The total prevalence of CL was 1.5% with 543 (1.8%) males and 345 (1.2%) females ( Table 1). The rate of CL increased from 0.9% in 2012 to 3.5% in 2018. Overall as well as throughout, slightly more males were examined for skin problems except for the slightly higher number of females in 2016 and 2017. Among the CL positives (n = 888), 543(61.1%) were males and 345(38.9%) females. There was a slightly higher prevalence of CL positivity in males than in females in all the years considered although the difference was not statistically significant. The proportion of CL-positive individuals was different in different age groups for both genders. In general, there was an increasing trend of CL cases with age and in both genders; those over 14 years were the most affected. While 3.4%, 11.6%, and 46.2% of under-5, 5-14 and > 14-year-old males were CL-positive, the corresponding proportions of females were 2.7%, 9.5%, and 26.7% (Table 2).

Patient place of residence
The CL cases in the Hospital were from about thirtythree districts including Dessie town. Nineteen of these are South Wollo Zone districts; all the districts in the Zone had CL cases. The remaining thirteen are from North Wollo (9), North Shoa (2), and Oromia Special Zone (2) which share borders with South Wollo. One district having three CL cases belongs to Tigray Region and is located at North Wollo boundary. Quantitatively, 789 (89.0%) of the CL cases were from South Wollo, 74 (8.0%) from North Wollo, 13 (2.0%) North Shewa, and 9 (1.0%) from Oromia. The proportion of CL cases from Dessie town was the highest (291 (33.0%) followed by Kutaber (183 (21.0%) and Tehulederie (72 (8.0%) that are within the immediate vicinity ( Fig. 1).

Repeated and new CL cases, type, and location of lesion
Among patients of CL who were examined in the Hospital, 521(58.7%) were new cases -318 (58.6%) males and 203 (58.8%) females. The rest 367(41.3%) were previously treated cases. That is, males and females who took treatment more than once in the Hospital for CL were respectively 225 (61.3%) and 142 (38.7%). The number and proportion of newly (once) and repeatedly treated (relapse) individuals in the hospital had shown The data showed that more repeat-cases were observed in 2017 and 2018 than the rest of the years ( Table 3). Out of the 888 patients, 792 (89.2%) were localized cutaneous leishmaniasis (LCL), 35 (3.9%) diffuse cutaneous leishmaniasis (DCL) and 61(6.9%) were MCL cases. The data showed that the majority of the cases were localized, particularly at the nose and cheek. The duration of lesions, treatment, and treatment outcomes was not available within patient medical records. Similarly, data for the four different diagnostics (clinical, skin slit smear, FNAC, and culture) that are reported practiced in the hospital were not available.

Questionnaire data
The questionnaire data showed that out of 354 respondents, 216 (61.0%) were males and 138 (38.9%) females. Twenty-four (6.8%) of the respondents were under-5 children, 52 (14.7%) 5-14 years and the rest 278 (78.5%) over 14 years. Most (141) of the respondents were secondary school students, 89 at primary school, 72 joined college and university education, and the rest 52 had no Among the variables tested in the univariate analysis most were significantly associated with CL positivity except gender, age, sleeping under the tree, and being a merchant or living outside Dessie town or Kutaber (Table 4).
In the multivariable model, Dessie town residence, lack of CL knowledge, living in muddy and grassy houses or cracked-wall houses, lack of bed net, irrigation nearby, possession of dogs or other animals, and travel history to CL endemic areas were significantly associated with CL occurrence (Table 5).

Discussion
The overall prevalence of CL among patients who visited the BMH Dermatology Department from 2012 to May 2018 was 1.5%. This is much lower than the prevalence reported by similar health facility-based studies in Ethiopia such as Addis Ababa 14.2% [33]. Although the nationwide prevalence of CL is lacking in Ethiopia, there are some community-based cross-sectional surveys showing 2.3-14.0% prevalence [15,23,24,27] from various parts demonstrating remarkable heterogeneity. A meta-analysis and systemic review [37] reported a pooled prevalence of 19.0%. The relatively lower prevalence of CL in the study area might be due to a number of reasons. Previously, some people might follow traditional medication options. Others used to visit the hospital, diagnosed but referred to other hospitals where leishmaniasis treatment service is readily available. However, recently, the hospital has started serving CL patients as any other skin disease cases (personal communication, hospital staff). During the time of the study, several people were attending the hospital for CL treatment from different nearby as well as distant areas. The increased prevalence of CL cases in 2017 and up to May 2018 corroborates this explanation. The CL prevalence increased from 0.9% in 2012 to 3.5% in 2018 to May.
The percentage of CL-positive males was 1.8% which was comparable with females (1.2%) agreeing with a report from different localities of Ethiopia at different times [19,24,30], and in some countries like Saudi Arabia in the Middle East [38] where no significant gender disparity was recorded. Further, a study in Argentina from South America [39] reported no significant difference between males and females suggesting equal exposure to infection.
On the other hand, a number of findings indicated that the male gender is a risk factor for CL in different countries (reviewed in ref no. [38,40], and [41] including Ethiopia [22]). The authors attribute this to variations in the activities of males and females and thus exposure to the sandfly vectors, provided there is no significant difference in gender attendance to health services. Males may have closer contact with the habitat of the sandfly through occupation and leisure activity. Cultural activities that mostly males are faced with outdoor activities including farming, keeping cattle, staying around gorges, and/or farmland for a long period and the presence of  endemic sites that mostly males could travel there for work. Furthermore, the gender difference in CL incidence is attributable to sex hormonal effects or immune responses [42], as it has been noted in some other parasitic diseases [43]. Differences in exposure or access to healthcare per se do not necessarily explain genderbased differences for infection or clinical disease. The role of gender in reference to infectious diseases is demonstrated in an animal model as well [44]. In general, there is a strong correlation between gender and incidence of leishmaniasis and the disease is more frequent among men than women although the current study did not suggest that.
The secondary data collected from the hospital records showed that the total percentage of CL positive individuals was different between age groups. Out of 888 CLpositive individuals, 6.1%, 21.1%, and 72.9% were under-5, 5-14, and ≥ 15, respectively, although the number of examined in each group is unknown. The primary data also revealed an almost similar pattern in relation to age although in the logistic regression age was not identified as a significant risk factor. This might be due to the outdoor activities performed by individuals of this age group in farms and the likes than individuals below this age as they are actively working groups.
Nevertheless, the age-related pattern of CL attack is variable across studies in Ethiopia. In some, the more predominantly affected group is the younger 0-9 [24], in others, the 11-20 [27], > 14 [22], or those 10-19 years old [23]. The data showed that the majority (89.2%) of the CL cases were localized, particularly at the nose and cheek. Diffuse and mucocutaneous accounted only for 3.9% and 6.9%, respectively. It is known that sandfly vectors bite in the face, which is exposed, and cutaneous lesions appear in the site of promastigote inoculation following the bite. In other localities and countries, the face is noticed as the most infected organ by the Leishmania although some other studies showed leishmaniasis lesions in the hands and feet as well [45].
The current hospital data indicates that CL cases were observed in 35 different districts including Dessie town. There was a greater variation of CL prevalence amongst the different districts. The data showed that the highest CL prevalence was in Dessie town. The high prevalence of CL in Dessie town and Kutaber district could be due to the higher altitude of 2500 m above sea level together with other factors such as recent urbanization, especially in Dessie. CL is distributed mainly at high and mid altitudes ranging from 1400 to 2700 m above sea level, which is most favorable for the proven vector sandfly species. In a recent mapping study in Ethiopia [46], CL occurrence was significantly associated with an altitudinal range between 810 and 3563 m above mean sea level increasing slope values producing higher CL correlations. The results show the widespread endemicity of CL in Dessie town and its surroundings calling for future field-based epidemiological surveys. Previous reports concerning the endemicity of CL in Kutaber, which is some 10 km from Dessie town [18] and "Dessie area" [19], exist, but this report is the first of its kind for CL from Dessie town proper in recent times. The urbanization process of the town, which is a well-known risk factor for CL, is ongoing [47].
The data showed that the CL prevalence was highest for farmers (68.5%) and least among civil servants. Farming activity could put farmers at greater risk of CL than non-farmers such as government employees. In the farm field and irrigation areas, and possible travel to endemic areas and other seasonal activities farmers could be more exposed to sandfly bites. Students are second in CL prevalence because most students are families of farmers and engaged in similar activities as their parents.
Lack of CL knowledge, living in houses constructed of mud and grass, presence of cracks in house wall, sleeping on the floor, lack of using bed net during sleeping, presence of plants and irrigation areas nearby houses, presence of dogs and other animals in the compound or living houses, and travel history to endemic areas are significantly associated with CL. The responses of the participants on the knowledge-based questionnaire showed that the majority of the respondents were not aware of the meaning, transmission, and associated risk factors of CL, and this lack of knowledge was significantly associated with CL. This result is in agreement with reports and reviews by various authors [48][49][50][51].

Conclusion
In the study area, CL prevalence was progressively increasing within the past six and half year. Moreover, the study demonstrated that CL remained endemic in Dessie town and its surroundings without being well documented in the literature. Common CL risk factors incriminated elsewhere are prevalent in the study area.
The study revealed that the community lacked awareness about the meaning, transmission, and associated risk factors of CL. To better control and monitor CL, awareness creation is necessary. The hospital lacked organized documentation system about CL patients and requires improvement particularly in the specific diagnostics ordered for a particular patient, drugs prescribed, and treatment outcomes as well as detailed clinical features, for instance, the duration of the lesions including whether dry, wet, painful, or painless. Comprehensive future studies on the epidemiology and public health burden of the disease in the area are essential.