Visceral leishmaniasis elimination in Ethiopia: insights into epidemiology, transmission, treatment access, and outcome – a systematic review and meta-analysis

Abstract

Background

Ethiopia is a high-burden country for visceral leishmaniasis (VL) and has committed to regional elimination by 2030. Current evidence on disease epidemiology, transmission dynamics, treatment access, and clinical outcomes is essential for transitioning from control to elimination efforts. This systematic review and meta-analysis synthesizes the available literature on the prevalence, associated risk factors, accessibility and effectiveness of treatment, transmission patterns, clinical outcomes, and ecological factors influencing VL distribution, including vector habitats, climate conditions, and land-use patterns.

Methods

Literature searches were conducted using PubMed, SCOPUS, EMBASE, CINAHL, HINARI, ScienceDirect, Google Scholar, and grey literature sources following PRISMA 2020 reporting guidelines. Study quality was assessed using the Joanna Briggs Institute (JBI) tools. Random-effects meta-analyses were used to estimate pooled prevalence and associated risk factors. In addition to systematic review and meta-analysis, we conducted a descriptive analysis of WHO surveillance data to assess the trend of caseload, treatment access, and cured treatment outcomes using R software version 4.5.1 and ArcGIS version 10.7, respectively.

Results

The study revealed that pooled VL prevalence (32 studies, 103,962 participants) was 15.29% (95% CI: 9.25–22.51) in institution-based and 10.38% (95% CI: 5.38–16.74) in community-based settings, with substantial variation according to detection methods employed. Male sex (OR = 1.86), sleeping outdoors (OR = 2.85) or on the ground (OR = 1.92), and travel to endemic areas (OR = 3.99) were significantly associated with VL infection. Family history of VL (OR = 2.27) and rural residence (OR = 2.27) increased VL risk substantially. Environmental factors, including proximity to termite mounds (OR = 3.60), presence of acacia trees (OR = 3.75), and hyraxes near dwellings (OR = 1.90), were significantly associated with increased VL risk. Spatial analysis revealed persistent VL transmission with intermittent outbreaks concentrated in high and very high VL-risk areas, primarily in the Kolla and Lower Kolla agroecological zones. Geographic inequities in treatment access were evident, with higher case numbers treated in northern endemic foci and increasing trends in southern and eastern foci. The sodium stibogluconate and paromomycin combination regimen demonstrated high cure rates.

Conclusion

Visceral leishmaniasis (VL) in Ethiopia shows heterogeneous prevalence, highest with leishmanin skin tests. Male sex, farming, family history, outdoor sleeping, and proximity to termite mounds, acacia trees, and domestic animals were major risk factors. VL distribution aligned with sandfly vectors, particularly Phlebotomus orientalis, across lowlands and transitional highlands. National treatment data (2017–2024) showed generally favorable outcomes, highlighting the need for targeted elimination strategies based on local prevalence, risk factors, and effective case management.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-026-13275-w.

Introduction

Visceral leishmaniasis (VL, also known as Kala-azar) is a fatal neglected tropical disease affecting the world’s most vulnerable populations. Visceral leishmaniasis is caused by parasites of the Leishmania donovani complex, with Leishmania donovani being the primary causative agent in East Africa. The East African VL foci constitute a major global hotspot; in 2022, the region accounted for approximately 73% of reported global VL cases [1]. In Ethiopia, approximately 3.2 million individuals live at risk of VL, with 3,000 to 4,500 new cases reported annually, likely representing an underestimate due to limited active surveillance [2, 3]. The disease disproportionately affects children in rural and marginalized populations, perpetuating cycles of poverty, who are also nutritionally vulnerable, as reported in VL patient populations from Brazil, East Africa, and South Asia [4]. While VL is considered predominantly anthroponotic in Ethiopia, evidence for zoonotic transmission cannot be excluded [5–8]. Primary sandfly vectors include Phlebotomus orientalis, which predominates in northern endemic foci in Ethiopia [9–16], while Ph. celiae and Ph. martini [17] occur in southern foci in Ethiopia. Ph. alexandri and Ph. Rodhaini are considered secondary [14, 18]. The northern VL foci are associated with Acacia-Balanite dominated savannas and vertisols, a deep-cracking black cotton soil [19]. In contrast, the southern VL foci are linked to termite mound-associated habitats [13, 17, 20].

Environmental and socioeconomic factors substantially influence VL risk [21–23]. Behavioral and host-related factors have also been associated with elevated VL risk [2, 4, 12, 21, 24–27]. The arid and semiarid agroecological zones, semi-nomadic pastoralism, and irrigated agricultural schemes coincide with major VL foci, amplifying the disease’s socioeconomic impact [3, 25].

Fine-scale epidemiological data to guide effective interventions remain limited in most endemic regions [3, 28, 29]. The potential role of animal reservoirs, vector bionomics, and community-level disease dynamics are identified as gaps within the WHO framework for VL elimination as a public health problem in East Africa by 2030 [30]. Previous reviews on VL in Ethiopia and the eastern African region were limited by broad pooled prevalence estimates, minimal exploration of ecological, diagnostic, or spatial determinants of transmission, and lacked Ethiopia-specific analyses or integration of environmental, vector, and potential reservoir data [31]. These limitations restrict the ability to understand the heterogeneous transmission dynamics of VL across endemic areas and to guide targeted interventions for control and elimination. Furthermore, additional data from the World Health Organization were included to complement the systematic review findings by providing national-level information on treatment access and outcomes, which are often underreported in primary studies, thereby enabling a more comprehensive assessment of visceral leishmaniasis burden and control efforts.

This systematic review and meta-analysis synthesizes a larger and more recent body of evidence on the prevalence, associated risk factors, ecological and spatial determinants, and treatment access patterns across Ethiopia. By integrating epidemiological, ecological, and treatment data, this study provides a more comprehensive understanding of VL transmission patterns and generates evidence that can support more targeted strategies toward VL elimination in Ethiopia.

Methodology

Protocol registration

The Systematic Review and Meta-Analysis (SRMA) protocol was registered in the PROSPERO database (CRD420251107488) and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 2020 checklist [32] (Fig. 1).

Fig. 1.

PRISMA flow diagram for new systematic reviews, which included searches of databases and registers

Study area and time

This study was conducted in Ethiopia. It included studies from all relevant endemic regions of the country, covering historical and recent data from the 1940s through November 2024. Data collected included prevalence, risk factors, ecological and spatial determinants, and treatment access and outcomes. Ethiopia’s ecological features, as previously described [33], are mainly influenced by altitude, which significantly affects livelihoods, vector distribution, and health risks. The country is divided into six major agroecological zones: Unsuitable Wurch (above 3,700 m above sea level [masl]), Wurch (3,200 to 3,700 masl), Dega (2,400 to 3,200 masl), Weyna Dega (1,500 to 2,300 masl), Upper Kola (500 to 1,500 masl), and Lower Kola/Bereha (below 500 masl). These zones determine the environmental suitability for sandfly vectors and the spatial variations of visceral leishmaniasis transmission across Ethiopia.

Literature search strategy

Relevant primary studies were identified from electronic databases (PubMed, SCOPUS, CINAHL, EMBASE, HINARI, Cochrane Library, Web of Science, Google, and Google Scholar) following the PRISMA reporting guideline-recommended PICO (Patient/Population, Intervention/Exposure, Comparison, Outcome) framework. Gray literature, including reports from the Ethiopian Ministry of Health, the Ethiopian Public Health Institute, non-governmental organizations, and WHO monthly and programmatic reports, was formally requested. Manual searching of reference lists in selected studies was conducted. Identified publications and grey literature (non–peer-reviewed sources such as reports from the Ministry of Health) were compiled and imported into EndNote reference management software, then exported to Rayyan for duplicate removal and additional processing. Each of the authors examined the literature individually for suitability to the inclusion criteria by looking at the titles, abstracts, and full texts.

Population and scope

The review focused primarily on studies that reported patients, both adults and children, who acquired VL in Ethiopia or after travel to Ethiopia, while also capturing the broader ecological context by including all publications that examined sandflies and potential reservoir hosts. This dual emphasis ensured that human case data were considered alongside entomological and animal-host evidence, allowing the synthesis to reflect both clinical burden and the environmental and zoonotic dimensions of VL transmission in Ethiopia.

Intervention or exposure

Any primary study that investigated environmental, socioeconomic, or host-related factors that might increase exposure to VL infection was eligible. Studies describing housing conditions, land use, occupational or migratory behaviors, household socioeconomic status, or biological host characteristics were treated as sources of exposure information, with the aim of identifying the constellation of determinants that shape risk in affected communities.

Study design

A systematic review and meta-analysis (SRMA) was conducted to synthesize evidence from published and grey literature on visceral leishmaniasis in Ethiopia. No restrictions were placed on study design or setting; both community- and institution-based studies were considered. Eligible studies included randomized trials, cohort and case-control studies, cross-sectional surveys, outbreak reports, and descriptive case series that reported data on VL occurrence, risk factors, vectors, or reservoirs.

Main outcomes

The review synthesized multiple outcome domains to build a comprehensive picture of VL in Ethiopia. Measures of disease burden magnitude, prevalence, proportion, and incidence were collected to quantify human impact. Estimates of association, such as odds ratios, relative risks, attributable risks, and population attributable risks, were extracted to identify and rank risk factors. Spatial information was compiled at the district level to map where VL was reported, and reports of infection in sandflies and non‑human mammals were recorded and mapped to reveal overlaps or gaps between human cases and non‑human infections. Where possible, spatial overlays compared human case distributions with vector and reservoir infection sites. Temporal trends in case counts and the geographic distribution of diagnostic and treatment facilities were summarized and related to Ethiopia’s main agroecological zones, and identified VL risk areas.

Eligibility criteria, data extraction, and quality assessment

Included studies were those that clearly specified the study setting or area, described the diagnostic method used to confirm VL cases, and, when relevant, reported infections in sandflies or non‑human mammals; publications in English or any language spoken in Ethiopia were accepted, and both institutional and community settings qualified. Studies that failed to present the outcome variables of interest, such as prevalence, measures of association for exposure to VL, or spatial data were excluded. Data extraction prioritized clear documentation of setting, case confirmation methods, outcome measures, and spatial coordinates or administrative locations; quality assessment procedures focused on the clarity of case definition and diagnostic methods, the appropriateness of study design for the reported outcomes, and the completeness of spatial and exposure data to ensure that included studies could meaningfully inform the synthesis.

Study selection

Studies identified from published and unpublished sources were uploaded to Rayyan (https://new.rayyan.ai/) for screening and selection. Two independent reviewers (GMA, DE) performed study screening and selection, with an independent collaborator (AA) observing and mediating any disagreements. Reviewers first conducted title and abstract screening, followed by full-text review for final inclusion decisions. Results were compiled and presented using a PRISMA flow diagram.

Data extraction

The two reviewers extracted data independently using a pre-designed data extraction Excel form. The collaborator validated the consistency and completeness of extracted data with disagreements resolved through consensus discussion. Data related to VL cases load, treatment, treatment outcome, and district level showing treatment access (obtained from the WHO- Ethiopia Country Office), and species of sandflies and non-human mammals with Leishmania infection were developed and overlaid with agroecological zones and environmental factor-based VL risk maps [34]. Administrative boundary shapefiles for administrative boundaries were obtained from the Ethiopian Central Statistics Service (CSA), and ecological zone data were obtained from the CSA and the Ministry of Agriculture geospatial database. Furthermore, data related.

Quality assessment

Each selected article was independently assessed for methodological quality and risk of bias by two reviewers (GMA and DEW) using the appropriate Joanna Briggs Institute (JBI: https://jbi.global/critical-appraisal-tools) critical appraisal checklist according to the study design (analytical cross-sectional, cohort, and randomized controlled trials). In cases of disagreement, a consensus was reached through discussion among the reviewers, and when necessary, a third reviewer (EM) was consulted before proceeding with the analysis.

Data processing and statistical analysis

Systematic review and meta-analysis

Data from the included studies were analyzed using R statistical software (version 4.5.1). Pooled prevalence estimates and associations between visceral leishmaniasis (VL) and potential risk factors were calculated using standard meta-analytic techniques. Statistical heterogeneity was assessed using Cochran’s Q test and the I² statistic [35]. Heterogeneity was considered low if p > 0.10 and I² < 50%, in which case a fixed-effects model based on inverse variance weighting was applied. When heterogeneity was moderate to high (p ≤ 0.10 and I² ≥ 50%), pooled estimates were obtained using the DerSimonian–Laird random-effects model to account for between-study variability. Although the 50% I² threshold is a conventional guideline, heterogeneity estimates in this analysis were substantially above this threshold, confirming the robustness of the model choice. Heterogeneity was considered low if p > 0.10 and I² < 50%, in which case a fixed-effects model based on inverse variance weighting was applied. When heterogeneity was moderate to high (p ≤ 0.10 and I² ≥ 50%), pooled estimates were obtained using the DerSimonian–Laird random-effects model to account for between-study variability. Although the 50% I² threshold is a conventional guideline, heterogeneity estimates in this analysis were substantially above this threshold, confirming the robustness of the model choice.

Subgroup analyses were conducted to explore potential sources of heterogeneity, including study region, population type, and study design. Publication bias was assessed using visual inspection of funnel plots and Egger’s regression test, with asymmetry considered indicative of potential small-study effects. Sensitivity analyses were performed by sequentially excluding each study to evaluate its influence on the overall pooled estimates.

Data on VL endemic districts, sand fly vectors, and nonhuman mammalian reservoirs were extracted from the included studies. Spatial map generation was performed using ArcGIS (version 10.7.1). All features were overlaid on agroecological zones and environmental factor–based VL risk maps to assess spatial patterns of VL risk.

Descriptive analysis of national WHO data

National VL case data, including the number of treated cases, were obtained from the WHO-Ethiopia country office covering eight years (2017–2024) and were analyzed descriptively to assess temporal patterns in disease caseload, treatment practices, and cured treatment outcomes. Spatial maps of treatment centers were generated using R software version 4.5.1 and ArcGIS version 10.7.

Results

Systematic review & meta-analysis results

Literature search and study selection

Database searches yielded 4,293 records. After duplicate removal and screening, 32 studies were included for pooled prevalence estimation and VL foci mapping [36–67]. Additionally, 21 articles were used to identify risk factors for VL in Ethiopia [37, 40, 43–46, 50, 53, 55, 58, 59, 61–63, 66, 68–73] (Fig. 1)and 15 articles contributed data on sandfly species and non-human mammals found infected with Leishmania donovani [5–9, 73–82] (Supplementary Table 1).

Pooled VL prevalence by detection method and study setting

In community-based studies, the pooled prevalence of visceral leishmaniasis (VL) was 10.38% (95% CI: 5.38–16.74), with extreme heterogeneity across studies (I² = 99.4%, p < 0.001). Subgroup analysis by diagnostic method showed significant variation in prevalence estimates (p < 0.0001). Studies using the leishmanin skin test (LST) reported a pooled prevalence of 17.26% (95% CI: 6.33–32.08), while those using the direct agglutination test (DAT) reported 15.70% (95% CI: 0.04–49.87). Studies employing the rK39 immunochromatographic test (rK39-ICT) reported a comparatively lower prevalence of 8.30% (95% CI: 5.37–11.78) (Fig. 2). The wide confidence interval observed for DAT indicates considerable uncertainty, likely reflecting the limited number of studies and substantial variability between them.

Fig. 2.

Forest plot for the pooled community-level prevalence of visceral leishmaniasis in Ethiopia, stratified by the method of exposure detection

In institution-based studies, the pooled prevalence was 15.29% (95% CI: 9.25–22.51) with substantial heterogeneity (I² = 99.6%, p < 0.001). Studies employing rK39-ICT reported a higher prevalence (16.42%, 95% CI: 9.11–25.34) compared to those using the DAT method (12.77%, 95% CI: 1.36–32.86). Other infection detection methods had insufficient studies for pooled analysis (Fig. 3).

Fig. 3.

Forest plot for the pooled institution-level prevalence of visceral leishmaniasis in Ethiopia, stratified by the method of exposure detection

Factors associated with visceral leishmaniasis

Male sex was significantly and consistently associated with increased VL risk compared to females (OR = 1.86; 95% CI: 1.39–2.48; p < 0.001). Occupational analysis revealed that housewives (OR = 0.43; 95% CI: 0.23–0.83; p = 0.012) and merchants (OR = 0.32; 95% CI: 0.12–0.88; p = 0.026) had significantly lower odds of VL compared to farmers, reflecting differential exposure to sandfly vectors. Behavioral risk factors significantly associated with VL included sleeping on the ground (OR = 1.92; 95% CI: 1.17–3.15; p = 0.010) or outside the house (OR = 2.85; 95% CI: 1.84–4.42; p < 0.001) a history of travel to endemic areas significantly increased VL risk (OR = 3.99; 95% CI: 1.59–10.02; p = 0.003).

At the household level, a family history of VL substantially increased infection risk (OR = 2.27; 95% CI: 1.55–3.31; p < 0.001). The presence of domestic animals, specifically dogs (OR = 2.11; 95% CI: 1.43–3.12; p = 0.0002) and cattle (OR = 1.96; 95% CI: 1.21–3.16; p = 0.0058), significantly increased VL odds. Rural residence substantially elevated VL risk compared to urban areas (OR = 2.27; 95% CI: 1.45–3.53; p = 0.0003).

Environmental factors also played a substantial role. The presence of hyraxes (OR = 1.90; 95% CI: 1.02–3.56; p = 0.045), termite mounds (OR = 3.60; 95% CI: 1.94–6.68; p < 0.001) and acacia trees (OR = 3.75; 95% CI: 1.91–7.37; p = 0.0001) within 100 m of dwellings significantly increased VL likelihood (Table 1).

Table 1.

Pooled Odds Ratios of Factors Associated with Visceral Leishmaniasis in Ethiopia

Variable	Pooled COR	CI	P Value	Number of studies	Heterogeneity (I^2)
Sex
Female	1			15	75.2% [58.9%; 85.0%]
Male	1.83	[1.35; 2.48]	< 0.0001
Occupation
Farmers	1
Government Employee	0.3304	[0.0898; 1.2154]	0.0956	3	0.0% [0.0%; 89.6%]
Student	0.7901	[0.5039; 1.2390]	0.3047	5	53.5% [0.0%; 82.9%]
Housewives	0.4330	[0.2258; 0.8303]	0.0117	2	0.0%
Daily Laborer	1.1248	[0.6387; 1.9808]	0.6837	2	0.0%
Merchant	0.3231	[0.1193; 0.8754]	0.0263	1	NA
Age group (in years)
>= 15	1			7	67.8% [28.5%; 85.5%]
<15	1.3295	[0.9707; 1.8209]	0.0760
Educational Status
Unable To Read or Write	1
Can Read or Write	0.6785	[0.3806; 1.2098]	0.1887	1	NA
Primary	0.5165	[0.2148; 1.2421]	0.1400	2	0.0%
Secondary And above	0.5482	[0.1588; 1.8932]	0.3418	1	NA
Sleeping Place within the house
On Bed	1			3	63.9% [0.0%; 89.7%]
On Ground	1.9174	[1.1675; 3.1489]	0.0101
Sleeping Outside of The House
Never	1			4	65.9% [0.0%; 88.4%]
Yes	2.8533	[1.8422; 4.4193]	< 0.0001
Travel to VL- endemic areas
No	1			3	93.5% [84.5%; 97.3%]
Yes	5.1896	[1. 2847; 20.9637]	0.0208
Temporary labor Migrant
Non-migrant	1			2	0.0%
Migrant	0.9142	[0.5164; 1.6185]	0.7582
Indoor Spraying last six months
Yes	1			2	0.0%
No	0.7539	[0.4948; 1.1486]	0.1885
Bed Net Use
Yes	1			6	40.3% [0.0%; 76.3%]
No	0.9551	[0.6203; 1.4707]	0.8348
Family History Of VL
No	1			5	0.0% [0.0%; 79.2%]
Yes	2.2683	[1.5523; 3.3146]	< 0.0001
Dwelling place
Urban	1			5	41.2% [0.0%; 78.3%]
Rural	2.2652	[1.4535; 3.5300]	0.0003
House Wall made of
Cement	1			3	42.7% [0.0%; 82.7%]
Mud Wall	0.7621	[0.5050; 1.1500]	0.1956
Dog presence in the household
No Dog	1			7	59.4% [6.5%; 82.3%]
Dog	2.1124	[1.4323; 3.1154]	0.0002
Termite Hill within 100 m of household
No	1			3	76.0% [21.0%; 92.7%]
Yes	4.1228	[1.9128; 8.8863]	< 0.0003
Acacia Tree is within 100 m the House
No	1			4	73.2% [24.6%; 90.5%]
Yes	3.7471	[1.9052; 7.3694]	0.0001

Ecology and geographic distribution of VL

The endemic district spatial overlaid with agroecology showed that most of the 58 VL endemic districts concentrated in the Upper and Lower Kola (Bereha) agroecological zones. Endemic foci were concentrated in the northwestern, western, southwestern, and southeastern lowland regions, with potential expansion into adjacent Weyna Dega midland zones (Fig. 4).

Fig. 4.

The agroecology of visceral leishmaniasis in Ethiopia

An overlay of endemic districts with the environmental factor-based VL risk map [34] demonstrated that VL foci are concentrated within areas classified as high and very high VL risk (Fig. 5). Ten districts within these high and very high-risk areas, marked with red cross-hatching, experienced recent outbreaks, indicating persistent transmission with intermittent recrudescence (Fig. 5).

Fig. 5.

The distribution of visceral leishmaniasis endemic districts over the Ethiopia environmental factor-based VL risk map

Spatial distribution of sandflies and non-human mammals

The spatial distribution of sandfly species overlaid with agroecology (Fig. 6) showed congruence with VL endemicity patterns. Evidence of L. donovani infection was documented in Phlebotomus orientalis, Ph. martini, Ph. celiae, Ph. duboscqi, Ph. rodhaini, and Sergentomyia multidens. These species were predominantly detected in the hot and temperate climatic zones within the Upper and Lower Kola and portions of the Weyna Dega agroecological zones. Additionally, serological evidence of human exposure to Ph. orientalis bites, indicated by anti-saliva IgG positivity, was documented primarily in lowland and mid-altitude areas (Supplementary Table 1).

Fig. 6.

Map overlying the spatial distribution of sand fly species with Agroecological zones. Where: Pmr: Ph. Martini, Pce: Ph. Celiae, Pdu: Ph. duboscqi, Po: Ph. Orientalis, Po Saliva IgG: Ph. Orientalis saliva IgG, Pr: P. rodhaini

Similarly, reports on non-human mammalian infection with Leishmania donovani and L. infantum originated from lowland (Upper Kola) and mid-altitude (Weyna Dega) areas, with geographic overlap corresponding to vector distribution hotspots (Fig. 7).

Fig. 7.

Map overlying the spatial distribution of nonhuman mammalian species found infected by L. donovani and L. infantum with ecological zones

WHO descriptive analysis results

Treatment access and geographic inequalities

The geographical distributions of health facilities showed that there were geographical disparities in access to VL diagnosis and treatment. A total of 28 treatment centers were reported for the 58 districts with visceral leishmaniasis (VL) cases. The WHO Ethiopia Country Report (2025) identified substantial variation in treatment center availability across endemic areas (Fig. 8).

Fig. 8.

Visceral leishmaniasis endemic districts in Ethiopia (Red pin) and closest diagnostic and treatment facility (Yellow cross sign), as determined by a systematic review and meta-analysis, and the WHO Ethiopian Country Report, 2025

Temporal trends in confirmed VL cases

The reported number of treated VL cases showed spatial and temporal variation across regions. Amhara Regional State contributed the largest number of reported cases to the national VL case burden, with major peaks in 2018–2020 and 2023–2024. Tigray Regional State exhibited intermittent peaks, while Somali and South Ethiopia Regional States showed increasing case numbers after 2022. Afar and Oromia Regional States reported consistently low case numbers throughout the study period (Fig. 9).

Fig. 9.

Monthly trend of treated visceral leishmaniasis cases by regional states from 2017 to 2024, the report of the WHO-Ethiopia country office

Treatment outcomes by regimen

Three VL treatment regimens were reported: (1) combination therapy using pentavalent sodium stibogluconate and Paromomycin (SSGV/PM) or Ambisome and Miltefosine (AmB/MF), and (2) monotherapy with Ambisome (AmB). Overall, treatment outcomes were favorable, with low rates of treatment failure across all regimens. SSGV/PM and Ambisome monotherapy were the most widely used, with SSGV/PM demonstrating consistently high cure rates. Ambisome monotherapy achieved high cure rates but was associated with higher mortality, likely reflecting preferential use in severe cases. AmB/MF was prescribed least frequently, with outcomes similar to other regimens. (Table 2).

Table 2.

Treatment outcomes by regimen and regional state, 2017–2024. used and treatment outcomes by treatment sites in a regional state from 2017 to 2024, reported from the WHO-Ethiopia country office

Treatment regimen by site	Cured	Death	Defaulter	Referral	Failure
Afar regional state
SSGV/PM	26	1	4	0	1
SSGV	4	0	0	0	0
AmB	39	6	0	3	0
AmB/MF	0	0	0	0	0
Amhara regional state
SSGV	157	2	1	12	1
AmB	875	60	5	72	16
AmB/MF	372	25	3	23	6
SSGV/PM	2689	41	14	95	1
Oromia regional state
SSGV	85	1	0	1	1
AmB	120	3	1	2	0
AmB/MF	0	0	0	0	0
SSGV/PM	324	4	0	12	0
South Ethiopia regional state
SSGV	89	4	1	4	1
AmB	97	7	2	17	0
AmB/MF	2	0	0	0	0
SSGV/PM	271	11	8	1	0
Somali regional state
SSGV	71	1	0	0	1
AmB	107	6	0	0	0
AmB/MF	0	0	0	0	0
SSGV/PM	338	2	0	0	0
Tigray regional state
SSGV	58	1	2	1	0
AmB	353	13	5	14	0
AmB/MF	0	0	0	0	0
SSGV/PM	1055	16	19	19	2

Where: SSG^V = Pentavalent Sodium Stibogluconate monotherapy, AmB=Liposomal Amphotericin B monotherapy, AmB/MF = Liposomal Amphotericin B and Miltefosine combination, SSG^V/PM = Sodium Stibogluconate and Paromomycin combination

Discussion

Ethiopia represents a high-burden VL country in eastern Africa, with endemic foci accounting for approximately 73% of the global VL case burden. Ethiopia is among the four countries contributing nearly 60% of the global VL burden [83]. As a signatory to WHO commitments, Ethiopia has adopted a strategic framework for VL elimination as a public health problem in East Africa by 2030. The framework identifies research gaps and prioritizes understanding transmission dynamics. This systematic review and meta-analysis present current evidence on pooled VL prevalence, temporal trends, contributing transmission factors, treatment access, treatment outcomes, ecology of VL foci, vectors, and associated non-human mammals, as well as their spatial distribution.

Prevalence findings

The review documented substantial variation in pooled VL prevalence across diagnostic methods and study settings. The interpretation of reported prevalence requires consideration of the diagnostic tools employed. The leishmanin skin test (LST) assesses the history of exposure and current asymptomatic infection, whereas the rK39-ICT and direct agglutination test (DAT) detect antibody responses in both asymptomatic individuals and those with active clinical VL. The higher prevalence reported with LST compared to DAT and rK39-ICT in community-based studies likely reflects the differential sensitivity of these tools for detecting asymptomatic infection or prior exposure [84]. Community-level prevalence estimates were comparable to those reported from West Armachiho, Mehoni, and Tigray [85] but higher than estimates from other Ethiopian settings and a previous national meta-analysis, underscoring substantial within-country heterogeneity [86, 87]. The observed heterogeneity in pooled prevalence based on detection methods emphasizes the importance of standardized, validated approaches to enable optimal estimation of VL exposure across endemic foci to support elimination efforts.

Risk factors

Multiple individual-level factors were significantly associated with VL infection. Male sex was associated with a higher risk of visceral leishmaniasis, a pattern observed across the included studies and consistent with reports from East Africa and other endemic regions. This pattern is commonly attributed to behavioral and occupational exposures, including outdoor nighttime activities, farming, herding, and seasonal labor migration [ 88–95]. Occupational patterns further supported this interpretation, with significantly lower odds observed among housewives and merchants compared to farmers, consistent with findings from neighboring Kenya [96].

Environmental and household factors significantly contributed to VL risk. The presence of termite mounds, acacia trees, hyraxes, and domestic animals, combined with behaviors such as sleeping outdoors or on the ground, was significantly associated with VL infection. These findings are consistent with evidence from Eastern Sudan, where high tree density and proximity to domestic animals were linked to increased VL risk [95]. Similar associations have been documented in previous meta-analyses, highlighting the contribution of domestic animals (particularly cattle), outdoor sleeping habits, and close human-animal-vector proximity to VL transmission [31, 90, 96].

Termite mounds and acacia trees create microhabitats with stable humidity and temperature conditions favoring sandfly survival [97] and density Acacia-Balanites woodlands are strongly associated with high densities of Phlebotomus orientalis [91, 98, 99]. Domestic animals, particularly dogs and cattle, may attract sandfly vectors or, in some epidemiological contexts, serve as potential reservoir hosts or amplifying hosts. Studies from South Asia demonstrate that cattle influence sandfly abundance and infection dynamics [96]. Although dogs and cattle are not confirmed reservoirs in Ethiopia, their presence likely increases vector-host contact frequencies.

The finding that a family history of VL more than doubled infection risk aligns with previous East African meta-analyses [31, 100] and studies from endemic regions in India. This pattern likely reflects shared household microenvironments, clustering of sandfly exposure risk, and localized transmission dynamics within endemic foci.

Travel history to endemic areas increased VL risk nearly fourfold, consistent with evidence from Kenya [97] and Ethiopia [2]. Migrant workers traveling to lowland agricultural zones such as Humera, Metema, and Abdurafi face particular vulnerability due to outdoor nighttime exposure, potentially limited acquired immunity, and prolonged residence in high-vector-density environments [2, 101, 102].

Ecological and spatial distribution patterns

This review emphasizes strong geospatial alignment between VL endemicity, sandfly vector distribution, and agroecological zones. The sandfly species Phlebotomus orientalis, Phlebotomus martini, Phlebotomus celiae, Phlebotomus duboscqi, and Phlebotomus alexandri are documented in the Lower and Upper Kolla zones, characterized by Acacia-Balanites woodlands, black cotton soils, and termite mound associations [11, 9]. Phlebotomus orientalis dominates northern foci, thriving at 500–1,500 masl with seasonal density peaks linked to land surface temperatures exceeding 25 °C [10, 23]. Termite-mound-associated species (Phlebotomus martini, Phlebotomus celiae) predominate in southern foci, while Phlebotomus duboscqi and Phlebotomus alexandri extend into Rift Valley ecological zones [20, 103]. Notably, vector presence in transitional highland areas such as Libo Kemkem (Weyna Dega zone, up to 1,800 masl [81] and linked the lowland extension to 1,100 masl [10] suggests ecological shifts possibly driven by climate change or agricultural land-use expansion [12, 13].

Zoonotic considerations

While VL in Ethiopia is predominantly anthroponotic, the possibility of zoonotic contributions cannot be entirely excluded. Leishmania DNA has been detected in wild rodent species (Arvicanthis spp., Mastomys spp.) and bats [5, 7, 104], with serological evidence of exposure documented in domestic dogs, bovines, and donkeys [8, 74]. Although these animals are likely incidental rather than primary reservoir hosts, in contrast to canine reservoirs documented in South Asian endemic regions [105], their geographic overlap with sandfly vector hotspots warrants integrated One Health surveillance [77, 106]. Furthermore, agricultural and community livelihoods dominated by pastoralism and semi-agrarian practices in areas where sandflies and infected non-human mammals coexist create human-animal-vector interface zones. This convergence exacerbates socioeconomic burdens through reduced productivity, underreporting of cases, and recurrent outbreaks across endemic regions [107–109].

Pattern and trends of treatment outcomes

Analysis of the national VL treatment data from 2017 to 2024 revealed that Amhara regional state consistently reported the highest number of treated cases, while Somali and South Ethiopia regional states demonstrated increasing trends after 2022. Treatment outcomes were generally favorable across all regimens, with the sodium stibogluconate/paromomycin combination being the most frequently employed and demonstrating consistent efficacy. Higher mortality rates associated with liposomal amphotericin B likely reflect its preferential use in severe or complicated cases rather than reduced inherent efficacy.

Conclusion

This systematic review and meta-analysis demonstrate that VL in Ethiopia remains a persistent public health challenge characterized by intermittent outbreaks and substantial spatial heterogeneity in prevalence, risk factors, and treatment accessibility. VL transmission is shaped by intersecting behavioral, environmental, and socioeconomic determinants, while persistent inequities in diagnosis and treatment access remain evident. The documented treatment success with the short-course sodium stibogluconate/paromomycin regimen is encouraging; expanding access to this effective treatment will substantially support national VL elimination objectives. Strengthening harmonized surveillance tools, implementing targeted surveillance in emerging and high-burden areas, ensuring equitable access to effective treatment, and implementing integrated prevention approaches addressing both behavioral and environmental risk factors will be essential for supporting national VL elimination efforts and achieving the WHO 2030 elimination framework targets.

Limitations of this study include high heterogeneity across studies, variation in diagnostic methods, pooling of ORs from mixed study designs, lack of population denominators to estimate incidence, and limited data for sensitivity analyses. Although studies published in English and languages spoken in Ethiopia were included, the review may still have excluded relevant studies in other languages due to feasibility and resource constraints for translation and data extraction. These factors should be considered when interpreting the findings.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

VL

Visceral Leishmaniasis

WHO

World Health Organization

SSG

sodium stibogluconate

PM

Paromomycin

DAT

Direct Agglutination Test

LST

Leishmanin Skin Test

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

OR

Odds Ratios

CI

Confidence Intervals

JBI

Joanna Briggs Institute

CSA

Central Statistics Service

Author contributions

GMA, DBD, DEW, EM, EH, AA, GS, and TN conducted article searches and data extraction for all components of the results. GMA led the protocol development, wrote the methods section of the manuscript, produced mapping and prevalence estimating components of meta-analysis, trend analysis, and their interpretation, while DEW led the factors associated result production of the meta-analysis component of the study. EM, TA, EH, and AA drafted the introduction and discussion section of the manuscript; HBK and MW managed and curated the WHO VL data. EG supervised the overall work and provided critical feedback on each section. BAE and MB provided methodological support and reviewed the manuscript. FA, TC ES, IG, FA, SC, SD, RG, KEB, and EG reviewed the final draft. All authors read and approved the final manuscript.

Funding

This research was funded in whole by the Science for Africa Foundation to the Developing Excellence in Leadership, Training and Science in Africa (DELTAS Africa) programme [Grant number Del-22-005] with support from Wellcome Trust and the UK Foreign, Commonwealth & Development Office, and is part of the EDCTP2 programme supported by the European Union.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This systematic review and meta-analysis were registered on PROSPERO under the registration number CRD42024571850. The study protocol was reviewed and approved by the ALERT/AHRI Ethics Review Committee. Data from the WHO Ethiopia office were accessed with full support and formal permission from the office.

Consent for publication

Consent to publish this manuscript from the participants was deemed not applicable since the manuscript does not contain identifying data from any individual person.

Competing interests

The authors declare no competing interests.