The magnitude of flaviviruses infection in Ethiopia: a systematic review and meta-analysis

Abstract

Background

Flaviviruses are a well-known mosquito borne viruses and their nucleic acid is single-stranded RNA which is enclosed with the capsid and envelop. Flavivirus infections can cause a wide variety of clinical manifestations and complications in humans. Outbreaks have been recorded due to DENV and YFV infections in Ethiopia in different time. The aim of the present systematic review and meta-analysis (SRMA) is to determine the overall magnitude of flaviviruses infection in Ethiopia.

Methods

Initially reported under the registration number CRD42023492735, this SRMA was subsequently carried out in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) standards. The search for appropriate resources took place from December 10, 2023, to January 10, 2024, using a number of electronic databases, including PubMed, Google Scholar, Wiley Online Library, African Journal Online, Science Direct, ResearchGate, and Web of Sciences. Additionally, the quality of the included studies was evaluated using the Joanna Briggs Institute (JBI) assessment tool. The statistical analysis was conducted using Stata version 14.0. In addition to funnel plot and Egger’s test, which were used to detect publication bias, Cochran’s Q test and I2 statistics were employed to measure heterogeneity. The pooled prevalence was estimated via a random effect model.

Results

The overall pooled prevalence of flaviviruses infection in Ethiopia was found to be 12.67% (95% CI: 11.00–14.35). The highest pooled prevalence of 23.49% (17.53–29.45) was observed in DENV followed by ZIKV, 13.63% (12.73-40.00). The sero-prevalence markers of IgG and IgM for flaviviruses were detected among 12.18% and 7.41% patients, respectively. The highest pooled prevalence of flavivirus was demonstrated in studies conducted in health care facilities (17.32%). The sensitivity test exhibits that no single study had an impact on the overall pooled effect magnitude. The symmetry of the funnel plot and the Egger’s test statistics indicated publication bias.

Conclusions

The pooled prevalence of flaviviruses infection in Ethiopia was high and DENV was the commonest flavivirus observed in the country. Nationwide surveillance of flaviviruses infection among human population should be conducted. Prompt prevention and control strategies should be implemented particularly in outbreak prone areas.

Supplementary information

The online version contains supplementary material available at 10.1186/s12879-025-11443-y.

Introduction

Flaviviruses, members of the family Flaviviridae and genus Flavivirus, are well-known mosquito-borne arboviruses. Their genetic material consists of single-stranded RNA enclosed within a capsid and surrounded by an envelope [1]. Along with a number of other related viruses, the genus Flavivirus contains some of the most dangerous viruses that infect humans, including Dengue virus (DENV), Yellow Fever virus (YFV), Zika virus (ZIKV), and West Nile virus (WNV) [2, 3].

The same blood-feeding arthropod vectors, mostly mosquitoes, spread these viruses, and vertebrates, especially humans and non-human primates, are their primary hosts [2–4]. A study in the Gambella region of Ethiopia looked at the seroprevalence of West Nile virus (WNV) in cattle and found that livestock can be reservoir hosts for flaviviruses, which could lead to human infection [5]. It means that viruses can spread from people to animals and back again, and they can also appear and reappear as new and old infectious diseases [6].

Although most infected people do not exhibit any symptoms, flavivirus infections can cause a wide range of clinical symptoms and complications in humans [7, 8]. The majority of arboviral infections share certain clinical characteristics [9, 10]. The quality of life in relation to health is also significantly impacted by some of these virus types [11]. The clinical signs and symptoms of infections caused by these viruses can be very different. They can range from a fever that isn’t specific to more serious conditions like vascular leak syndrome, encephalitis, conjunctivitis, arthralgia, hepatonephritis, and even death in some cases [12, 13]. Guillain-Barré Syndrome and microcephaly are the two most prevalent neurological side effects linked to Zika virus (ZIKV) infection [11]. Numerous flavivirus species have been found to be responsible for febrile illnesses in various demographic groups [14]. The clinical severity of flavivirus infections is significantly influenced by the presence or absence of chronic comorbidities, including diabetes, hypertension, cardiovascular disease, or immunosuppressive disorders. People with pre-existing medical conditions are frequently at greater risk for serious consequences, such as complications, prolonged illness, and an increased chance of hospitalisation or death. On the other hand, people who are otherwise healthy might have less severe symptoms or heal faster. When evaluating risk, developing treatment plans, and putting public health strategies for controlling flavivirus outbreaks into practice, this relationship emphasises how crucial it is to take underlying health status into account [15, 16].

The four main flaviviruses that mosquitoes carry are known around the world to be major causes of human disease that have big effects on public health. These viral pathogens are still a threat, and there is a chance that they could cause widespread outbreaks if multiple arboviruses come together. This is a new phenomenon called arboviral syndemics [17, 18]. In many parts of Africa, infections brought on by common flaviviruses pose a serious threat to public health. Particularly in regions with a high mosquito density and inadequate healthcare infrastructure, these viruses significantly increase the burden of disease. Frequent outbreaks strain local health systems, disrupt economic productivity, and increase morbidity and mortality, especially among vulnerable populations such as children, the elderly, and those with underlying health conditions. The growing incidence of flavivirus infections highlights the urgent need for strengthened surveillance, vector control, and public health preparedness across the continent [19, 20]. The viruses have spread throughout the continent and are the cause of the region’s recurring outbreaks [21]. A systematic review and meta-analysis (SRMA) that included studies from other regions showed a strong correlation between the socioeconomic status of populations and the prevalence of the most common arboviral diseases and their respective vectors, despite the fact that extensive research has not yet been conducted in Africa [22–24].

Ethiopia has seen notable outbreaks of dengue virus (DENV) and yellow fever virus (YFV) infections at different times [13, 25]. Numerous flavivirus species have been found in Ethiopia’s various regions, both in healthcare facilities and in the general public. This systematic review is the first attempt to date in Ethiopia to gather and examine research on any of the four prevalent human-source flavivirus infections. This SRMA’s main goal is to determine the overall prevalence of flavivirus infections in Ethiopia in order to provide important information about their distribution and effects on public health.

Methods

Reporting and protocol registration

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed in the conduct and reporting of this SRMA to guarantee openness, accuracy, and thorough reporting [26]. To ensure transparency and methodological rigour prior to data collection, the protocol for this review was pre-registered with the International Prospective Register of Systematic Reviews (PROSPERO) under registration number CRD42023492735.

Data sources and search strategies

To find studies pertinent to the prevalence of common flaviviruses among Ethiopian populations, a thorough and methodical review of several electronic databases was conducted. PubMed, Google Scholar, Wiley Online Library, ScienceDirect, ResearchGate, and Web of Science were among the databases that were searched. The goal of this extensive search was to find a variety of grey and peer-reviewed publications discussing the prevalence and occurrence of flavivirus infections in Ethiopia. The review specifically concentrated on research describing the prevalence and epidemiological trends of important flaviviruses in the nation, including the Dengue, Zika, West Nile, and Yellow Fever viruses.To find studies pertinent to the topic, a thorough and methodical review of several electronic databases was carried out. To make sure that all pertinent materials were included, further searches outside of electronic databases were carried out. Unpublished theses, dissertations, and locally presented studies were gathered through a systematic review of university repositories and the annual research conference proceedings. Additionally, by looking through the reference lists and appendices of the first few chosen articles, a snowballing technique was used to find more relevant literature. Important studies that might not have come up in the main database searches were found with the aid of this supplemental technique. Over the course of one month, from December 10, 2023, to January 10, 2024, the combined electronic and snowball search was conducted.

MeSH phrases were used for the search, and the review question’s combined key terms were used. The following combination keys were used to gain access to all possibly eligible research; “Prevalence”, “sero-prevalence”, “Sero-positivity”, “detection”, “epidemiology”, “burden”, “magnitude” “Flavivirus”, “Dengue virus”, “Zika virus”, “West Nile Virus”, Yellow fever virus”, “Ethiopia”,. The Boolean operators’ terms “OR” and “AND” was used as necessary.

Eligibility criteria

Inclusion criteria

A set of precise inclusion criteria was used for this SRMA, in order to guarantee the quality and applicability of the chosen studies. Studies that documented the frequency of Flavivirus infections and/or their genetic variations were considered eligible. Only research involving human subjects of all ages and carried out within the borders of Ethiopia was taken into account. Grey literature, including theses, dissertations, and conference proceedings, as well as original research articles published in peer-reviewed journals, were included. Additionally, to maintain consistency in interpretation, only studies published in the English language were selected for analysis.

Exclusion criteria

Editorials, commentaries, letters to the editor, author replies, case reports, narrative reviews, and conference abstracts that lacked adequate information or for which the authors did not reply to requests for the full text were all excluded from this SRMA in order to preserve methodological rigour and concentrate on pertinent, high-quality data. Furthermore, research that did not use human participants or that did not offer quantitative information on the frequency of Flavivirus infections was not taken into account.

Study selection

EndNote version 20 software was used to effectively manage and arrange the literature found during the search process. All publications obtained from university repositories, conference proceedings, electronic database searches, and the reference lists of pertinent studies found through snowballing were imported and consolidated more easily thanks to this tool. Duplicate records were methodically found and eliminated following the initial import to guarantee that every study was only included once during the review procedure. After de-duplication, a thorough screening process was conducted. Using the predetermined inclusion and exclusion criteria, two independent reviewers evaluated each record in three stages: title, abstract, and full-text. The purpose of this multi-step review procedure was to reduce selection bias and guarantee objectivity. A third reviewer was consulted to offer an unbiased opinion in situations where the two reviewers disagreed about a study’s eligibility. For the final systematic review and meta-analysis, this method improved the transparency and dependability of the article selection process while guaranteeing consensus.

Quality assessment

Independent reviewers have evaluated the quality of each included study to guarantee the consistency and dependability of the results. The purpose of this critical evaluation was to assess the overall validity and methodological rigour of the chosen research. The Joanna Briggs Institute (JBI) critical appraisal checklist—which has been modified for cross-sectional study designs—was used for this purpose. Eight structured criteria make up the JBI tool, which evaluates important elements like the suitability of the sampling strategy, the precision with which study subjects and settings are defined, the validity and reliability of the measurement instruments, the detection and control of confounding variables, and the suitability of statistical analyses. Every study was assessed separately in light of these standards, and any disputes amongst reviewers were settled by discussion or, if required, the participation of a third reviewer. The credibility of the systematic review and meta-analysis was increased by this methodical approach, which made sure that only research of acceptable methodological quality were included in the final analysis [27]. The final analysis only included studies that satisfied the Joanna Briggs Institute (JBI) rating criteria, which set a minimum quality level of 50% or higher. This threshold was put in place to make sure that only research with respectable methodological quality added to the systematic review and meta-analysis’s conclusions. Studies that met or beyond this criterion were deemed to be of good quality since they demonstrated enough design, execution, and reporting rigour to back up trustworthy findings. The overall validity and robustness of the synthesised results were improved by this quality-based inclusion criterion.

Data extraction process

A standardised extraction sheet created in Microsoft Office Excel was used to extract the data. The purpose of this page was to compile and arrange pertinent data from every study that qualified. The most pertinent findings regarding the incidence of flaviviruses were jointly identified and documented by three independent reviewers. By suggesting certain subheadings based on their examination of the literature, each reviewer helped to shape the Excel sheet’s structure. The reviewers cross-checked each other’s extracted data to guarantee accuracy and consistency. Any disagreements were settled by dialogue and, if required, by rerunning some steps of the extraction procedure in order to come to an agreement. The primary data gathered from every study included the following: the first author’s name, the year of publication, the study location, the study setting, the study design, the study population’s characteristics, the sample size, the diagnostic techniques employed, and the quantity, kind, and frequency of flaviviruses found. A thorough and trustworthy dataset for further research was guaranteed by this painstaking procedure.

Statistical methods and analysis

Once the data extraction was completed, the extracted data were exported to Stata version 14.0 for statistical analysis. Heterogeneity among the included studies was assessed using Cochran’s Q test and the I² statistic. Heterogeneity was considered present if the I² value exceeded 50% [28]. A Q test p-value of less than 0.05 also suggested heterogeneity. A 95% confidence interval and the pooled prevalence of flaviviruses were then estimated using a random-effects model [29]. A forest plot, which successfully depicted the estimates from each research as well as the total pooled effect, was used to graphically summarise the results. Both the Egger test and the funnel plot techniques were used to assess the possible existence of publication bias. Notable asymmetry in the funnel plot and a statistically significant Egger test p-value of less than 0.05 revealed publication bias, suggesting that smaller studies with non-significant results might have gone unreported or were not published. Subgroup analyses were carried out based on variables like geographic location, year of publication, study setting (e.g., community-based or hospital-based), type of immunoglobulin measured, and particular flavivirus species examined in order to investigate possible sources of variation and heterogeneity across studies. To make sure the results were reliable and unaffected by any one study, sensitivity analyses were also carried out by methodically eliminating individual studies in order to evaluate their impact on the overall pooled prevalence estimate.

Results

Description of included studies

A total of 2521 articles were found through database searches and other sources. Out of those articles, 1532 articles were eliminated due to redundancy. After evaluating the titles and abstracts of 989 publications, 879 investigations were excluded. Following that, the eligibility criteria were applied to 110 full-text papers for review. Approximately 82 full-text articles were then eliminated. Ultimately, only twenty-eight publications were deemed possibly suitable and included in this review for the final analysis (Fig. 1).

Fig. 1.

Diagram showing the flow of the research that were included in the meta-analysis and systematic review of Ethiopia’s flavivirus prevalence

Characteristics of the included studies

A total of 28 original articles conducted in different regions of Ethiopia were included in this SRMA. The majority of the included studies were reported from Southern Nations, Nationalities and Peoples’ Region (SNNPR) of Ethiopia, (9; 32.14%), followed by the Oromia Region, (6; 21.43%). Amhara and Somali regions were represented by a single study. One study was conducted in multiple regions and localities of the country and it addresses the four types of flavivirus. In terms of the study design, all studies were cross-sectional studies and one study uses cross-sectional but retrospective approach. Most studies used immunoglobulin (IgG and IgM), four studies used nucleic acid and one study used antigen (NS1 Ag) as a diagnostic markers.

The included studies reported flavivirus that ranged from 0.43% for zika virus to 67.74% for dengue virus. The lowest prevalence of flavivirus was reported from a study which was conducted in multiple sites of the country in 2018, while the highest prevalence was exhibited from a study which was conducted in Dire Dawa city administration in 2019. About 67.86% (19/28) of the included articles use all age group population, 17.86% (5/28) studies encompass study participants who were greater than 5 years of age, 10.71% (3/28) research works involve adult population category whereas one study was conducted among children. In overall the incorporated studies used four types of diagnostic methods namely Immunofluorescence testing (IFT), Enzyme linked immune-sorbent assay (ELISA), Plaque Reduction Neutralization Test (PRNT) and polymerase chain reaction (PCR). Majority of the studies that were recruited in this study uses IFT (12/28), and then followed by ELISA (8/28), PCR (4/28) and PRNT (4/28). Most (71.43%) of the studies included in this systematic review and meta-analysis were conducted in health care facilities. The number of study subjects in each of the incorporated studies were between 57 (the lowest sample size) and 1643 (the highest sample size) (Tables 1 and 2).

Table 1.

Distribution, diagnostic method and patient characteristics of studies on flaviviruses in Ethiopia

Author	Study area	Region	Pub year	Study design	Diagnosis	Method Used	Participants
Nigussie et al. [30]	Borena	Oromia	2020	Cross–sectional	IgG	IFT	All patients
Nigussie et al. [30]	Borena	Oromia	2020	Cross–sectional	IgM	IFT	All patients
Tsegaye et al. [31]	Multiple	Multiple	2018	Cross–sectional	IgG	PRNT	> 5 years
Endale et al. [32]	South Omo Zone	SNNPR	2020	Cross–sectional	IgG	ELISA	> 5 years
Eshetu et al. [33]	Gamo Gofa zone	SNNPR	2020	Cross–sectional	IgG	IFT	All patients
Eshetu et al. [33]	Gamo Gofa zone	SNNPR	2020	Cross–sectional	IgM	IFT	All patients
Asebe et al. [34]	Lare and Itang	Gambella	2021	Cross–sectional	IgG	ELISA	Adult
Mulchandani et al. [35]	South Omo Zone	SNNPR	2019	Cross–sectional	IgM	ELISA	All patients
Yusuf and Ali [36]	Somali	Somali	2016	Cross–sectional	Nucleic acid	PCR	All patients
Geleta [37]	Borena	Oromia	2018	Cross–sectional	IgG	IFT	All patients
Geleta [37]	Borena	Oromia	2018	Cross–sectional	IgM	IFT	All patients
Eshetu et al. [38]	Arba Minch Zuria	SNNPR	2020	Cross–sectional	IgG	IFT	All patients
Eshetu et al. [38]	Arba Minch Zuria	SNNPR	2020	Cross–sectional	IgM	IFT	All patients
Ferede et al. [39]	Metema and Humera	Tigray	2018	Cross–sectional	IgG	ELISA	All patients
Ferede et al. [39]	Metema and Humera	Tigray	2018	Cross–sectional	IgM	ELISA	All patients
Tsegaye et al. [31]	Multiple	Multiple	2018	Cross–sectional	IgG	PRNT	> 5 years
Degife et al. [40]	Dire Dawa	Dire Dawa	2019	Cross–sectional	Nucleic acid	PCR	All patients
Shimelis et al. [41]	Hawassa	SNNPR	2023	Cross–sectional	NS1 Ag	ELISA	children
Woyessa et al. [42]	Dire Dawa	Dire Dawa	2014	Cross–sectional	Nucleic acid	PCR	All patients
Sisay et al. [43]	Dire Dawa	Dire Dawa	2022	Cross–sectional	Nucleic acid	PCR	All patients
Eshetu et al. [44]	Arba Minch Zuria	SNNPR	2020	Cross–sectional	IgG	IFT	All patients
Eshetu et al. [44]	Arba Minch Zuria	SNNPR	2020	Cross–sectional	IgM	IFT	All patients
Tsegaye et al. [31]	Multiple	Multiple	2018	Cross–sectional	IgG	PRNT	> 5 years
Nigussie [45]	Borena	Oromia	2021	Cross–sectional	IgG	IFT	All patients
Nigussie [45]	Borena	Oromia	2021	Cross–sectional	IgM	IFT	All patients
Asebe et al. [34]	Lare and Itang	Gambella	2021	Cross–sectional	IgG	ELISA	Adult
Tsegaye et al. [31]	Multiple	Multiple	2018	Cross–sectional	IgG	PRNT	> 5 years
Akelew et al. [46]	Gondar	Amhara	2022	Cross–sectional	IgG	ELISA	Adult

Table 2.

Characteristics of included articles describing flaviviruses in Ethiopia

Author	Study approach	Study setting	Sample size	Positive cases	Prevalence	Flavivirus type
Nigussie et al. [30]	Prospective	Health facility	519	65	12.52	Yellow Fever
Nigussie et al. [30]	Prospective	Health facility	519	38	7.32	Yellow Fever
Tsegaye et al. [31]	Prospective	Community	1643	10	0.61	Yellow Fever
Endale et al. [32]	Prospective	Community	313	155	49.52	Yellow Fever
Eshetu et al. [33]	Prospective	Health facility	529	79	14.93	Yellow Fever
Eshetu et al. [33]	Prospective	Health facility	529	38	7.18	Yellow Fever
Asebe et al. [34]	Prospective	Community	135	4	2.96	Yellow Fever
Mulchandani et al. [35]	Prospective	Community	165	6	3.64	Yellow Fever
Yusuf and Ali [36]	Retrospective	Health facility	57	33	57.89	Dengue
Geleta [37]	Prospective	Health facility	519	119	22.93	Dengue
Geleta [37]	Prospective	Health facility	519	41	7.90	Dengue
Eshetu et al. [38]	Prospective	Health facility	529	133	25.14	Dengue
Eshetu et al. [38]	Prospective	Health facility	529	42	7.94	Dengue
Ferede et al. [39]	Prospective	Health facility	600	126	21	Dengue
Ferede et al. [39]	Prospective	Health facility	600	114	19	Dengue
Tsegaye et al. [31]	Prospective	Community	1643	9	0.55	Dengue
Degife et al. [40]	Prospective	Health facility	62	42	67.74	Dengue
Shimelis et al. [41]	Prospective	Health facility	407	9	2.21	Dengue
Woyessa et al. [42]	Prospective	Health facility	88	50	56.82	Dengue
Sisay et al. [43]	Prospective	Health facility	60	13	21.67	Dengue
Eshetu et al. [44]	Prospective	Health facility	529	39	7.37	West Nile
Eshetu et al. [44]	Prospective	Health facility	529	24	4.54	West Nile
Tsegaye et al. [31]	Prospective	Community	1643	15	0.91	West Nile
Nigussie [45]	Prospective	Health facility	519	38	7.32	West Nile
Nigussie [45]	Prospective	Health facility	519	14	2.70	West Nile
Asebe et al. [34]	Prospective	Community	150	41	27.33	Zika
Tsegaye et al. [31]	Prospective	Community	1643	7	0.43	Zika
Akelew et al. [46]	Prospective	Health facility	200	41	20.5	Dengue

Pooled prevalence of flaviviruses in Ethiopia

Twenty-eight articles were included in this SRMA to estimate the pooled prevalence of flaviviruses in Ethiopia. The included studies revealed a greater disparity in the prevalence of flaviviruses. The prevalence ranges from 0.43% (95%CI: 0.11–0.74%) reported from multiple region of Ethiopia (zika virus) to 67.74% (95%CI: 56.11–79.38%) reported in Dire Dawa city administration (dengue virus). The overall pooled prevalence of flaviviruses in Ethiopia from the random effects model was 12.67% (95% CI: 11.00–14.35). There was a high level of heterogeneity between studies (I² = 98.4% and Q test = 12.67, p < 0.001) (Fig. 2).

Fig. 2.

Forest plot showing the pooled prevalence of flaviviruses in Ethiopia from random-effect model analysis

Subgroup analysis of flavivirus prevalence in Ethiopia

The subgroup analysis by flavivirus type demonstrated that the highest pooled prevalence of 23.49% (17.53–29.45) was observed in DENV followed by ZIKV, 13.63 (12.73-40.00), YFV, 11.95 (5.86–18.04), and WNV, 4.43 (1.79–7.07). About 18 studies were published in the year between 2019 and 2023 and their pooled prevalence of flaviviruses in these articles was 14.71% (95%CI: 11–18.40). By considering the study setting the highest pooled prevalence of flavivirus was demonstrated in studies that were conducted in health care facilities (17.32%). By different regions of Ethiopia indicated that the highest pooled prevalence of 48.67% (21.4-75.93) was observed in Dire Dawa city administration followed by 19.96% (17.70-22.22) in the Tigray region, 14.92% (8.96–38.8) in the Gambella region and 13.21% (7.65–18.78) in SNNPR region. On the other hand, the lowest prevalence of 0.58 (0.40–0.76) was observed in the studies which were conducted in multiple region of Ethiopia. In addition, the subgroup analysis based on population type revealed that 11.35% (5.15–17.59) prevalence was indicated among adult population. Similarly, the result from subgroup analysis based on diagnostic method demonstrated that studies conducted using PCR method showed the highest pooled prevalence of 50.91% (30.29–71.53) and then followed by prevalence of 18.03% (9.12–26.93) using ELISA and 10.46% (7.32–13.59) using IIFT (Table 3).

Table 3.

Subgroup analysis of flaviviruses by virus type, year of publication, Immunoglobulin type, study setting, region, population type and diagnostic method in Ethiopia

Subgroup		No of studies	Pooled prevalence (95% CI)	Heterogeneitytest (I2)	P-value
Virus types	Yellow fever Dengue West Nile Zika	8 13 5 2	11.95 [5.86–18.04] 23.49 [17.53–29.45] 4.43 [1.79–7.07] 13.63 [12.73-40.00]	98.6 98.8 94.5 98.2	< 0.000 < 0.000 < 0.000 < 0.000
	Total pooled	28	12.67 (11.00-14.35)	98.4	< 0.000
Publication year	2014–2018 2019–2023	10 18	10.22 [8.24–12.20] 14.71 [11-18.40]	98.6 97.3	< 0.000 < 0.000
	Total pooled	28	12.67 (11.00-14.35)	98.4	< 0.000
Immunoglobulin type (Ig)	IgG IgM	15 8	12.18 [10.25–14.10] 7.41 [4.54–10.28]	98.7 93	< 0.000 < 0.000
	Total pooled	23	10.54 (8.92–12.16)	98.4	< 0.000
Study setting	Health facility Community	20 8	17.32 [13.71–20.98] 5.18 [3.52–6.83]	97.3 98.4	< 0.000 < 0.000
	Total pooled	28	12.67 (11.00-14.35)	98.4	< 0.000
Region	Oromia	6	9.65 [5.31–14.58]	96%	< 0.000
	Multiple	4	0.58 (0.40–0.76)	0%	0.394
	SNNPR	9	13.21 (7.65–18.78)	98.0%	< 0.000
	Gambella	2	14.92 (8.96–38.8)	97.4%	< 0.000
	Tigray	2	19.96 (17.70-22.22)	0	0.386
	Dire Dawa	3	48.67 (21.4-75.93)	94.8	< 0.000
	Total pooled	26	11.72 (10.07–13.38)	94.8	< 0.000
Population Type	All	19	6.45 (3.88, 9.02)	84.9%	< 0.001
	> 5 year	5	1.9 (0.73, 3.08)	36.4%	0.194
	Adult	3	11.35 (5.15–17.59)	70.4%	0.034
	Total pooled	27	13.19 (11.47–14.92)	98.4%	< 0.000
Diagnostic method	IIFT	12	10.46 (7.32–13.59)	0.0%	0.635
	PRNT	4	0.58 (0.40–0.76)	0	0.394
	ELISA	8	18.03 (9.12–26.93)	98.5%	< 0.000
	PCR	4	50.91 (30.29–71.53)	92.6%	< 0.000
	Total pooled	28	12.67 (11.00-14.35)	98.4%	< 0.000

Publication bias

A funnel plot was used to visually assess the chosen studies for potential publication bias. The funnel plot’s asymmetry suggested the existence of publishing bias. Additionally, the results of Egger’s test indicated a publication bias that was only marginally significant(p-value = < 0.001) (Fig. 3) (Table 4).

Fig. 3.

Funnel plot of flaviviruses in Ethiopia

Table 4.

Egger’s test statistics of the prevalence of flavivirus in Ethiopia

Std_Eff	Coef.	Std. Err.	T	P>|t|	95% CI
Slope	−1.19	0.39	−3.03	0.005	−1.99, −0.38
Bias	8.85	0.84	10.59	0.000	7.14, 10.57

Trim and fill analysis of pooled prevalence of flavivirus in Ethiopia

A trim and fill analysis was carried out since publication bias was present. Following the inclusion of fourteen studies, Ethiopia’s pooled prevalence of flavivirus was 2.087%.(95% CI: 0.263 − 3.911) (Table 5).

Table 5.

Trim and fill analysis of the prevalence of flavivirus in Ethiopia

Method	Pooled est.	95% CI		Asymptotic		No. of studies
		Lower	Upper	z-value	p-value
Fixed	1.382	1.208	1.555	15.598	0.000	28
Random	12.675	11.004	14.346	14.865	0.000
Test for heterogeneity: Q = 1667.733 on 27 degrees of freedom (p = 0.000)
Moment-based estimate of between studies variance = 16.402
Trimming estimator: Linear
Meta-analysis type: Fixed-effects model
Iteration	Estimate	Tn	# To trim		Diff
1	1.382	394	14		406
2	0.907	397	14		6
3	0.907	397	14		0
Filled
Meta-analysis
Method	Pooled est.	95% CI		Asymptotic		No. of studies
		Lower	Upper	z-value	p-value
Fixed	0.907	0.736	1.078	10.375	0.000	42
Random	2.087	0.263	3.911	2.242	0.025
Test for heterogeneity: Q = 3185.035 on 41 degrees of freedom (p = 0.000)
Moment-based estimate of between studies variance = 30.196

Sensitivity analysis

A random effects model was used to conduct a sensitivity analysis in order to evaluate the effect of many investigations on the total estimate. After excluding individual studies, the pooled prevalence that was left was within the 95% confidence interval (CI) of the entire pooled estimate. This indicates that the total pooled effect magnitude was not influenced by any particular study (Table 6).

Table 6.

Sensitivity analysis of the magnitude of flaviviruses infection in Ethiopia

S. no.	Study omitted	Estimate	95% CI
1	Nigussie et al. [30]	12.62	10.93	14.30
2	Nigussie et al. [30]	12.89	11.19	14.59
3	Tsegaye et al. [31]	13.86	11.91	15.82
4	Endale et al. [32]	11.17	9.61	12.73
5	Eshetu et al. [33]	12.49	10.82	14.17
6	Eshetu et al. [33]	12.90	11.20	14.60
7	Asebe et al. [34]	13.08	11.37	14.79
8	Mulchandani et al. [35]	13.05	11.35	14.76
9	Yusuf and Ali [36]	12.03	10.38	13.68
10	Geleta [37]	12.11	10.46	13.75
11	Geleta [37]	12.86	11.16	14.56
12	Eshetu et al. [38]	11.99	10.36	13.62
13	Eshetu et al. [38]	12.86	11.16	14.55
14	Ferede et al. [39]	12.16	10.52	13.81
15	Ferede et al. [39]	12.27	10.62	13.92
16	Tsegaye et al. [31]	13.92	11.94	15.89
17	Degife et al. [40]	11.76	10.13	13.38
18	Shimelis et al. [41]	13.19	11.47	14.92
19	Woyessa et al. [42]	11.82	10.18	13.45
20	Sisay et al. [43]	12.51	10.83	14.19
21	Eshetu et al. [44]	12.89	11.19	14.59
22	Eshetu et al. [44]	13.05	11.34	14.76
23	Tsegaye et al. [31]	13.67	11.79	15.55
24	Nigussie [45]	12.89	11.19	14.59
25	Nigussie [45]	13.17	11.45	14.90
26	Asebe et al. [34]	12.24	10.58	13.92
27	Tsegaye et al. [31]	14.03	12.01	16.04
28	Akelew et al. [46]	12.38	10.70	14.06

Discussion

This systematic review shows the epidemiologic distribution of flavivirus and their pooled prevalence in the Ethiopian context. Infections due to the common flaviviruses cause an important public health problem in different parts and localities of Ethiopia [31, 47]. Even though there is a number variation, studies that aimed to determine the prevalence of one or more of the common flaviviruses were conducted in the northern, southern, western and eastern parts of Ethiopia. The laboratory confirmed cases of flavivirus infection was documented during an outbreak and non-outbreak period in various areas of the country.

The documented studies use various biomarkers and several diagnostic methods for the detection of flaviviruses in the health facility and community set up. In the present SRMA majority of the studies that aimed to determine the four flaviviruses used IFT (12/28), followed by ELISA (8/28) whereas a systematic review that focuses only on dengue, chikungunya and zika sero-prevalence at global level reported that more than half of the included studies used ELISA [48]. There were also variations in the study population in which the majority of studies uses all age group, some others uses adult population and populations who were greater than 5 years of age.

The pooled prevalence of the four flaviviruses in Ethiopia was 12.67% which is in agreement (14.5%) with a single study conducted in several localities in Kenya that have close border to Uganda, South Sudan and Ethiopia [20]. This can be evidence for the circulation of flavivirus in the region and for the possible transmission of the viruses across the countries.

Among the four flaviviruses the highest pooled prevalence was detected in DENV (23.49%) followed by ZIKV, YFV, and WNV. SRMA that was done in Africa to investigate the etiology of non-malaria febrile illness demonstrated that the highest rate of DENV infection as a cause of the illness [14]. Additionally the pooled prevalence of dengue virus in the present study was higher than the two previously conducted similar studies in Africa which reports 7% [49] and 14% overall prevalence [21]. Furthermore, the pooled prevalence of 27% DENV in the Sudan demonstrated comparable with the present SRMA [50]. In the contrary, the present pooled prevalence for DENV is lower than the report in the world (38%) in which the highest burden was indicated from South-East Asia Region [51].

The pooled prevalence of WNV in the present SRMA was lower than a similar study that was conducted in Nigeria which included studies conducted about seven decade ago [52] and it was also lower than SRMA conducted in the Eastern Mediterranean region [53]. Djibouti and Sudan are the two neighboring countries to Ethiopia that reports WNV infection to their population. In Djibouti apart from the human population the virus has been detected among animal and primary vector [53]. This might be an evidence for the transmission of the WNV from Djibouti and Sudan in which the virus is circulated among human, animal and the primary vector.

Even though there were only two studies included, the pooled prevalence of ZIKV in Ethiopia was 13.63% which is comparable to the global estimate of the virus in which the American region and the African region had shown the highest and the lowest rate, respectively [51]. In comparison with the pooled prevalence of the ZIKV in SRMA conducted among population in blood donation, Africa is represented by a single study from Cameroon [54], the present pooled prevalence for the virus becomes high. The high sero-prevalence of ZIKV in Gambella region might have an association with blood donation and the situation might hamper the blood donation practice of the region and the country. There was an evidence for high sero-prevalence of ZIKV in the Sudan that share border with the Gambella region of Ethiopia [55].

Even though Ethiopia is represented by only one study, a systematic review that was conducted in Africa reported 4% overall concurrent malaria and dengue fever infection. In East Africa, the study similarly revealed a 4.7% regional frequency [56]. This investigation also showed that the only species that simultaneously caused dengue fever and malaria was P. falciparum. More than 60% of Ethiopia’s plasmodium species, P. falciparum, cause malaria, and malaria affect more than half of the country’s land area. Given this, it is possible that the pooled prevalence of DENV is higher than predicted.

According to the subgroup analysis of the present study the pooled prevalence of YFV was found to be 11.95%. This figure is in consistence (9.4%) to a similar study held in Africa [57]. Since community and health facility based studies were evenly included in this report, the present systematic study indicates that YFV infection has become prominent public health problem in Ethiopia particularly in localities found in the Gambella region (Lare and Itang), SNNPR (Gamo Gofa and South Omo zone) and Oromia region (Borena zone). These areas have close proximity to each other and share borderline demarcation with arboviral infection endemic neighbor countries like Kenya and South Sudan. There was an evidence for the detection of the virus from the vector in West Africa (Senegal) and East Africa (Kenya) [57]. South Omo zone is one of YF viral infection endemic area in Ethiopia which is evidenced by the resurgence of YF outbreak after certain period of time in the area [13]. According to Tsegaye et al., YF and other flaviviruses have recently circulated in the population and have shown the potential to become epidemic in Ethiopia. This situation is evidenced by the infection of young children with YF and Zika virus [31].

The pooled prevalence of flavivirus in health care facilities was found to be 17.32% which is higher than the prevalence that was demonstrated from the community based studies (5.18%). In this SRMA the pooled prevalence of flavivirus among adult population was 11.35% which is higher than results from the overall prevalence among all population and population whose age were greater than five years. This might be due to high level of exposure of the adult population category to the mosquito vector that transmit these flaviviruses during occupational activities and these population group had the chance of moving from place to place for different purposes. Furthermore, the higher incidence of flavivirus infections observed in adults may be influenced by factors such as a larger proportion of samples collected from adults and the potential that certain studies were conducted in regions with increased transmission rates.

Conclusion

According to the present SRMA estimate the pooled prevalence of flaviviruses infection in Ethiopia was high, given that study participants from health care facilities demonstrate the highest sphere share of the infection. Among the four common flaviviruses infection the highest pooled prevalence was observed in DENV followed by ZIKV, YFV and WNV. The overall prevalence of flaviviruses according to IgG detection was found to be higher than the reports from IgM finding. The highest pooled prevalence of flavivirus was observed in Dire Dawa city administration followed by the Tigray, the Gambella, the SNNPR and the Oromia region. In the country nationwide surveillance of flaviviruses infection among human population should be conducted. Prompt prevention and control strategies should be implemented particularly in outbreak prone areas of the country.

Limitation of the study

Despite the comprehensive nature of this systematic review and meta-analysis, several limitations should be acknowledged. Most of the available studies were concentrated in specific regions of Ethiopia, leading to geographic and temporal gaps and an underrepresentation of data from other parts of the country. Additionally, substantial heterogeneity was observed across studies due to differences in study design, population demographics, diagnostic assays (e.g., ELISA, PRNT, RT-PCR), and case definitions, complicating direct comparisons and potentially affecting the reliability of pooled estimates. Furthermore, there is a possibility of publication bias, as studies reporting significant findings are more likely to be published; although funnel plots and statistical tests were employed to assess this, the limited number of studies reduced the power to detect such bias accurately.

Abbreviations

DENV

Dengue Virus

ZIKV

Zika Virus

WNV

West Nile Virus

YFV

Yellow Fever Virus

SRMA

Systematic Review and Meta-Analysis

SNNPR

Southern Nations, Nationalities and Peoples’ Region

IFT

Immunofluorescence testing

ELISA

Enzyme Linked Immune-Sorbent Assay

PRNT

Plaque Reduction Neutralization Test and PCR: polymerase chain reaction

Author contributions

Daniel GW conceptualize and design the study and search for articles, screen, and extract data, and evaluate the quality of the articles included. In addition, perform statistical analysis and write the manuscript. Saba GT, Hussen E, Melaku AB and Alemu G. were involved in data extraction. Daniel GW, Ousman M, and Habtu D participated in the quality assessment of the included studies. Zewdu M, Melaku AB, Ermias A and Daniel GW perform statistical analysis and interpretation of the data. Bruktawit E, Zewdu M, Mihret T, and Agumas S involve in checking the validity and monitor the overall process, in addition they were participated in critical review activity. All the authors read and approved the final manuscript.

Funding

There is no specific funding for this review.

Data availability

Data is provided within the manuscript or supplementary information files.

Declarations

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

Ethical approval and consent to participate are not required because this is a systematic review and meta-analysis that does not involve human or animal experiments.