Hepatitis B Virus in West African Children: Systematic Review and Meta-Analysis of HIV and Other Factors Associated with Hepatitis B Infection

While Hepatitis B virus (HBV) and the human immunodeficiency virus (HIV) are endemic in West Africa, the prevalence of HBV/HIV coinfection and their associated risk factors in children remains unclear. In this review, we sought to assess HBsAg seroprevalence among 0- to 16-year-olds with and without HIV in West African countries and the risk factors associated with HBV infection in this population. Research articles between 2000 and 2021 that reported the prevalence of HBV and associated risk factors in children in West Africa were retrieved from the literature using the Africa Journals Online (AJOL), PubMed, Google Scholar, and Web of Science databases as search tools. StatsDirect, a statistical software, was used to perform a meta-analysis of the retained studies. HBV prevalence and heterogeneity were then assessed with a 95% confidence interval (CI). Publication bias was evaluated using funnel plot asymmetry and Egger’s test. Twenty-seven articles conducted across seven West African countries were included in this review. HBV prevalence among persons aged 0 to 16 years was 5%, based on the random analysis, given the great heterogeneity of the studies. By country, the highest prevalence was observed in Benin (10%), followed by Nigeria (7%), and Ivory Coast (5%), with Togo (1%) having the lowest. HBV prevalence in an HIV-infected population of children was (9%). Vaccinated children had lower HBV prevalence (2%) than unvaccinated children (6%). HBV prevalence with a defined risk factor such as HIV co-infection, maternal HBsAg positivity, undergoing surgery, scarification, or being unvaccinated ranged from 3–9%. The study highlights the need to reinforce vaccination of newborns, screening for HBV, and HBV prophylaxis among pregnant women in Africa, particularly in West Africa, to achieve the WHO goal of HBV elimination, particularly in children.


Introduction
Hepatitis B virus (HBV) infection remains a major cause of acute and chronic liver disease with significant associated morbidity and mortality, worldwide. The World Health Organization (WHO) estimated that 296 million people were living with chronic HBV Studies conducted among West African populations residing outside Africa were excluded. Systematic reviews, commentaries and editorials, case report studies, surveillance reports, conference abstracts, animal studies, and articles describing the sero-prevalence of HBV only among adult populations were excluded. Studies with insufficient or inaccessible data, pre-prints, studies with a sample size < 100, and studies that did not describe their sampling technique or that used purposive sampling were excluded.

Data Sources and Study Screening Strategies
A systematic search, using key search words, was conducted on Africa Journals Online (AJOL), PubMed, Google Scholar, and Web of Science databases, supplemented by a manual search of retrieved references. The keywords used were HBV prevalence, children, HBV, HBV risk factors, HIV co-infection, and the names of the 14 countries of West Africa (Benin, Burkina Faso, Côte d'Ivoire or Ivory Coast, Cape Verde or Capo Verde, Gambia, Ghana, Guinea, Guinea-Bissau, Liberia, Mali, Mauritania, Niger, Nigeria, Senegal, Sierra Leone, and Togo). Prior to screening, all identified articles were imported into Covidence, a systematic review management tool [18]. Two reviewers (D.B.F. and A.M.S.) independently screened the titles and abstracts for study eligibility and full-text review. Reviewers then independently read the full text and selected studies for inclusion. After each step of the review, the two reviewers met and reached consensus on the final selection of studies for inclusion.

Description of the Study Area
West Africa is comprised of 14 countries (see above) and has one of the world's fastestgrowing populations, with more than 420 million people, equivalent to 5% of the total world population and with a median age of 18 years [19].

Data Extraction
After full-text review, the two reviewers, individually, extracted data from all eligible and retained articles, using the Covidence tool. Any differences observed by a reviewer in data extractio, were reconciled prior to conducting the meta-analysis. In the event of disagreement, a third reviewer (MIK) was consulted to establish consensus. Information extracted from the articles included: sociodemographic characteristics, sample size, the prevalence of HBsAg, risk factors including HIV co-infection, mother's HBsAg status, HBV contact(s) in the family, blood transfusion, circumcision, or other scarification, surgery, and receipt of HepB vaccine, including a birth dose.

Quality Assessment
The Newcastle-Ottawa scale (NOS) for cross-sectional studies quality assessment tool was used to assess the quality of each study [20]. The tool includes multiple sections. The first section focuses on the sample selection of each study. The second section deals with the comparability of the study including type of study, diagnostic method, and year of publication. The last section focuses on the statistical analysis of each study. A total NOS score ranges from 0 to 10. Study scores of 9-10 points are considered very good, 7-8 as good and 5-6 as satisfactory.

Data Analysis
StatsDirect, a statistical software (Version 3.0.0, StatsDirect Ltd., Cheshire, UK) was used to conduct the meta-analysis of the proportions of HBsAg in the retained studies [21]. Individual study proportions were assessed with a 95% confidence interval (CI) and the pooled effect. Sources of variation among studies were assessed by sub-group analysis, using the following grouping variables: study setting (hospital/other health care setting versus community setting), year of study, year of publication, and the prevalence of HBsAg among children, aged 0-16 years, stratified by "with" group risk factors, "without" group risk factors, and individual risk factors including HIV status, with HBsAg positive mother, and HBV vaccination including birth dose status.
We also assessed the study type, country, sample type, and testing method. Heterogeneity across the studies was assessed by the Quoran (Q) statistic test and the I 2 statistic. High heterogeneity (I 2 > 73% and p het < 0.05) represents the percentage of total variation across studies, attributable to heterogeneity rather than to chance [22]. Sources of heterogeneity were analyzed through sub-group analysis and one sensitivity analysis of HIV/HBV co-infection.
Publication bias was evaluated using funnel plot asymmetry and Egger's test [23]. In this test, p > 0.05 indicates an absence of evidence of publication bias (not significant). A random-effects model (REM) was used to pool HBV prevalence. Some representative results of the REM are presented graphically using forest plots.
We also assessed publication bias analysis in different subgroups. A sensitivity analysis was carried out by excluding 2 studies likely to strongly bias the results of HBV prevalence. Analyses of HBV prevalence were assessed with a 95% CI and p < 0.05 was considered as significant. All calculations were done using the StatsDirect software version 3.

Study Selection
The databases and other literature search tools yielded a total of 1162 articles. Figure 1 summarizes the flow chart for the selection of studies. Duplicate articles (n = 713), due to the diverse search tools used, were removed. Titles and/or abstracts were first used to screen the articles, according to the systematic review and meta-analysis criteria. After screening, 157 articles were further excluded because of a sample size < 100 and/or not within the publication period. However, data were extracted from four general population studies with more than 100 children and adolescents in each of the studies [24][25][26][27]. Of the 278 articles fully screened, 253 were excluded, based on the inclusion and exclusion criteria, as described in the Methods section. A total of 27 articles met the eligibility criteria for the meta-analysis. The included studies were conducted in seven of the fourteen West African countries, Nigeria (thirteen articles), Senegal (five articles), Burkina Faso and Ghana (three articles each), Benin, Togo, and Ivory Coast (one article each), as shown in Table 1.

Characteristics of the Systematic Review Studies
Four were cohort studies, nineteen were cross-sectional, three were retrospective cohort studies, and one was case-control. A total of 11,304 children, aged 0 to 16  Multiple techniques were reported to detect HBsAg, with direct enzyme-linked immunoassay test (ELISA) being the most common, followed by rapid diagnostic assays (RDT) using serum or plasma serological markers of HBV.
Nineteen (n = 19), 70% of the studies were conducted in hospital settings and eight (30%) in community settings. Most of the studies involved children only, although we extracted data from four studies of the general population that included >100 children, 0-16 years old [24][25][26][27].
Nine (=9), 33% of studies included persons living with HIV (PLWH), including a total of 2317 children, and three studies included high-risk children born to HBsAg-positive mothers. Most studies were conducted in urban areas, with only four studies being in mixed urban and rural areas. Seven (n = 7), 26% of studies included HepB vaccine status; however, most studies were missing vaccination information (Supplementary Data S2). Adult population (119) Wrong intervention (107) Missing testing data (8) No abstract (3) Wrong outcomes, study design (6) Other Viruses infection (6) Studies included in qualitative synthesis (n = 28)

Studies included in quantitative synthesis
(meta-analysis) (n = 27)   (3/210 = 1.4%) and Benin 103 (10/103 = 9.7%), (Supplementary Data S1). We observed a decrease in the prevalence of HBV between 2000 and 2021. HBV prevalence by country and study periods calculated by the StatsDirect software are presented in Figure 2 and Figure 3, respectively.   Multiple techniques were reported to detect HBsAg, with direct enzyme-linked immunoassay test (ELISA) being the most common, followed by rapid diagnostic assays (RDT) using serum or plasma serological markers of HBV.
Nineteen (n = 19), 70% of the studies were conducted in hospital settings and eight (30%) in community settings. Most of the studies involved children only, although we extracted data from four studies of the general population that included >100 children, 0-16 years old [24][25][26][27].
Nine (=9), 33% of studies included persons living with HIV (PLWH), including a total of 2317 children, and three studies included high-risk children born to HBsAg-positive mothers. Most studies were conducted in urban areas, with only four studies being in mixed urban and rural areas. Seven (n = 7), 26% of studies included HepB vaccine status; however, most studies were missing vaccination information (Supplementary Data S2).

Overall Pooled HBV Prevalence in Children in West Africa and Publication Bias
HBV prevalence among children, 0-16 years old, varied widely in West Africa ( Figure 4A). Crude overall HBV prevalence in the pooled sample of 11,304 children was 5% (95%, CI 4-7%). The seroprevalence data are presented in the random effect model because of the substantial heterogeneity of HBV prevalence across the studies (I 2 > 94.2%, 95% CI = 93% to 95.1%). Egger's test was significant (p < 0.001) for HBV prevalence, suggesting publication bias in the studies ( Figure 4B).

Sensitivity Analysis
To further assess the strength of the HBV prevalence results, we conducted a leavetwo-out sensitivity analysis by removing two studies that included two high-risk populations: a study of HBV-HIV co-infected children with high HBV prevalence [37] and a study of infants born to HBsAg positive mothers [27]. The overall HBsAg positive pooled prevalence rate was 5% with heterogeneity prior to the exclusion of these two studies (I 2 of 94.2%, 95% CI = 93% to 95.1%, p < 0.0001). However, after the exclusion, the pooled prevalence rate only dropped to 4% (95% CI 3% to 6%) with heterogeneity (I 2 of 91.2%, 95% CI = 88.8% to 92.9%), indicating that the pooled HBV prevalence was not affected by the two studies and that the results are robust (Supplementary Data S3).

HBV Prevalence in Children with or without Risk Factors
Overall, HBV prevalence in children ranged from 4% (95%CI: 0.03 to 0.06) in children without risk factors, ( Figure 5A) to 9% (95%CI: 5 to 15%) in children with at least one risk factor, such as HIV/HBV co-infection or being born to an HbsAg positive mother ( Figure  5B).

Sensitivity Analysis
To further assess the strength of the HBV prevalence results, we conducted a leave-twoout sensitivity analysis by removing two studies that included two high-risk populations: a study of HBV-HIV co-infected children with high HBV prevalence [37] and a study of infants born to HBsAg positive mothers [27]. The overall HBsAg positive pooled prevalence rate was 5% with heterogeneity prior to the exclusion of these two studies (I 2 of 94.2%, 95% CI = 93% to 95.1%, p < 0.0001). However, after the exclusion, the pooled prevalence rate only dropped to 4% (95% CI 3% to 6%) with heterogeneity (I 2 of 91.2%, 95% CI = 88.8% to 92.9%), indicating that the pooled HBV prevalence was not affected by the two studies and that the results are robust (Supplementary Data S3).

HBV Prevalence in Children with or without Risk Factors
Overall, HBV prevalence in children ranged from 4% (95%CI: 0.03 to 0.06) in children without risk factors, ( Figure 5A) to 9% (95%CI: 5 to 15%) in children with at least one risk factor, such as HIV/HBV co-infection or being born to an HbsAg positive mother ( Figure 5B). A funnel plot of HBV prevalence in children with and without HIV shows a not strictly symmetrical display of the prevalence reported by the individual studies ( Figure  3). However, the random effects model (DerSimonian Laird) suggests that there is evidence of publication bias, as revealed by the Egger's test, with a bias = 4.447819 (95% CI = 3.043189 to 5.852449) and p < 0.0001.

Risk Factors of HBV Infection in Children
An analysis of children by individual risk factor shows HBV prevalences of 12%, 8%, 5%, and 1% in children born to HBsAg positive mothers, HBV/HIV co-infected, unvaccinated with a birth dose, and vaccinated, respectively, as shown in Figure 6. A funnel plot of HBV prevalence in children with and without HIV shows a not strictly symmetrical display of the prevalence reported by the individual studies ( Figure 3). However, the random effects model (DerSimonian Laird) suggests that there is evidence of publication bias, as revealed by the Egger's test, with a bias = 4.447819 (95% CI = 3.043189 to 5.852449) and p < 0.0001.

Risk Factors of HBV Infection in Children
An analysis of children by individual risk factor shows HBV prevalences of 12%, 8%, 5%, and 1% in children born to HBsAg positive mothers, HBV/HIV co-infected, unvaccinated with a birth dose, and vaccinated, respectively, as shown in Figure 6. A funnel plot of HBV prevalence in children with and without HIV shows a not strictly symmetrical display of the prevalence reported by the individual studies ( Figure  3). However, the random effects model (DerSimonian Laird) suggests that there is evidence of publication bias, as revealed by the Egger's test, with a bias = 4.447819 (95% CI = 3.043189 to 5.852449) and p < 0.0001.

Risk Factors of HBV Infection in Children
An analysis of children by individual risk factor shows HBV prevalences of 12%, 8%, 5%, and 1% in children born to HBsAg positive mothers, HBV/HIV co-infected, unvaccinated with a birth dose, and vaccinated, respectively, as shown in Figure 6.

HBV Prevalence in Hospital and Community Settings
Subgroup analysis showed higher HBV prevalence in studies conducted in hospital settings [7% (95% CI. 4 to 11%)] compared to studies conducted in community settings among children without any risk factors [3% (95%CI: 2 to 4%)] ( Figure 7A,C).

HBV Prevalence by HIV Status in Children
HBV/HIV co-infected children had an overall HBV prevalence of 9% (95% CI: 4 to 15%), with an Egger bias of 6.253247 (95% CI = 2.533005-9.97349; p = 0.0054), corresponding to low heterogeneity between studies, as opposed to studies of HBV mono-infected children who had an overall HBV prevalence of 4% (95% CI, 2 to 6%) and an Egger bias of 3.7865 (Figure 8).

Discussion
A systematic review of HBV prevalence in children in West Africa aged 0-16 years yielded 27 studies meeting inclusion criteria for a meta-analysis. The studies, however, only included data from seven of West Africa's fourteen countries. Most studies were cross-sectional; three-quarters used the Enzyme-Linked Immunosorbent Assay (ELISA) as the HBV diagnostic tool, and most were conducted in a hospital setting (18/27) in an urban area. Indeed, most healthcare facilities in West Africa are in urban areas.
Overall, HBV prevalence was 5%, based on the random analysis, given the heterogeneity of the studies. This meta-analysis revealed that HBV prevalence in children in West Africa is moderate, according to the WHO's criteria for HBV endemicity [51]. HBV prevalence by country ranged from 1% in Togo to 10% in Benin. The prevalence was also variable inside the same country, such as in Nigeria or Ghana, for which there are a high number of published data available. Differences in health system resources across West African countries, although already generally limited in all of West Africa, may still explain some of the discrepancies. In fact, some countries such as Mali and Benin lack resources to screen pregnant women or provide vaccine doses at birth for newborns. Findings from a study with timely dose of hepatitis B Birth dose (HepB-BD) in Africa demonstrated the variability of birth dose implementation and the challenges countries face in immunizing babies [52]. In the absence of efforts for preventing mother-to-child transmission, it is anticipated that HBV prevalence in children in these countries would stay higher and fall short of the WHO targets.

Discussion
A systematic review of HBV prevalence in children in West Africa aged 0-16 years yielded 27 studies meeting inclusion criteria for a meta-analysis. The studies, however, only included data from seven of West Africa's fourteen countries. Most studies were cross-sectional; three-quarters used the Enzyme-Linked Immunosorbent Assay (ELISA) as the HBV diagnostic tool, and most were conducted in a hospital setting (18/27) in an urban area. Indeed, most healthcare facilities in West Africa are in urban areas.
Overall, HBV prevalence was 5%, based on the random analysis, given the heterogeneity of the studies. This meta-analysis revealed that HBV prevalence in children in West Africa is moderate, according to the WHO's criteria for HBV endemicity [51]. HBV prevalence by country ranged from 1% in Togo to 10% in Benin. The prevalence was also variable inside the same country, such as in Nigeria or Ghana, for which there are a high number of published data available. Differences in health system resources across West African countries, although already generally limited in all of West Africa, may still explain some of the discrepancies. In fact, some countries such as Mali and Benin lack resources to screen pregnant women or provide vaccine doses at birth for newborns. Findings from a study with timely dose of hepatitis B Birth dose (HepB-BD) in Africa demonstrated the variability of birth dose implementation and the challenges countries face in immunizing babies [52]. In the absence of efforts for preventing mother-to-child transmission, it is anticipated that HBV prevalence in children in these countries would stay higher and fall short of the WHO targets.
Due to the significant burden of HBV-related diseases in SSA, where >8% of the general population is chronically infected, HBV infection is still ubiquitous across West Africa. This can impact HBV prevalence in these countries [53] and the prevalence is around 5% among pregnant HBV/HIV co-infected women [54]. There has been a significant decrease in the prevalence of HBsAg among children in West Africa since the year 2000. This is clearly demonstrated by comparing the 12% HBsAg prevalence in children and adolescents < 19 years old in West African SSA in 1990 [51], and the prevalence of HBV infection in children at 2.53% in 2019, amounting to 360,000 infected children each year [55]. We believe that the decrease may be related to the introduction of infant HBV vaccination programs in most African countries and including the HepB-BD in a few countries.
Studies conducted in hospital settings showed higher prevalence of HBV infection (7%) and this is likely related to the studies being conducted on sick children, who are also tested at higher rates than non-hospitalized children. By contrast, healthy children in the community setting had an HBV prevalence of only 3% and may represent early success of the newborn HepB vaccination programs. Prevalence varied from 3% to 9% among infants with HBV/HIV co-infection, those born to mothers who tested positive for HBsAg, and those who had other risk factors, such as having undergone surgery, scarification, or not receiving vaccinations. Other risk factors were not available in many studies. In one Nigerian study, risk factors such as previous history of jaundice (p = 0.26), blood transfusion (p = 0.24), past history of surgery (p = 0.47), or scarification marks (p = 0.17) were not associated with HBV prevalence [26].
HIV and HBV infections have many similarities, such as common routes of transmission, high prevalence in certain geographical regions, same at-risk groups, and the risk of mother-to-child transmission. All these factors contribute to a significant association between HBV and HIV co-infection in the pediatric population [56,57]. We found an overall HIV/HBV co-infected prevalence of 9%. However, in a sensitivity analysis, we removed one study from Burkina Faso (2001) that reported a co-infection rate of 40%, and yielded a prevalence of 6%, ranging from 1.15% in Nigeria (2021) to 10% in Benin (2015) (Figure 3). The finding from this study is consistent with prior studies. In studies from South Africa, HIV/HBV co-infection prevalence in children ranged from 5-17%, with the higher prevalence occurring in the industrialized towns where mining activity is associated with increased sexual activity. In Nigerian studies, HBV/HIV co-infection prevalence varied by geopolitical region, ranging from 5.8 to 19% [13,29,38,40,58]. Finally, a Tanzanian study of HBV/HIV co-infected children reported 1.2% prevalence [59]. Reported HBV/HIV co-infection prevalence in children has been somewhat higher in Nigeria (7.8%) [29], 10.4% in Zambia [60], and 12.1% in Ivory Coast [61]. The generational effect benefits of HepB vaccination may be a factor in the relatively lower HBV-HIV co-infection in later studies compared to earlier studies. It is also important to note that HIV/HBV co-infection considerably increases the risk of mother-to-child transmission, if the mother is infected and untreated.
Perinatal and childhood acquisition of HBV not only leads to increased risk of chronicity but also strongly predict worse long-term outcomes for liver cirrhosis and hepatocellular carcinoma [16]. HIV/HBV co-infection in childhood further places children at high risk for associated morbidity and mortality, similar to adults, hence the urgent need to implement HepB vaccination at birth, routine screening, and follow-up. In fact, despite more than two decades since the introduction of HepB vaccination programs, the overall prevalence of HBV infection still remains high in many settings in SSA [53,62,63].
The difference in HBV prevalence infection between vaccinated and unvaccinated children is still substantial (2% versus 6%). While there is a residual possibility of HBV infection despite newborn HepB vaccination, often, it is unclear if infection occurred before or after completion of the HepB vaccine series [41]. In SSA, horizontal transmission in children, aged 6 months to 5 years, is common due to close interactions with infected household contacts and playmates. We examined the impact of HepB newborn vaccination compared to the evolution of the epidemic in each country. In Nigeria, overall HBV prevalence was 7% and a case-control study found that HBsAg prevalence was significantly lower among vaccinated (1.4%) compared with unvaccinated (4.8%) children [49]. In Senegal and Ghana, where all the studies were conducted after newborn HepB vaccination had been introduced, overall HBV prevalence was only 2%. In Togo, one in ten women of childbearing age was infected with HBV, but less than 2% of infants under five years of age who received the HepB vaccine at birth were infected with HBV. These results are similar to other studies that found significant reductions in HBsAg positivity post-vaccination with a protective efficacy of between 67-94% [64,65]. Such a high level of vaccine efficacy is likely to positively impact and prevent community transmission of HBV.
In 1992, the World Health Assembly adopted a resolution recommending the introduction of HepB into national immunization programs [66]. In 2016, the WHO and the World Health Assembly established the target of controlling and eliminating HBV worldwide by 2030. The WHO recommends that all infants receive their first dose of HepB as soon as possible after birth, ideally within 24 h, followed by two or three doses of HepB at least 4 weeks apart to complete the vaccination series [67]. Despite this recommendation, HepB vaccination at birth has not been widely implemented in most national vaccination programs in SSA, particularly in West Africa, despite the Expanded Program of Immunization (EPI) [68]. Indeed, in most West African countries, parents must still pay for the first HepB dose, which, at~$US 8 per dose, is prohibitive, given that 85% of the population in SSA live on less than $5.50/day [69]. In 2021, coverage with timely HepB-BD was 42% globally and only 17% for infants in the WHO African region [70]. A total of 114 countries worldwide had introduced HepB-BD in their routine immunization schedule [71]. Yet, this number includes only 14 (30%) of 47 countries in the WHO African region [72]. Limited countries have data on Studies of the Effectiveness of HepB-BD Vaccination in Africa with only two studies published to date. In 2001-2002, Ekra and colleagues conducted a nonrandomized controlled trial in four health centers in Abidjan, Cote d'Ivoire [73] and a second effectiveness study was conducted from 2009-2016 in a single center in Tokombéré district, Cameroon [74]. Based on the results of those published studies, the authors assumed a high-level transmission rate in the absence of vaccination and highlighted the benefit of the addition of HepB-BD to the three-dose HepB vaccination schedule for infants. In addition, a residual risk of motherto-child transmission of hepatitis B virus infection despite timely birth-dose vaccination in Cameroon has been reported. In addition, studies from Hawaii, Taiwan, and China provide assurance that routine infant HepB immunization beginning with HepB-BD vaccination is a highly effective public health strategy that will progressively protect generations to come from HBV-related liver disease, HCC, and premature mortality [75,76].
The universal vaccination strategy implemented in Thailand provides evidence of the effect of newborn HepB vaccination (HepB-BD) in eliminating HBV infection [77]. As more African countries seek to implement HepB-BD, attention to disparities in implementation need to be addressed particularly in rural and underprivileged settings [63]. Maternal education and community engagement are essential to scale up HepB-BD in SSA through the Global Alliance for Vaccines and Immunisation (GAVI). A study in Nigeria demonstrated that missed doses were largely avoided when staff completed a vaccine checklist before releasing mother-child pairs [78]. There must also be a focus on offsetting the costs of distributing the HepB-BD throughout SSA.
In endemic regions of Africa and Asia, in-utero infection of the fetus, vertical transmission, constitutes the main mode of HBV transmission [79]. A meta-analysis indicated that maternal viral load was an important risk factor for mother-to-child transmission (MTCT) in HBeAg-positive mothers, and maternal viral load was dose-dependent with HBV MTCT incidence [80]. Other studies have shown that a high viral load in HBsAg-positive mothers can lead to vaccination failure in the newborn, even if combined immunoglobin treatment and vaccination at birth are delivered [81].
HBsAg screening for all pregnant women is critical. Focusing on HBsAg positive women and providing prophylactic treatment to women with high viral loads can be an effective approach to reduce transmission to the infant [82]. A Senegalese study of HBV in children born to HIV-positive mothers showed a low rate of HBsAg (2.6%) if the mother was treated with lamivudine (3TC) or tenofovir (TDF), compared to 7.9% if the mother was untreated [46]. Another Senegalese study showed low HBV prevalence, but only 56% of children had a sero-protective level >10 UI/L [34]. It is, therefore, strongly recommended to vaccinate children and to adhere to the necessary doses to protect them against HBV infection.

Limitations
This meta-analysis has several limitations, particularly the substantial heterogeneity of eligible studies. Prior HBV prevalence meta-analyses in Africa have also had high levels of heterogeneity [83,84]. Data about children, aged 0-16 years, from four general population studies were used and were the only data available from several countries. However, it was challenging to analyze risk factors for children separately from adults in these studies. The quality of the studies varied between countries, as well as within the same country. Studies showed that protective levels of HbsAb antibodies decrease with age after vaccine introduction [30,43]; however, most of the studies included in this review did not report prevalence by age group.

Conclusions
In conclusion, this study using the most recent data available estimated HBV prevalence in children aged 0-16 years in West Africa, revealing a decrease in prevalence over the past two decades; yet, a persistently high prevalence in high-risk child populations still exists. The studies are robust and cover periods before and after the introduction of HepB vaccination in different West African countries. The meta-analysis identified wide variation in HBV prevalence, depending on the country, as well as in subgroups of HBV/HIV co-infected children, children born to HBsAg-positive mothers, and vaccinated or unvaccinated children. Overall, pooled prevalence and pooled subgroup prevalence remain moderate in West Africa and reaffirm the likely impact of newborn HepB vaccination. The study reinforces the need to implement HepB-BD and screening and prophylaxis of HBV in pregnant women. These interventions are essential to prevent mother-to-child transmission in order to achieve the WHO goal of targeted HBV elimination.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.