Initially, a total of 2729 studies were identified from the databases and manual searching. From this, 1310 of the studies were removed due to duplication. The remaining 1419 articles were screened by their title and abstract and 1312 of the studies were excluded. Further, 107 full-text articles were refined and 47 of them were excluded due to being unrelated to the current study, studies on immigrants, review articles, studies conducted before 2010, and studies on immunization. Finally, a total of 60 [19, 28-86] studies were fulfilled the inclusion criteria and enrolled in the study [Fig. 1].
Fig. 1 PRISA flow diagram for identification and selection of articles for inclusion in the review
Characteristics of included studies
A total of 60 articles were included in this systematic review and meta-analysis, with an overall sample size of 106125 that conducted on the prevalence of HBV in Ethiopia. All the included studies were cross-sectional studies. The most recent study was conducted in 2019. Regarding the included studies on prevalence of HBV, more than of the half of the studies were obtained from Amhara region 22 (36.7%) [19, 28, 30, 35-37, 43, 45, 48, 49, 52, 57, 61, 62, 66-68, 70, 74, 77, 78, 81], Oromia region 12(20%) [31, 33, 40, 46, 47, 55, 59, 60, 63, 82, 84, 86], and Southern Nations, Nationalities and Peoples Region (SNNPR) 9(15%) [29, 42, 51, 64, 65, 69, 73, 75, 85]. Except. One study (prospective study design), all the included studies were conducted using cross-sectional study designs. The sample size across the studies was ranged from 108  to 35435  that obtained from the Amhara region. In addition, the quality of each of the included studies was evaluated using nine items risk of bias too tool (Table 1).
Table 1 Characteristics of the included studies in the systematic review and meta-analysis for the prevalence of hepatitis B virus in Ethiopia, 2019.
Key: SAC: Self-Administrative City; SNNPR: Southern Nations, Nationalities and Peoples’ Region; ELISA: Enzyme-Linked Immunosorbent Assay; RDT: Rapid Diagnostic Test; IA: Immunoassay; *: not stated; VCT: volunteer for counseling and testing
Prevalence of HBV in Ethiopia
There was a wide HBV prevalence variation among included studies which is ranged from 1% in Amhara region to 36% in Addis Ababa city. Based on the random-effects model, the overall pooled prevalence among 106125 was 6% (95% CI: 5% to 6%) with heterogeneity index (I2) of 97.77% (p<0.001) (Fig. 2).
Fig. 2 The pooled prevalence estimate of HBV in Ethiopia from 2010 to 2019.
Since this meta-analysis exhibited a considerable heterogeneity, subgroup analysis was done using study group, study quality, region/ city where the studies were conducted, year of publication, data collection year, sampling technique, screening method, and the setting was considered to identify the possible sources of heterogeneity among the studies. The subgroup analysis indicated that the heterogeneity level was slightly reduced among pregnant women (I2=52.2%), HIV positive study participants (I2=64.59%), studies conducted in SNNPR (I2=67.13%), studies conducted using probability sampling technique (I2=80.13%), and among studies those did not clearly indicated their sampling techniques (I2=% 41.91%). Geographically, the highest and the lowest prevalence of HBV were obtained from the Addis Ababa city 10% (95% CI: 6%, 15%) and Amhara region 4% (95% CI: 3%, 5%), respectively. Studies conducted on nonprobability sampling accounted for the highest 7% (95% CI: 5%, 9%) followed by studies conducted through a survey. Concerning screening techniques, studies conducted using ELISA accounted for the least prevalence estimate of 4% (95% CI: 4%, 5%) (Table 2).
Table 2 Subgroup analysis of the HBV pooled prevalence estimation in Ethiopia, 2019.
Key: ELISA: Enzyme-Linked Immunosorbent Assay; IA: Immunoassay; RDT: Rapid Diagnostic Test; SNNPR; Southern Nations, Nationalities and Peoples Region
Meta-regression and sensitivity analysis
A meta-regression analysis was done on categorical variables including the year of study, year of publication, study group, region, sample size, sampling technique, quality score, and screening method. Among these, the year of the data collection was marginal. The remaining covariates including region/ city (p=0.04), study group (p=0.005), screening method (p=0.017) and quality of papers (p=0.001) showed statistically significant association with HBV prevalence (Table 3).
Table 3 Meta-regression analysis of factors for the heterogeneity of HBV prevalence in Ethiopia, 2019.
Publication bias and small study effects
The presence of publication bias was evaluated using funnel plots and Egger’s test. Each point in funnel plots represents a separate study and asymmetrical distribution indicates the presence of publication bias . First, studies’ effect sizes were plotted against their standard errors and the visual evaluation of the funnel plot indicated no publication biases as the graph appear symmetrical (Fig. 3). The subjective evidence from the funnel plot was objectively confirmed using the Egger’s weighted regression statistics. According to the symmetry assumption, the p-value of 0.747 declares the absence of small study effects among the included studies.
Fig. 3 Funnel plot of the prevalence of HBV in Ethiopia from 2010 to 2019.