As described in Fig.1, 751 studies were identified regarding EBF in Ethiopia through PubMed, Google Scholar, Science Direct, Scopus, CINAH, Web of Science, and others in the first step. Then 200 studies were excluded because of duplication. From the remaining 551 studies, after duplication was removed, 299 articles were excluded after reviewing their titles based on assessment since they were not relevant to this review. The remaining 252 studies were screened by abstracts yielding an additional 189 studies to be excluded. Moreover, 63 full-text articles were assessed for eligibility based on the preset inclusion criteria. From these 18 articles were excluded due to the inclusion criteria. Among excluded studies, three of the studies were excluded because they didn't report our outcome of interests [73, 74,75]. Two studies were conducted among HIV-infected mothers [76, 77]. The rest 12 articles were excluded because of publication year; one Study published in 2009, two studies were published in 2012[79,80], three of the studies were published in 2013[81,82,83], six studies were published in 2014[84,85,86,87,88,89] and one study excluded by quality assessment criteria. Finally,45 studies fulfilled the inclusion criteria and included in the systematic review and meta-analysis.
As shown in Table1, all of these 45 studies were published between January 1/2015 to December 1/2020. A total of 33000 breastfeeding women were included to estimate the pooled prevalence of exclusive breastfeeding in the current Meta-analysis. Regarding study design most 42 of the studies are cross-sectional study design. The sample size of the studies ranges from 226 - 5,227. The lowest prevalence (29.29%) of EBF was observed in a study conducted in Addis Ababa, Ethiopia  whereas the highest prevalence (87.84%) was observed in a study conducted SNNP and Tigray region, among rural mothers in Ethiopia . From nine regions of Ethiopia, seven regions and two council cities were represented in this meta-analysis.
Fifteen of the studies were from Amhara [34-43,55,68,116,117,119], two from Addis Ababa [52,53], four from Affar [33,58,63,115], four from Oromia [45,64,65,122], twelve from SNNP [46-48,51,52, 54,57, 59, 60,61,62,66], one from Tigray , two from Somalia[49,121],two from Harari[44,120],one from nationwide,one from Diredawa and one from SNNP and Tigray. No studies reported from Benishangul Gumiz and Gagmbela in this review research. Regarding the quality score of each primary study, the score was between the lowest 3 and highest 8(additional file 1) and almost all primary studies had sufficient response rate.
Pooled Prevalence of exclusive breastfeeding in Ethiopia
The pooled prevalence of EBF in Ethiopia from 45 included studies is 60.42% (95% CI: 55.81-65.02%). The I2 test result indicated high heterogeneity among included studies (I2 98.8%, p = <0.001), and because of this high heterogeneity the random effect meta-analysis model was used. In this meta-analysis research, Hagos et al (87.84%) and shiferaw et al (29.29%) reported the highest and lowest prevalence of exclusive breastfeeding respectively (Fig.2).
Besides, we conducted a Univariate meta-regression by considering the year of publication, sample size, and region in Ethiopia as covariates to identify the possible sources of heterogeneity, and unfortunately, none of them were found to be statistically significant(Table2). Besides, publication bias was assessed subjectively and objectively using funnel plot and Begg's and Egger's tests respectively. Even if the funnel plot showed the presence of publication bias (Fig.3), Begg's and Egger's tests showed no statistical significance for estimating the prevalence of EBF (p = 0.304) and (p = 0.314) respectively.
A subgroup analysis was done by an administrative area (region) in Ethiopia, the year of publication, and the Total sample size of primary studies to compare the prevalence of exclusive breastfeeding across different studies. Accordingly, this systematic review and meta-analysis showed that the highest prevalence reported in a study conducted in SNNP and Tigray 87.84(85.19, 90.48), a study published during (2015-2016) 64.60 %( 52.90, 76.30) and among studies with sample size with less than 500 (64.15(58.61, 69.68)) respectively (Table3).
Association between maternal employment and exclusive breastfeeding in Ethiopia: We examined the association between maternal employment and EBF practice using 19 studies [33-50,118] in this meta-analysis and the finding showed that the practice of EBF was significantly associated with maternal employment. Those employed mothers were 57% less likely to practice exclusive breastfeeding in comparison to unemployed mothers [OR] 0.43; 95% CI (0.31, 0.61).High heterogeneity (I2 = 85.0% and p-value < 0.000) was observed across the included studies and a random effect meta-analysis model was applied to examine the association between maternal employment and EBF in Ethiopia (Fig.4). We also assessed publication bias subjectively using the funnel plot and objectively using Begg's and Egger's tests. While the funnel plot showed the presence of publication bias, Begg's and Egger's tests showed the absence of significant publication bias (p- value=0.363 and p=0.684) respectively (Fig.5).