Objective This study aims to assess the prevalence of early initiation of breast feeding (EIBF) and associated factors among mothers having children less than 2 years of age in Ethiopia. Design Community-based cross-sectional study. Setting In this analysis, data from 2019 Ethiopia Mini Demographic and Health Survey (EMDHS) was used. The survey included all the nine regional states and two city administrations of Ethiopia. Participants We extracted data of 2054 mothers who had last-born children and those mothers who ever breast fed or still breast feeding their children during the survey from the 2019 EMDHS datasets. Main outcome measures We performed a two-stage multilevel mixed-effects logistic regression to identify individual and community-level determinants of EIBF. In the final model, variables with a p-value less than 5% and an adjusted OR with a 95% CI were reported as statistically significant variables with EIBF. Result The prevalence of EIBF among mothers having children aged 0-23 months was 73.56% (95% CI: 71.65% to 75.47%). Women who delivered at a health facility (adjusted OR (AOR)=1.98; 95% CI: 1.39 to 2.79) and have children with birth order second-fourth (AOR=1.76; 95% CI: 1.24 to 2.49) were more likely to initiate early breast feeding than their counterparts. On the other hand, women who gave birth by caesarean section (AOR=0.21; 95% CI: 0.13 to 0.33), had multiple births (AOR=0.35; 95% CI: 0.13 to 0.92) and had postnatal check-up (AOR=0.62; 95% CI: 0.44 to 0.91) were less likely to practise EIBF as compared with their counterparts. Region of residence of women was also significantly associated with EIBF. Conclusion In this study, the overall prevalence of EIBF was good. Place of delivery, mode of delivery, postnatal check-up, type of birth, birth order and region were factors significantly associated with EIBF. Therefore, government and stakeholders need to show commitment to improve access and utilisation of basic maternal health services to increase the practice of EIBF.
This study used a secondary data of the 2019 EMDHS. Originally, the survey sample was stratified and selected in two stages. The country is stratified into nine regions and two city administrations. Then, each region was stratified into urban and rural areas. In the first stage of selection, 305 enumeration areas (EAs) were selected using probability proportional to EA size according to the sampling frame created for the upcoming Ethiopian population and housing census. Consecutively, a list containing household in all selected EAs was developed. In the second stage, a fixed number of households (30) per cluster was selected from the newly created household listing using an equal probability systematic selection. Additional data on survey sampling strategies are provided in the DHS handbook.18 Generally, all women in childbearing age (15–49 years) who were either permanent residents of the selected households or visitors who slept in the household the night before the survey were eligible for the survey. The source population for the present study were mothers who breast fed and had children less than 2 years of age. All mothers of last-born children born in the 2 years preceding the survey (both surviving and dead) were included in the analysis, whereas mothers who had never breast fed their children were excluded from the study. Hence, data of 2054 mothers who had last-born children and those mothers who ever breast fed or still breast feeding were extracted from the 2019 EMDHS datasets for this analysis. Figure 1 depicts the method of selection of study participants that we followed to identify eligible mothers for this study (figure 1). Eligibility assessment for early initiation of breast feeding (EIBF) among women having children aged 0–23 months in Ethiopia, 2019. The data for this study come from a standardised community survey of the 2019 EMDHS data, which was conducted in Ethiopia from March 2019 to June 2019. It included all the nine regional states and two city administrations of Ethiopia. EIBF is the outcome variable of the study. It is defined as giving breast milk within the first hour of birth to the last child born in the last 2 years preceding the survey.18 During the survey, all women were asked how long after their babies were born to breast feed for the first time. According to the self-report of the child’s mother, it is coded as ‘1’ if the child is breast fed within the first hour after birth and ‘0’ otherwise. The covariates in this study include variables at the individual and community level. The selection of explanatory variables is based on data from previous similar studies and the availability of variables in the 2019 EMDHS dataset. Maternal age, educational level, marital status, religion, family size, wealth index and possession of radio/television were individual-level sociodemographic and economic characteristics of the mothers included in the analysis. Obstetric and variables related to the use of health services were parity, number of antenatal care (ANC) visits, place of delivery, mode of delivery, type of delivery assistance, postnatal check-up, type of birth and counsel on breast feeding during the first 2 days of delivery. Other individual-level factors included in the analysis were child-related characteristics. These were age of child in months, sex of child, number of living children, birth order, preceding birth interval and child lives with whom. Place of residence, region, community-level women education, community-level health facility delivery, community-level ANC utilisation and community poverty level were community-level variables included in the analysis. Community-level factors, which were not directly obtained from the survey dataset, were derived by aggregating individual-level factors (table 1). Description of individual and community-level variables of early initiation of breast feeding among mothers of children under 2 years in Ethiopia, 2019 ANC, antenatal care; DHS, Demographic and Health Survey; HH, household; PCA, principal component analysis; SNNPR, Southern Nations, Nationalities, and People’s Region. STATA V.14 was used to clean, recode and analyse the 2019 EMDHS kids’ data. Sample weight was applied to adjust for sampling error and for non-responses. Descriptive statistics were used to present the distribution of background characteristics among the study participants. We employed a two-level multilevel mixed-effects logistic regression analysis so as to account for the hierarchical nature of the EMDHS data and to identify the true association between the individual and community-level factors and EIBF. Consequently, four models containing variables of interest were fitted: null model (without any explanatory variables), model I (with only individual-level variables), model II (with only community-level variables) and model III (with both level variables). Intraclass correlation coefficient (ICC) and a proportional change in variance (PCV) were tested to determine the clustering effect and the degree to which community-level factors explain the unexplained variance of the null model. Moreover, goodness of fit of the model was assessed by the deviance and the model with the lowest deviance was used as the best-fitted model. Finally, multicollinearity between explanatory variables was checked by variance inflation factor (VIF) and a mean value of VIF <10 indicates absence of multicollinearity. Both individual and community-level factors having a p-value of less than 0.2 in the bivariable analysis were selected as candidate variables for the multivariable multilevel mixed-effects logistic regression analysis. Likewise, variables with a p-value less than 5% and an adjusted OR (AOR) with a 95% CI were reported as statistically significant variables with EIBF in the final model. Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.