Prelacteal feeding practice and its determinant factors among mothers having children less than 6 months of age in Bure district, Northwest Ethiopia: A community-based cross-sectional study

Objective The main aim of this study was to assess prelacteal feeding practice and its determinant factors among mothers having children less than 6 months of age in Bure district, Northwest Ethiopia. Design Community-based cross-sectional study design. Setting Northern Ethiopia. Participants The present study was conducted among 621 mothers who had children less than 6 months of age in Bure district, Northwest Ethiopia, from 1 March 2019 to 30 March 2019. Primary outcome Mothers prelacteal feeding practice, modelled using multivariable logistic regression. Secondary outcome Determinant factors of prelacteal feeding practice. Results This study found that the prevalence of prelacteal feeding practice was 11.6% ((95% CI 9.0% to 14.2%)). Delayed initiation of breast feeding (adjusted OR, AOR=5.4, 95% CI 2.2 to 13.5), mothers who did not get counselling of breast feeding (AOR=2.9, 95% CI 1.2 to 7.2), home delivery (AOR=6.9, 95% CI 2.2 to 21.5), primiparous mothers (AOR=4.1, 95% CI 1.4 to 12.2), a newborn with history of neonatal illness (AOR=3.3, 95% CI 1.3 to 8.5) and lack of postnatal care visits (AOR=3.9, 95% CI 1.3 to 11.8) were determinant factors of prelacteal feeding practice. Conclusions Delayed initiation of breast feeding, mothers who did not get counselling of breast feeding, home delivery, primiparous mothers, newborns with a history of neonatal illness and lack of postnatal care visits were determinant factors of prelacteal feeding practice. Therefore, healthcare workers should provide a home to home health education for mothers on the merits of early initiation of breast feeding, promote institutional delivery, enhance maternal health-seeking behaviour and encourage mothers to have postnatal care visits is recommended. A community-based cross-sectional study design was conducted in Bure district, Northwest Ethiopia from 1 March 2019 to 30 March 2019. This study was part of our previous study, which was published elsewhere. Both studies were targeted to understand the factors that determining optimal breastfeeding practices in Ethiopia where neonatal mortality was unacceptably high. Additionally, we have used a similar methodology to access the study participants.20 Bure district is located in Amhara Regional State, West Gojjam Administrative Zone, Ethiopia. It is one of the 15 districts of West Gojjam Zone. It is found 148 km southwest of the Regional State capital, Bahir Dar, and 400 km northwest of Addis Ababa, the capital city of Ethiopia. According to the Bure district health office, the total number of children less than 6 months was 2882, under 5 children, 17 956, women with childbearing age of, 31 221, and pregnancy/delivery was 3469. There are 5 functional health centres, 4 medium private clinics and 20 health posts that provide primary healthcare services to the catchment population during the data collection period. In 2018, the district health office indicates that institutional delivery was 38%, antenatal care (ANC) coverage was 90%, postnatal care (PNC) coverage was 85% and immunisation coverage was 95%. All mothers who had children less than 6 months of age in Bure district were considered as source population and mothers who were found in the randomly selected kebeles were considered the study population. All mothers who had children less than 6 months of age, permanent residents and who lived in the study area for at least 6 months were included. Mothers who were critically ill and unable to respond during the data collection period were excluded. The outcome variable in this study was PLF practice. PLF is the administration of fluid or food other than breast milk to the newborns except drugs, vitamins, minerals and vaccines before breastfeeding initiation, usually on the first 3 days of neonatal life after delivery.3 In this study, PLF was measured dichotomously as ‘yes’ (practised PLF) and ‘no’ (did not practise PLF). For the explanatory variable, we have collected data on sociodemographic variables and obstetric healthcare service utilisation-related variables. The conceptual framework was used to describe the determinant factors of PLF practice. Several literatures show sociodemographic characteristics affect PLF and obstetric healthcare service utilisation. On the other hand, infant-related factors affect PLF practice (figure 1).16 21 Conceptual frame work of predictors of prelacteal feeding practice. In this study, maternal age was assessed by asking the age of mothers and categorised into: ‘<21’, ‘22–29’, ‘30–38’ and ‘39+’. Maternal education status was categorised into: ‘unable to read and write’, ‘can read and write without formal education’, ‘primary (grade 1–8)’, ‘secondary (grade 9–12)’ and ‘college/university and above’. Father occupation was assessed by categorising into ‘unemployed’ and ‘employed’. Regarding obstetric health service utilisation-related variables, breastfeeding initiation; it was categorised into ‘within 1 hour’ and ‘after 1 hour’, ANC visit was categorised into ‘yes’ and ‘no’, the number of ANC visit was categorised into ‘one time’, ‘two times’, ‘three times’ and ‘greater than or equal to four times’. Concerning counselling on breast feeding during ANC visit, it was categorised into ‘yes’ and ‘no’. Parity was categorised into ‘primiparous’ and ‘multiparous’, place of delivery was categorised into giving birth at ‘health institution’ and ‘home’.21 Mode of delivery was categorised into giving birth via ‘vaginal delivery’ and ‘caesarean section delivery’, participated in an ANC care group was categorised into ‘yes’ and ‘no’; PNC visit was categorised into ‘yes’ and ‘no’; history of neonatal illness was categorised into ‘yes’ and ‘no’. Delayed initiation of breast feeding is defined as the initiation of breast feeding after 1 hour of delivery.3 Early initiation of breast feeding is defined as the initiation of breast feeding within 1 hour following delivery.3 A single proportion formula was used to estimate the sample size by considering the following assumption (p=14.2%)22 the proportion of the estimated level of PLF practice, and considering the assumption of 95% confidence level (Zα/2=1.96), 4% margin of error (d=0.04), and the design effect of (d=2), and adding the non-response rate of 10%, the final sample size was 621 mothers. The data collectors go to the study participants for a consequent three times and those who were absent are considered as non-respondent. A multistage stratified sampling technique was used to select the study participants at the community level. First, in the primary stage, Bure district has got 24 kebeles (the smallest administrative unit in Ethiopia having a total of 3000–5000 residents) (22 rural and 2 urban). In this study, the kebeles were considered clusters and they were stratified into urban and rural clusters. From 24 kebeles, 7 rural and 1 urban were selected using a simple random sampling technique. Second, in the secondary stage, the census was conducted to identify those mothers having children less than 6 months of age in each eight kebele using the folder of community health extension workers registration book. A total of 1269 mothers having children less than 6 months of age were identified in the selected kebeles. Later on, the sample was allocated proportionally based on the number of mothers who had children less than 6 months in each selected kebele. The sampling interval was determined by dividing the total number of mothers who had children less than 6 months in each kebele into the final sample size (1269/621) which gives a sampling interval of (k=2). Finally, mothers who had children less than 6 months were selected using systematic random sampling until the total sample size was reached. The first case was selected using the lottery method at each kebele. In the case of twin newborns lottery method was used to select the study participants (figure 2). Schematic presentation of the sampling procedure. The questionnaires were adopted after reviewing different relevant literature.16 18 19 Data were collected by face-to-face interviews using structured and close-ended questionnaires. To ensure quality, initially, the questionnaire was pretested 5% (31) on non-selected kebele. A total of 12 BSc midwife/nurses (8 diploma data collectors and 4 BSc supervisors) were recruited for data collection process. Two-day training was given to both data collectors and supervisors. Data were entered into Epi Data V.4.2.0 and then exported to SPSS V.23 software package for further analysis. Descriptive analysis results were presented in the form of tables, figures and texts using frequencies and summary statistics including SD and percentage. In this study, the outcome variable was PLF practice among mothers having children less than 6 months of age. PLF practice was coded as ‘1’, while not giving prelacteal foods was coded as ‘0’ for regression analysis. Model fitness was checked using a Hosmer-Lemeshow goodness-of-fitness test (0.93). Bivariate logistic regression analysis was used to determine the association of each independent variable with the outcome variable by using binary logistic regression. The adjusted OR along with 95% CI was estimated to identify predictors for PLF practice by using multivariate analysis in binary logistic regression. In this study, a p<0.05 was considered to declare a result as a statistically significant association. Neither patients nor public (all the people in the community) were involved in the development of this research.