Study design and setting
Hawela Tula is a rural catchment that is under the scope of Hawassa City Administration, which has another seven urban sub-cities. Hawela Tula has 1 urban and 11 rural kebeles (the smallest administrative unit in Ethiopia), and an estimated total population of 163,872, from which nearly 48% are females. There are six functional health centers and 17 functional health posts which provide primary healthcare services to the catchment population. Health education on infant feeding is among the services rendered at community level by health extension workers. The common staple food in the study area is false banana, which is called “Wese” in Sidama Afoo, the local language predominantly used in the study site. A cross-sectional study design that deployed quantitative interviews of mothers of children was conducted to estimate the prevalence of prelacteal feeding.
Study participants and eligibility criteria
Biological mothers who had a child or children aged less than six months in Hawela Tula were the source population of the study. Biological mothers who did not live for six consecutive months in the study area, who had communication problems (unable to speak or hear), and who had serious health problems (critically ill mothers) were excluded from the study.
Sample size and sampling
A single population proportion formula was used to estimate the sample size of this study with the assumptions of: 5% precision, 95% confidence, 10% non-response rate, and a design effect of 1.5. A 38.8% estimated prevalence of prelacteal feeding practice was considered from a previous study conducted in Raya Kobo district (rural based study), north eastern Ethiopia [5]. Accordingly, the final calculated sample size was 597. Proportionate allocation of samples to kebeles (based on the number of target respondents in each kebele) was made in a multistage fashion; initially four kebeles were randomly selected from the 11 rural kebeles. Following, initial numbering of all households with target respondents in the four kebeles was conducted to generate a sampling frame for the selection of households using systematic random sampling approach. In scenarios where there were more than one potential respondent in a household, simple random sampling was done to select one. Accordingly, 154, 190, 157 and 149 mothers were allocated to Fincawa, Tula Rural, Tullo, and Gamato Gale kebeles, respectively.
Data collection, processing and analysis
Structured and pretested questionnaire initially prepared in English and later translated to the local language widely used in the study area (Sidamu Afoo) was used to collect data. The questionnaire was adopted from the Ethiopian Nutritional Survey Tools and the Ethiopian Demographic and Health Survey [11, 20]. Data was collected by trained data collectors who are health professionals and experts in the local language. Face-to-face interview technique was conducted at the study participants’ house; revisiting was scheduled for unavailable or busy potential respondents.
To assess the outcome variable (prelacteal feeding), a single question which was narrated as “did you give any food/fluid to your newborn before the initiation of breastfeeding or in the first three days of birth?” was forwarded to mothers; a dichotomized possible response of “Yes” or “No” were offered to respondents. Therefore, newborns whose mothers responded “Yes” to the aforementioned question were regarded as prelacteal fed. Additionally, knowledge and misconception on breastfeeding and health services utilization history of mothers were assessed using closed-ended questions. Five questions (the first milk of the breast is not important to a newborn, giving fluids/liquids prior to initiating breastfeeding is important to the health of a newborn, a breastfed newborn will get hungry if not given additional food within 24 h of birth, a newborn will get thrush if its mouth is not cleaned with water after breastfeeding, and women with small breasts have difficulty producing enough breastmilk) with possible responses of “agree” or “disagree” were used to assess respondents’ misconception on breastfeeding. Respondents who responded “agree” to at least one of the five questions were considered to have misconception.
Data were entered into and analyzed using SPSS for windows version 20 software. Descriptive statistics like percentages and frequency distributions were displayed using tables and figures. Binary logistic regression analysis was performed to assess the association between the various explanatory variables and the outcome variable. Age of child, mother’s religion, mother’s age, mother’s occupation, mother’s level of education, mother’s number of children, antenatal visit, place of delivery, knowledge on the risks of prelacteal feeding, receipt of counseling service on breastfeeding, knowledge on breastfeeding, misconception on breastfeeding, and cultural belief on breastfeeding were included in the bivariate analysis. Additionally, multivariate logistic regression was performed to control for the effects of potential confounders. Variables with p -value < 0.25 in the binary logistic regression analysis were included in the multivariate logistic regression analysis. The Hosmer-Lemeshow goodness-of-fit was used to test for model fitness. Adjusted odds ratio with 95% confidence interval was reported. A p - value of less than 0.05 was used to declare presence of statistical significance.
Data quality assurance
To assure data quality, the following measures were undertaken: data collectors and supervisors were trained for two days on the objectives, relevance of the study, confidentiality of information, respondent’s right, importance of pretest, principles of informed consent, and techniques of interviewing. Furthermore, adjustments were made to the data collection tool after pretesting it in Alamura Kebele (one of the 11 kebeles not selected for the actual study) taking 5% of the sample size. Depending on the feedback generated from the pretest, modifications were made to improve the clarity, understandability, and simplicity of the messages embedded in the questionnaire. Completed questionnaires were checked for completeness and accuracy on a daily basis during the entire data collection.
Operational definition of terms
Prelacteal feeds: Foods, substances, or drinks other than human milk that are given to newborns before breastfeeding initiation or before breast milk comes in, usually on the first few days of life.
Prelacteal feeding: The administration of any foods or liquids other than breast milk to an infant during the first three days after birth.
Index child: The last baby of a mother, less than six months of age, about whom feeding practice is being assessed.
Knowledge on breastfeeding: mothers’ knowledge was categorized into three as follows
-
Good knowledge on breastfeeding: provision of correct answers to all (three) breastfeeding knowledge assessment questions
-
Moderate knowledge on breastfeeding: provision of correct answers to two of breastfeeding knowledge assessment questions
-
Poor knowledge on breastfeeding: provision of correct answers to one or none of breastfeeding knowledge assessment questions