Study area & period
The study was conducted in Ambo district from March 1 to April 7/2019. Ambo is one of the districts of west Shoa zone in Oromia Region of Ethiopia. It is located at 114 km west to Finfinee/Addis Ababa on the main road to Wollega. Bordered by T/kutaye to west, by Cobi and Ilfeta to north, by Dendi to east, by Wonchi district of Southwest Shewa Zone to south. Based on the information from Ambo district administration health office, the district has 32 rural and 1 urban totally of 33 “Gandas” (small administration units). The district has six health centers, 31 health posts, 66 health extension workers and 82 all category health workers. According to the projected population by the health office for 2018/19, the total population was estimated to be 138,011 among which 50.9% (70248) were women and 3.22% (4,444) were mothers of children aged less than 12 months/surviving infants (office 2018/19).
Study design
A quantitative community based cross-sectional study design supplemented by qualitative study was used.
Population
Source population
Quantitative: All mothers of infants in Ambo district.
Qualitative: All husbands, mother-in-laws, & traditional birth attendants for focus group discussions; and Abbaa-Gadaas/elders in Ambo district, health extension worker and maternal & child health coordinator of Ambo district for key informant interview.
Study population
Quantitative: Selected mothers of infants living in the selected Gandas who fulfill the inclusion criteria.
Qualitative: Selected husbands of mothers of infants in the selected Gandas, mother-in-laws, and traditional birth attendants in Ambo district for focus group discussions; and Abba Gadaas/elders, health extension worker and maternal & child health coordinator of the district for key informant interview.
Inclusion and exclusion criteria
Inclusion criteria:
a) All biological mothers who live in the ganda/kebele at least for the last six months & had under one-year old Child.
Exclusion criteria
Mothers who were acutely sick and unable to communicate during data collection.
Sample size calculation
The sample size (n) required for the study was calculated using the formula to estimate a single population proportion by considering the following assumptions. Zα/2 = critical value for normal distribution at 95% confidence level which equals to 1.96 (Z value at alpha = 0.05).
According to the study conducted in Debremarkos town of Amhara region in 2015, about 76.72%, 78.84% and 77.71% of mothers had good knowledge, favorable attitude and good practice of colostrum feeding respectively(27).
a) Hence, according to the study; P = 76.72% for knowledge:
d (w) = margin of error of 0.05 with 95% confidence level.
n = (Z α/2)2 *P (1-P)/ d2 = (1.96)2*.77(1-.77)/0.052
n = 272 Mothers with below one years old children (including non-response rate of 10%)
= 272 + (272*10%) = 299 Mothers of infants.
b) According to study conducted in Debremarkos town in Amhara region (2015),
P= (78.84%) for attitude.
d (w) = margin of error of 0.05 with 95% confidence level.
n = (Z α/2)2 *P (1-P)/ d2= (1.96)2*.79(1-.79)/0.052 = 254 (including 10% non-respondents)
= 282 Mothers of infants.
C) According to study conducted in Debremarkos town in Amhara region (2015):
P= (77.71%) for practice.
d (w) = margin of error of 0.05 with 95% confidence level.
n = (Z α/2)2 *P (1-P)/ d2
= (1.96)2*.78(1-.78)/0.052
= 264 (including 10% non-respondents)
= 291 Mothers of infants.
Here, from the study conducted in Debremarkos town of Amhara region, the result for knowledge of the respondents bequeathed representative sample size for this study. Since the sampling procedure during this study was affected by design effect, the sample size becomes, (299*1.5 = 449).
Concerning focus group discussion, till saturation of ideas, two FGDs were conducted. The first was conducted in A/Qora ganda at Awaro HP and the second was conducted at Meti HC including mother in laws, mothers of infants, husbands of a mothers of infants, & traditional birth attendants. Three participants for key informant interview were purposively selected and interviewed. The participants were the district MNCH-coordinator, HEW, and one religious leader in Ambo district.
Sampling procedure
Ambo district administration has 33 Gandas (the smallest administrative unit in Oromiya). Thirty percent (11) Gandas were selected as study Gandas by lottery method. These were: Y/Cebo, Bilo, G/Qora, G/wadessa, Golja, U/Qorke, Mexi, N/Warqa, Amaro, Tule, and Wadessa. Study population were assigned proportionally to the selected gandas/sub districts. From family folder/there is a list of mothers of infants. The list/sampling frame was already there, because these were targeted for immunization. Then, to select the study population, all source population of respective gandas were coded. Mothers of infants were selected by simple random sampling, lottery method. Since the households’ zone/Gare were clearly indicated on the cover of the family folder, the selected mothers were easily traced and interviewed. Where it was difficult, local guiders were hired in two kebeles for some days to reach those selected households with mothers of infants.
Study variables
Dependent variables:
-
Knowledge towards colostrum.
-
Attitude towards colostrum.
-
Practice of colostrum feeding within an hour to 2–3 days.
Independent variables:
-
Socio-demographic characteristic of the mother
-
Age
-
Sex of neonate
-
Marital status,
-
Educational status
-
Occupation of mother
-
Area of residence
-
Housing ownership
-
Possession of cow milk at birth
-
Average monthly income
-
Providers/health service-related variables
-
Whether the birth/pregnancy of current neonate wanted
-
Attendance of ANC
-
Number of ANC visit
-
Reception/treatment by health service providers
-
Maternal counseling by ANC provider on benefit of colostrum
-
Information regarding colostrum
-
Attendance of pregnant mothers’ forum
-
Place of delivery
-
Attendance of PNC
-
Obstetric and medical variables of the baby
-
Mode of delivery
-
Health status of the newborn at birth
-
Health status of the mother during pregnancy/at birth.
Operational Definition
Knowledge: In this study refers to awareness of mother of infant (s) about colostrum breast milk. It was evaluated by the mother's answer to the questions (11).
Good knowledge: Those mothers who answer ≥ 60% questions of knowledge related questions (11).
Poor knowledge: Those mothers who answer < 60% questions of knowledge related questions (11).
Attitude: The way that mothers of infants think towards colostrum feeding to their infants.
Positive attitude: Those mothers who answer ≥ 60% of attitude related questions.
Negative attitude: Those mothers who answer < 60% of attitude related questions.
Practice: The behavior of mothers of infants on colostrum feeding to their current infants within one hour to two-three days(9, 21, 31–33).
Data collection instrument and methods
Structured questionnaire for face to face administration and semi structured guiding questionnaire for quantitative & qualitative data collection were adapted and modified from previous similar studies respectively (9, 21, 31–33). For qualitative, in addition to semi-structured questionnaire, smartphone was used. One teacher from each selected Gandas as data collector and 11 supervisors who have BSc from respective PHCU were provided one day training/orientation. Qualitative data were also collected through FGD, and key informant interview. After the data collection for qualitative, the data were transcribed in to English to answer the semi-structured questions. The investigator and two supervisors facilitated the sessions of FGD but the investigator alone conducted the three in-depth interviews.
Data quality control
The tools for quantitative data collection were prepared in English language and translated in to Afaan Oromo and then re-translated back to English by language expert/professional for its consistency. Data collectors and supervisors were trained for one day on how to interview and record. Pre-test was done in A/Qora at 5% (5%*449 = 20) of actual respondents. After the pre- test, no unclear questions were founded to amend. But the data from the pre-test was not included in the analysis of actual study. Regularly, monitoring and supervision of the overall activity was conducted by the supervisors & investigator. To maintain the quality of data while entering data, Epidata version 3.1 was used.
Data processing and analysis
After checking the completeness and consistency of the quantitative data on the printed form of the questionnaire; it was cleaned, coded and entered into Epidata version 3.1. The data analysis was performed by using SPSS version 21.0 statistical package for analysis. Descriptive statistics was used to show the knowledge, attitude and practices of the mothers on colostrum feeding in the district. Binary and multivariable logistic regression analyses were carried out to identify the factors associated with colostrum feeding. The variables with p-value < 0.2 in binary logistic regression were included in multivariable regression analysis model. Variables at p-value < 0.05 in the final multivariable model were concluded as factors associated with the practice of colostrum feeding.
Qualitative data were transcribed into English text by the investigator. Thematic framework analysis approach was used to analyze the qualitative data. Finally, the result was presented in triangulation with the quantitative findings.