Study design, setting, and period
A community-based cross-sectional study was done at Dejen woreda from March 20 to April 04/2017. The woreda had latitude and longitude of 10°10′N38°8′E and an elevation between 2421 and 2490 meters above sea level. It is one of the woredas in the Amhara Region of Ethiopia.
Based on the 2007 national census conducted by the Central Statistical Agency of Ethiopia (CSA), the woreda has twenty-three kebeles and 102,359 total population. Seven Kebeles had selected for the study.
Study Participants And Sampling Procedures
All mothers who had an infant during the data collection period in the selected kebeles were the study populations. Whereas; every mother who fulfilled the inclusion criteria and had interviewed was the study unit. Through multi-stage sampling technique and simple random sampling method, a total of 884 infants had selected from seven kebeles as seen in figure (Fig. 1).
The sample size was determined through using both single and double population proportion
Formula with the assumption of; population proportion 50%, Confidence level 95%, and margin
of error 5%, none response rate 15%, and design effect 2
n= (zα/2)2 p (1-P)/d2
Where n = sample size
P = proportion take as 50% (no previous study)
d = margin of error
α = 5%
Zα/2 with 95% confidence level = 1.96
And hence 1.962 0.5(1-0.5)/0.052Equal to 384 samples were required
And hence considering none response rate of 15%, and the sample size of this study was calculated as; 384×15%=58 then 58 + 384 = 442 sample.
Multiplying 442 by the design effect 2 gives 884. For the associated factors, a double population proportion formula using EPI Info version 3.5.1 stat calc had used to calculate the sample size.
s.no
|
Variables
|
% prevalence
|
Confidence level
|
power
|
Prevalence of good practice
|
The ratio among exposed to unexposed
|
Odds ratio
|
Calculated sample size
|
1
|
Mother age group > 33years
|
16.6
|
95%
|
80%
|
40%
|
1:1
|
8.67
|
40
|
2
|
Diploma and above educational status of mothers
|
24.9
|
95%
|
80%
|
60%
|
1:1
|
3.24
|
136
|
3
|
Family size b/n 4–6
|
59.1
|
95%
|
80%
|
39.1
|
1:1
|
3.88
|
214
|
But the value was lower than the above sample size. Then the final sample size was 884 infants.
Data Collection Methods And Instruments
The data had collected through five diplomas and one Bsc nurse supervisor. The Data were collected through the home-to-home visits using a structured interviewer-administered questionnaire. The questionnaires had five parts, socio-demographic characteristics, Knowledge, Attitude, Practice related, and associated factors of mothers towards their infant sunlight exposure. Before the data collection, the questions were translated into the local language (Amharic) from English.
Variables Of The Study
The dependent variable was maternal practice level regarding infant sunlight exposure. Socio-demographic variables (Age of infant, Age of mothers, Mother Educational status, Husband educational status, Mothers Religion, Family size, Mothers Occupation, Father’s occupation), social factor ( media influence, health professional influence, infant grandmother influence), and health-seeking behavior of mothers (ANC follow up, place of delivery), Knowledge status, and Attitude status were the independent variables.
Operational Definition
Infant: a child whose age is less than twelve months old
Good practice: those mothers who responded to the practice-related questionnaires and scored above 8, the median value [19].
Poor practice: those mothers who failed to respond to the practice-related questionnaires scored 8, and less than 8, median values[19].
Good-knowledge: those mothers who responded to the knowledge-related questionnaires and scored above median value [20].
Poor knowledge: those mothers who responded to the knowledge-related questionnaires scored median and less than the median value [20].
Favorable attitude: - those mothers who responded to attitude-related questionnaires and scored above 40 median value [21–23].
Unfavorable attitude: - Those mothers who responded to attitude-related questionnaires scored 40 and below 40 median value. [21–23].
Data Quality Control
Two days of training had given for both the data collectors and supervisors. The questionnaires were prepared initially in the English version and translated to the Amharic version and back to the English version before data collection for consistency. The pre-test had done on 5% of the sample in the other woreda. The supervisor and investigator were checked the data for completeness, missing, and unwanted filling daily. Finally, The data had entered into Epidata version 3.1.
Data Processing And Analysis
During data collection, completeness, missing, and the unwanted filling had assessed daily. Then it was coded, categorized, and entered into EPI data version 3.1 and exported to IBM- SPSS version 20 for analysis. The association between each independent variable with the outcome variable had assessed using bivariate logistic regression. P-value < 0.2 in bivariate analysis covariate retained for multivariate analysis. Logistic regression analysis had performed to identify factors associated with the outcome variable. Finally, a p-value < 0.05 declare as statistical significance. The data had presented using tables, charts, and texts.