2.1. Study Area, Period and Design
This study was conducted in the Gamo Gofa Zone and Konso Woreda which is found in Zegen area People Zone. Both are found in Southern Nation, Nationalities and Peoples’ Regional (SNNPR) state of Ethiopia. Gamo Gofa zone is organized by 15 Woredas and two town administrations (Arba Minch and Sawula town). Arba Minch (the capital of Gamo Gofa Zone) is located at a distance of 505 kms from Addis Ababa. Gamo Gofa Zone is characterized by mountains that reach 4200 meters in height and it make an area relatively difficult to reach, where the infrastructures (roads, communications) are relatively low. Arba Minch (the capital) is located at a distance of 505 kms from Addis Ababa. Based on the 2007 census population projection, the 2017 population for the Gamo Gofa zone was 2,043,668 (male: 1,013,533 and female: 1,030,135) [11]
Konso peoples are a Cushitic speaking group who live in the administrative territory of Southern Nations, Nationalities and Peoples’ region. It is one of the 5 woredas in Segen Area Peoples’ Zone. A Woreda has 43 kebeles of 41 are rural and 2 are towns. Karat is the town of the Woreda situated about 595Kms South of Addis Ababa. The Segen River in the South, the Woito River in the West, Ale woreda in the North West, Derashe Woreda in the North, Burji Woreda in the South East and Borena in the East are borders for Konso [39]. The population of Konso Woreda is estimated to be 275,535 out of which 132,613 (48.13%) are males and 142,922 (51.87%) are female with annual growth rate of 2.9%. The report also affirmed that 95% of the Woreda population lives in the rural while the remaining 5% of the population dwells in urban areas. The major religions in Konso are Protestant, Orthodox and traditional African religion [12].
The study was conducted using community based cross sectional study design and the data was collected from primary sources on October 2018.
2.2. Population, sampling and sampling procedure
All head of households with under five children (their spouse or their guardian) in the study area were the source population. The household with under five children (their spouse or their guardians) in the in randomly selected Kebeles (small administrative unit) were the study population and individuals (spouses or guardians) who randomly selected and participated in study were sampled population. Spouses or guardians who were severely ill and unable to communicate during data collection were excluded from the study.
The sample size was determined by using Open Epi-Statcalc statistical software. Residence of the study participant was used as the most significant determinant of birth registration status from the study done by Mwango B. Chomba at Copper belt province of Zambia and the following assumption were considered [13]; the proportion of participants who have registered their child from urban residence (exposed group) are16.36% and proportion of participants who have registered their child from rural residence (non-exposed group) was7.89%, Confidence level of 95%, 80% power with a minimum detectable alternative of ± 5%, OR of 0.42, ratio of one to one among unexposed to exposed was used. Accordingly, the calculated sample size was 474 participants. Assuming a non-response rate of 5%, a total minimum sample size needed for this study was 474* 0.05+ 474 = 498 households. Therefore, the minimum sample size required for the study was 498 households.
There were 44 kebeles in 7 Woredas (Arba Minch Town Administration, Kamba, Kucha, Dita, Chencha, Arba Minch Zuria and Konso Woredas) which are considered for this study. According to World Health Organization assessment tool, 14 kebeles (30% of the kebeles) were randomly selected by using lottery method and then the sample was proportional allocated for each selected kebele. To select these sampled households, the lists of the households were obtained from the registration book of family folder in the kebele. Then, the numbers of proportionally allocated households in each kebele were selected by simple random sampling technique by using table of random method from the list. The lists of selected households were reviewed and then by traced their address by the help of the guider from each kebeles. The spouses or the guardians were interviewed.
2.3 Data Collection and analysis
The data was collected using a structured interviewer administered questionnaire which was developed by reviewing different literatures. The tool has three sections; the 1st section was containing socio-demographic variables including age, gender, marital status, religion, ethnicity, residence, education status, occupation status, family size and income; the 2nd section was containing questions assessing the awareness of individual on the availability of birth registration service and birth registration practice; the 2rd section was containing questions assessing practice of birth registration.
Questionnaires were originally written in English, however survey questions were offered in local languages, which were Amharic, Gamogna and Konso language. Hence, participants were allowed to answer in the language they found most comfortable. The tool was tested on 5% of the study participants in Mirab Abbaya district which was not included in this study, before the actual data collection the correction was incorporated.
The data was edited, coded and entered in to EpiData version 3.1 and exported to SPSS version 21.0 statistical software for analysis. Further, data cleaning (editing, recoding, checking for missing values, and outliers) was made after exported to SPSS. First, statistical assumptions of normality, heterogeneity and outliers were assessed both graphically and statistically. Then after, continuous variables were summarized using means and standard deviations while categorical variables were summarized using proportions to describe the characteristics of the study participants. Finally both bivariable and multivariable logistic regression analysis were conducted. A bivariable logistic regression analyses were performed for each independent variable with outcome of interest (practice of birth registration) to identify the associated variables. Upon the completion of the bivariable logistic regression analysis, variables with p-value <0.25 were selected for the multivariable logistic regression analysis and analysis were implemented. The adjusted odd ratios together with their corresponding 95% confidence intervalswere computed and interpreted. All variables with p-value < 0.05 at multivariable logistic regression analysis were considered as determinate of practice of birth registration in the study area.