Menu
Menu
Menu
Menu
Background: In Ethiopia, the Women Development Group program is a community mobilization initiative aimed at enhancing Universal Health Coverage through supporting the primary healthcare services for mothers and newborns. This study aimed to assess the association between engagement in women’s groups and the utilization of maternal and neonatal health services. Method: A cluster-sampled community-based survey was conducted in Oromia, Amhara, Southern Nations, Nationalities and Peoples, and Tigray regions of Ethiopia from mid-December 2018 to mid-February 2019. Descriptive and logistic regression analyses were performed, considering the cluster character of the sample. Results: A total of 6296 women (13 to 49 years) from 181 clusters were interviewed. Of these, 896 women delivered in the 12 months prior to the survey. Only 79 (9%) of these women including Women Development Group leaders reported contact with Women Development Groups in the last 12 months preceding the survey. Women who had educations and greater economic status had more frequent contact with Women Development Group leaders. Women who had contact with Women Development Groups had better knowledge on pregnancy danger signs. Being a Women Development Group leader or having contact with Women Development Groups in the last 12 months were associated with antenatal care utilization (AOR 2.82, 95% CI (1.23, 6.45)) but not with the use of facility delivery and utilization of postnatal care services. Conclusions: There is a need to improve the organization and management of the Women Development Group program as well as a need to strengthen the Women Development Group leaders’ engagement in group activities to promote the utilization of maternal and neonatal health services.
The Optimizing Health Extension Program intervention was implemented in 26 districts. The evaluation was conducted in these 26 intervention districts and 26 comparison districts that were selected to be similar to the intervention districts based on demographic and health service characteristics [17]. A baseline study was conducted before the intervention (mid-December 2016 to mid-February 2017) and an endline study was conducted two years after the baseline survey (mid-December 2018 to mid-February 2019). Our secondary analysis was unrelated to the original evaluation study of the OHEP intervention. Therefore, we used data from the endline survey and combined the data from the original intervention and comparison areas [17]. The study design was a cluster-sampled cross-sectional survey and was carried out in Oromia, Amhara, Southern Nations, Nationalities and Peoples, and Tigray regions of Ethiopia. Figure 1 shows a map of the study districts. Map of Ethiopia showing all regions (left) and the intervention and comparison districts within the four study regions (right), 2018/2019 [17]. A two-staged stratified cluster sampling method was used to select the study subjects. The 2007 Ethiopian Housing and Population Census data were used to identify and list enumeration areas in the 52 study districts. The plan was to select 200 enumeration areas proportional to the size of the districts. However, by the endline survey, nineteen clusters were excluded due to civil unrest, and 181 enumeration areas were finally considered for the study. In each enumeration area, 66 households were selected and all women who were eligible for interview (13 to 49 years old) and available on the day of the survey were invited to participate in the study. A total of 120 data collectors were recruited by the Ethiopian Public Health Institute and had, as a minimum, completed their first degree in health sciences. The data collectors were trained for two weeks, including in study procedures, questionnaires, interview and data collection techniques, quality assurance procedures, and study ethics. A field manual was prepared and used in the training. Regional coordinators and team leaders were assigned to supervise the interviews. The interviews were conducted face-to-face. On average, the interviews took thirty minutes. One interviewer interviewed on average six households per day. A questionnaire was prepared in English and translated into the local languages Amharic, Oromiffa, and Tigrigna, and after that back-translated into English. The questionnaire was based on earlier validated survey tools, including the Ethiopia Demographic and Health Survey tools which were used for the household survey and other tested tools, which were used for the evaluation of the integrated Community Case Management of Childhood Illnesses and Community Based Newborn Care programs [21,29]. The household survey comprised of two sections. The first was a household overview administered to the head of the household to collect information on age and sex of all current residents, the identification of the primary caregiver to any child under five years of age, characteristics of the house and its assets, access to healthcare, and location of each household based on geographical positioning system (GPS) assessment. The second was the women of reproductive age (13 to 49 years) section capturing information on the healthcare available to them, and their recent contact with WDG leaders and health providers. The decision to include a 1-year recall period was primarily related to the original OHEP evaluation to reduce recall bias for details. In addition, the majority of the respondents were illiterate; we decided to consider women who had a last birth history for children born in the past 12 months preceding the survey to minimize recall bias. Furthermore, the interview included care seeking and utilization of health services during pregnancy, delivery, and postpartum for the first 28 days after delivery periods [17,30]. The analysis considered three maternal and neonatal health dependent variables: antenatal care service utilization, use of facility delivery, and utilization of postnatal care services. Antenatal or prenatal care service utilization was the indicator of access and use of healthcare services during pregnancy. At least one antenatal care visit defined the proportion of pregnant women who received at least one example of antenatal care for a pregnancy in the 12 months preceding the survey. Use of facility delivery captured the proportion of live births (13 to 49 year old women) at a health facility that had been assisted by health professionals within the last 12 months preceding the survey. The utilization of postnatal care indicator defined the proportion of women who received postnatal care in the first month after delivery for a pregnancy that occurred in the 12 months preceding the survey. Utilization of postnatal care during the first month after delivery assumed that the use was for both mothers and newborns. This analysis included the WDG contact in the 12 months prior to the survey as the main contact outcome variable. Assessment of women’s contact with WDG leaders was categorized as follows: women who had WDG contact but who were not WDG leaders; women who were WDG leaders; and women who neither were WDG leaders nor had WDG contact. Additional information was also collected on women’s socio-economic characteristics, including their years of education, religion, marital status, birth order, and household wealth. Education was categorized into no education and educated. Educated refers to those who had at least one year of formal education or more. Religions included Christian Orthodox, Muslims, and Protestant. Birth order refers to the order a child is born in within their family. Birth order was classified into one child, two or three children, and four or more children. The household wealth index was based on durable assets, household building materials, utilities, and animals owned. The continuous variable produced by the first principal component was divided into five equally (20%) sized groups (quintiles) of households from quintile 1 (poorest) to quintile 5 (wealthier). Data were collected on personal tablet computers, and the Census and Survey Processing System (CSPro) was used. The collected data were regularly sent to the central server at the Ethiopian Public Health Institute. The server was password protected, and access to the data was limited to the study team. Data were cleaned and prepared for analysis. The quality of the information collected was ensured by using validated and pretested forms, a system of field supervision, and careful data quality control and management that included daily checks on completeness and consistency. A detailed survey manual with extensive standard operating procedures was prepared and used in training, piloting, and fieldwork. A descriptive analysis, including frequencies and percentages, was performed. Background factors, including contact with a WDG leader or being a WDG leader in the last 12 months, were cross-tabulated with antenatal care service utilization, use of facility delivery, and utilization of postnatal care. Fisher’s exact test was performed to determine the significant association for variables with small values. A multivariate analysis was thereafter conducted to assess whether there was an association between having a WDG contact or being a WDG leader and antenatal care service utilization, use of facility delivery, and utilization of postnatal care services. We controlled for factors that were significantly associated in the bivariate analysis. The “svyset” and “svy” prefixes were used to adjust for the cluster sample design to account for a clustering effect. The STATA statistical software version 14.0 (StataCorp LLC., College Station, TE, USA) was used for all analyses. Ethical review: Ethical approval was obtained from the Ethiopian Public Health Institute (protocol number SERO-012-8-2016), London School of Hygiene and Tropical Medicine (protocol number 11,235), and the Institutional Review Board (IRB) of Mekelle University, College of Health Sciences (protocol number 1433/2018). Support letters were also obtained from the Regional Health Bureaus in Amhara, Oromia, Southern Nations and Nationalities Peoples, and Tigray. Informed consent was obtained from all study participants.
N/A
