Menu
Menu
Menu
Menu
As part of the 2030 maternal and child health targets, Ethiopia strives for universal and equitable use of health services. We aimed to examine the association between household wealth, maternal education, and the interplay between these in utilization of maternal and child health services. Data emanating from the evaluation of the Optimizing of Health Extension Program intervention. Women in the reproductive age of 15 to 49 years and children aged 12–23 months were included in the study. We used logistic regression with marginal effects to examine the association between household wealth, women’s educational level, four or more antenatal care visits, skilled assistance at delivery, and full immunization of children. Further, we analyzed the interactions between household wealth and education on these outcomes. Household wealth was positively associated with skilled assistance at delivery and full child immunization. Women’s education had a positive association only with skilled assistance at delivery. Educated women had skilled attendance at delivery, especially in the better-off households. Our results show the importance of poverty alleviation and girls’ education for universal health coverage.
A cross-sectional study was conducted in 46 districts of four Ethiopian regions, i.e., Amhara, Oromia, Southern Nations, Nationalities and Peoples Region, and Tigray, from December 2018 to February 2019. These regions are the most populous in the country, where the Ethiopian Government initiated the Optimizing the Health Extension Program interventions. The survey was conducted jointly by the London School of Hygiene and Tropical Medicine, the Ethiopian Public Health Institute, and four Ethiopian universities; the University of Gondar, Jimma, Mekelle, and Hawassa Universities. The country has a three-tiered health system with primary healthcare units and secondary and tertiary levels of care. The population size of the study districts was on average 130,000 people, with 23% being women of the reproductive age and 20% children below the age of five years. One-third of the districts had a hospital. There were, on average, five health centers per district and five health posts under each health center [19]. This study used data from the evaluation of the Optimizing the Health Extension Program intervention that aimed at improving services utilization. A two-stage stratified cluster sampling technique was used to select study subjects. First, 194 enumeration areas, the primary sampling unit, were obtained based on the 2007 Ethiopian Housing and Population Census using probability proportional to the size of the districts. Second, all households within the clusters were listed. Sixty households per cluster were selected using systematic random sampling. All women of reproductive age (15–49 years old) and children under the age of five years, who lived in the selected households, were included in the study. A standard sample size formula was used to calculate the sample size. The sample size was estimated to be 6000 households per group (12,000 in total). The sample size determination was detailed elsewhere [20]. The questionnaire was developed based on existing large-scale survey tools in English, translated into local language and back translated and pretested. Data collectors were trained for 10 days including field training before the start of data collection. Information about antenatal care attendance and delivery by skilled birth assistance was collected from all reproductive-age women who had a live birth during one year preceding the survey. Immunization information was collected by combining data recorded on children’s vaccination cards and responses from the parents if the vaccination card was missing. The questionnaire also included information on sociodemographic data and household assets. Data were collected on personal digital assistants (Companion Touch 8), and tablets (Toshiba and Hewlett Packard) programmed with CSPro 7.1. through face-to-face interviews. Data collectors sent encrypted data from the field to the password-protected server at the Ethiopian Public Health Institute. Data managers conducted quality checks and provided feedback to field teams. Data were cleaned and checked for consistency and completeness. The analysis included three maternal and child health indicators: four or more antenatal care visits, skilled birth assistance, and full immunization of children aged 12–23 months. Four or more antenatal care visits were defined as the percentage of women of reproductive age with a live birth within the last 12 months preceding the survey who attended four or more antenatal care visits during pregnancy. Skilled birth assistance was represented in the percentage of women aged 13–49 years with a live birth within the last 12 months preceding the survey who were attended at delivery by skilled health personnel. Full immunization was defined as the percentage of children aged 12–23 months who had received one dose of BCG vaccine, three doses of polio vaccine, three doses of pentavalent vaccine, and one dose of measles vaccine [21]. All these outcomes were coded as 1 when the subjects had received the service or 0 when the subjects had not received the service. Covariates. The covariates included in this study were household wealth, which was created by dividing the household wealth index into three equal tertiles (Tertile 1, Tertile 2, and Tertile 3) to classify households as poor, middle, and better-off. The wealth tertile was created based on ownership of durable assets, access to utilities and infrastructure, and housing characteristics. The construction of the wealth tertile was done using principal component analysis as detailed in a previous publication [19]. Maternal education was categorized into two levels: no education (not attended formal education) and educated (primary or above). Other covariates included were maternal age in years (15–24, 25–34, and 35–49), birth order (1, 2–3, and 4 and above births), region (Amhara, Oromia, SNNPR, and Tigray), religion (Orthodox Christian, Muslim, Protestant, and others), and sex of the child. Wealth-education was also created by combining household wealth and maternal education and was categorized into six levels: tertile 1*no education, tertile 1*educated, tertile 2*no education, tertile 2*educated, tertile 3*no education, and tertile 3*educated. Descriptive analyses included frequency distributions of the determinants and covariates and outcomes of the service utilization. The utilization of services was cross-tabulated with socioeconomic and other background factors. Logistic regression was used to examine the associations between household wealth, maternal education, maternal age, birth order, region and religion and outcomes of the service utilization, and interactions between household wealth and maternal education. The results from the logistic regression analyses were presented as average marginal effects with 95% confidence intervals for the main effects and 90% confidence intervals for the interaction terms. The average marginal effects were used to estimate the discrete change for the factor’s levels from the reference. The Chi-square test was used to measure the significance of the change. Potential multicollinearity between the covariates used in the multivariate regression model was assessed using variance inflation factors. The Delta method was used for the standard errors to estimate the variation. The marginal effects were estimated using the margins command in Stata 14.1 for windows (GSW) (StataCorp LLC, College Station, TE, USA), which considered the interaction terms included in the model. Marginsplot command in Stata [22] was used to graphically display the results. During the analysis, all the commands were preceded with svy to account for clustering. Ethical review: Ethical approval was obtained from the Ethiopian Public Health Institute (SERO-012-8-2016; Version 001), London School of Hygiene and Tropical Medicine (LSHTM Ethic Ref: 11235), and the IRB office of College of Health Sciences of Mekelle University in Ethiopia (ERC 1434/2018). Written consent and assent were also obtained from the participants.
N/A
