Factors associated with the access and continuum of vaccination services among children aged 12–23 months in the emerging regions of Ethiopia: evidence from the 2016 Ethiopian demography and health survey

The 2016 Ethiopian Demographic and Health Survey (EDHS) was conducted in the nine national regional states and the two city administrations. The regions classified as emerging, Afar, Benishangul-Gumz, Gambella and Somali, are characterized by scattered pastoralists and semi-pastoralist societies suffering from extreme poverty. Absence of clear and detailed regulations, basic infrastructures and services are also their common chrematistics [14, 15]. On the other hand, developed regions such as Amhara, Oromia, Southern Nations, Nationalities and Peoples (SNNP), Tigray and the city administrations, such as Addis Ababa, Dire Dawa and Harari regions are relatively more densely populated [15]. Ethiopia is one of the Sub-Saharan countries found in the horn of Africa with a total population of 73.5 million. The total population of the four emerging regions was 6926, 933, with the largest in Somali (4,445,219) and the least in Gambella (307,096). Similarly, the total number of children 0–4 years was 449,699 in Somali and 42,044 in Gambella [16].

The sampling frame used for the 2016 EDHS was the 2007 Population and Housing Census (PHC) of the Central Statistical Agency report of Ethiopia [16]. The sample for the 2016 EDHS was designed to provide estimates of the key indicators of the country as a whole, separately for urban and rural areas, and for each of the nine regions and the two city administrations. The sample was stratified and selected in two stages and each region was stratified into urban and rural areas. Samples of the Enumeration Areas (EAs) were selected independently in each stratum of two stages. Implicit stratification and proportional allocation were used at each lower administrative level.

In the first stage, EAs in urban and rural areas were selected with probability proportional to the EA size (based on the 2007 PHC) and with independent selection in each sampling stratum. A household listing operation was carried out in all of the selected EAs in 2015. The resulting lists of households served as sampling frames for the selection of households in the second stage. Segmentation was done for some of the selected EAs with large households, and only one segment was selected for the survey with a probability proportional to size. Household listing was conducted only in the selected segment, that is, the 2016 EDHS cluster was either an EA or a segment of an EA. In the second stage, the selection of the households per cluster was done using systematic sampling technique. In this study, the 2016 Ethiopian demographic and health survey childhood datasets of the four emerging regions, namely Afar, Benishangul-Gumz, Gambella and Somali were used for analysis.

All women aged 15–49 years and permanently lived in the area and slept in the selected households the night before the surveys were eligible [10]. Children 12–23 months are the source population and the study included 642 mothers and their children aged 12–23 months and data on both were extracted from the 2016 EDHS datasets. Potential independent variables such as socio-demographic, economic, fertility history and health service utilization were also extracted and further recoding of the selected variables was done to match and compare with other similar studies.

Access and continuum of childhood vaccination were the dependent variables of the study. Socio-demographic characteristics (age, residence, religion, and marital status), and obstetric history of the women, like places of delivery, birth order, antenatal care, postnatal check-ups in 2 months after birth, number of live children, sex of children, and marital status were the independent variables. Vaccination refers to the administration of antigenic material (a vaccine) to stimulate an individual’s immune system to develop adaptive immunity to a pathogen [17] and its coverage is defined as the proportion of a given population that has been vaccinated in a given time period [18]. A fully vaccinated child was expected to receive one dose of BCG, three doses of pentavalent, Pneumococcal Conjugate (PCV), Oral Polio Vaccines (OPV), two doses of Rotavirus and one dose of measles vaccines [19]. Each vaccine had five response categories, namely “no”, “vaccination date on the card”, “reported by mothers”, “vaccination marked on card” and “do not know”. The vaccination status of children was recoded as 0 and 1 for each antigen. “No” responses were recoded as “0” and labeled “not received the vaccine”, while the other responses “vaccination date on card, reported by mothers, vaccination marked on card” were recoded together as “1” and labeled “received the vaccine”. Besides, “do not know responses” were excluded from analysis. As a result, the vaccination status of children was recoded as “0” for “not vaccinated” and “1” for “vaccinated” for each antigen on the basis of the reports of women and information on the child vaccination card. Accordingly, access and the continuum of childhood vaccination were recoded as “0” for “no” and “1” for “yes” for each child. Access is the provision of services in shorter waiting times, more flexibly, electronically, by telephone or alternative methods of communication [20]. As a result, access to childhood vaccination was determined based on pentavalent-1 vaccination status of children. Continuum of care reflects the extent to which a series of discrete health care events are being experienced by people coherently and interconnected over time [20]. Thus, continuum of childhood vaccination was measured by measles vaccination status of children.

The extracted EDHS data included socio-demographic characteristics of the women, obstetric history and service utilization child-specific information for all births in the past 5 years of women in the reproductive age group. The 2016 EDHS collected information on childhood vaccination status from vaccination cards and women’s verbal reports. The interviewer copied the vaccination dates directly into the questionnaire if the cards were available. However, the interviewer asked respondents to recall the vaccines given to their children when there were no vaccination cards. The cleaned and recoded data were analyzed using STATA version 14. Descriptive statistics such as means, medians, SDs, frequencies and proportions of variables were presented using graphs, texts, and tables. Bivariable and multivariable logistic regression analyses were conducted to identify factors associated with access, utilization, and continuity of vaccination. Variables with p-values < 0.2 [21, 22] during the bivariable analyses were fitted into the multivariable logistic regression analyses. However, p-value of 0.2 did not mean that there was a 20% chance that the null hypothesis was correct [22, 23], rather variables with p-value of < 0.2 during the bivariable analyses might have a chance to be significantly associated with the outcome variable during the multivariable regression analyses. Adjusted Odds Ratio (AOR) and 95% Confidence Interval (CI) with p-value < 0.05 were used to identify variables associated with the outcome variables.

A total of 642 mothers/caregivers of children 12–23 months of age were included in these analyses. The mean age of the participants was 27.86 (+ 6.31SD) years, and the median age was 28 years with an Inter Quartile Range (IQR) of 8 years. Furthermore, more than half of the participants were in the age range of 25–34 years; over two-thirds (68.2%) were Muslims; more than 80 % (83.2%) lived in rural areas. Only 8.9 and 19.5% of the respondents and their husbands had secondary and above education, respectively. Similarly, 5.1% of the respondents and 13.2% of their husbands had professional jobs; 93.3% were married (Table 1).

More than half (57.5%) of the women received antenatal care services. Accordingly, 44 and 65.6% of the ANC users had four and/or more visits and started the service at the second trimester, respectively. Similarly, only 27.6 and 8.1% of the women had health facility delivery and PNC in 2 months of postpartum period, correspondingly. Nearly 27, 62 and 7% of the participants had ≥6, ≥3 and history of terminated pregnancy, respectively (Table 2).

Of the children, 55.6% were female and60.6% of the caregivers were in the poorest wealth status. The mean and median age of the children were 16.29 (+ 3.29SD) months and 16 months with the IQR of 6 months, respectively. About 45% of the children were in the age range of 12–15 months. More than three-fourths (76.5%) of participants had no mobile phones, and almost 85% did not listen radio programs. Nearly 40 and 90% of children had average birth weights and wanted by their families, respectively (Table 3).

The vaccination status of children 12–23 months of age was analyzed for each antigen in the emerging regions of Ethiopia. Overall, 25.1% (95%CI: 21.80, 28.30) of the children were fully vaccinated in the region. About 57 and 47%of the children received BCG and measles vaccines, accordingly. Seventy-four-point-four and forty-seven-point 2 % of children received OPV 1 and 3, respectively. About 62.0, 41.3, 55.8, 37.5, 52.8, and 45.3% had received Pentavalent 1, 2, PCV 1, 2, Rota 1 and 2, correspondingly. Furthermore, pentavalent and BCG to measles dropout rates were 33.4 and 17.5%, respectively. As a result, access to childhood vaccination was 62% (95%CI: 58.40, 65.70) and the continuum of vaccination 46.9% with (95%CI: 43.30, 50.80) (Fig. 1).

The binary logistic regression model showed that educational status of the mothers, residence, weight of children, ANC follow ups and places of birth were significantly associated with access to childhood vaccination. Mothers who lived in rural areas were 67% less likely (AOR = 0.33; 95%CI: 0.14,0.76) access to their children vaccination compared with that of mothers lived in urban areas. Accordingly, participants who had formal education were 1.99 times (AOR = 1.99; 95%CI: 1.20, 3.31) more likely to get access to vaccination for their children compared with women who had no formal education. On the other hand, children who had average birth weights were1.67 times (AOR = 1.67; 95%CI: 1.03, 2.72) more likely to get access to vaccination compared with children who had more birth weights. The odds of childhood access to vaccination among women who had ANC follow ups were 4.13 times (AOR = 4.13; 95%CI: 2.75,6.19) higher than those of their counterparts. Women who gave their last birth at health facilities were 1.58 times (AOR = 1.58; 95%CI: 1.19, 2.82) more likely to have chances to access child vaccinations compared with women who gave birth at home. Similarly, women with middle (AOR = 2.79; 95%CI: 1.09, 7.14) and rich wealth (AOR = 1.57; 95%CI: 1.19, 3.14) had 2.79 and 1.57-times higher access to childhood vaccination compared with women in poor wealth status (Table 4).

Potential predictor variables associated with the continuum of vaccination services among children 12–23 months of age were identified using the multivariable logistic regression analyses. Children who had less birth weights were 49% less likely (AOR = 0.51; 95%CI: 0.33, 0.81) to continue vaccination services compared with children who had more birth weights. The odds of the continuum of childhood vaccination among women who had ANC follow ups were 3.68 times (AOR = 3.68; 95%CI: 2.48, 5.47) higher than women who had no ANC. Women in rich wealth status were 2.07 times (AOR = 2.07; 95%CI: 1.15, 3.71) more likely to continue childhood vaccination compared with women in poor wealth status (Table 5).