Inequalities in measles immunization coverage in Ethiopia: a cross-sectional analysis of demographic and health surveys 2000–2016

Measles is a highly infectious disease, causing epidemics in high income countries and is a leader in the mortality among children in low income countries [1, 2]. In 2014, measles led to 114, 900 deaths globally, 63% (73,914) of which were in Africa [3, 4]; it is the main cause of child mortality in sub-Saharan Africa [5]. There has been progress in the reduction of measles mortality in the last decade in Africa [6], however, the disease remains an issue in the region [5, 7, 8], reflecting the challenge of sufficient herd immunity levels in areas where financial resources are limited.

Childhood immunizations are low-cost, safe and effective interventions for reducing infant and under-five morbidity and mortality [9]. The majority (90%) of the unimmunized population are in low and middles income countries, many of whom are in South-East Asia and Africa [10]. In 2015, 20.8 million infants did not receive the first dose of the Measles Containing Vaccine (MCV1) through routine immunization; Ethiopia accounted for 3.4% of unimmunized infants [11] and 9% of the global measles mortality [12, 13].

Ethiopia’s childhood immunization coverage has improved through a combination of strategies that include the Reaching Every District (RED) approach, health extension programs and implementation of Enhanced Routine Immunization Activities (ERIAs) [14‐16]. With the increased measles immunization approach, the incidence rate of measles was expected to decrease [17]. Despite improvement of childhood vaccinations, measles outbreaks continue in most of Ethiopia. Even though MCV1 coverage increased from 59% in 2005 to 84% in 2014 [14, 15, 18], the incidence rate of confirmed measles cases per 100,000 increased from 0.6 in 2005 to 11.2 in 2014 [15]. The persistent occurrences of measles outbreaks despite vaccination is thought to be attributed to spatial heterogeneity of measles vaccination [15, 19].

Vaccine coverage can vary due to geographical, socioeconomic or demographic factors; therefore, inequalities of coverage can emerge based on household economic status, mother’s education, and place of residence [20]. The 2016 WHO State of Inequality: Childhood Immunization report demonstrated a 20 percentage point difference between the most and least educated sub-groups, suggesting that high levels of education-related absolute inequality continue despite gains in national coverage in Ethiopia [20].

National immunization coverage rates tend to mask disparities among sub-populations within a country. To understand who the unimmunized children are, it is important to not only investigate the national averages, but also look at the distribution of immunization within a country’s sub-groups [21]. Inequalities by household wealth, mother’s education and urban or rural residence are common in childhood immunization [21].

Continued immunization coverage without focus on inequalities will not address gaps in coverage. It is important to have an equity-focused approach that concentrates on barriers to supply and demand of immunization. This reinforces that the hardest to reach populations are at the center of the expanded coverage strategies while also sustaining the gains made to date [22, 23].

Not only must a country achieve high national coverage, it must also reach the children who remain unimmunized [21]. In addition to the moral imperative of equitable coverage, there are economic and cost-effective advantages to addressing inequalities in immunization coverage [21, 24].

Assessing the state of inequality in immunization identifies gaps and informs planning strategies regarding increased coverage in unvaccinated or under-vaccinated population subgroups [20]. A few studies in Ethiopia focused on socio-economic inequalities in immunization using different dimensions of inequality and summary measures [25, 26]. This study expanded on previous studies and assessed socio-economic inequality using five dimensions of inequality; economic status, educational status, place of residence, sex and sub national regions. Furthermore, inequality was assessed using simple, complex, relative and absolute measures. Finally, it filled the gap of evidence in trends of inequality. The main aim of the study was to assess prevalence and trends of inequality in childhood measles immunization in Ethiopia from 2000 to 2016.

Using the Ethiopia Demographic Health and Survey data, the magnitude and trends of socio-economic, sex related and location-based inequality in childhood measles immunization coverage were investigated between 2000 and 2016. A general over-time increase was evident for the national measles immunization coverage during the first three surveys, which leveled off by the last survey in 2016. Overall national levels of measles immunization showed a median increase of 1.73% per year (Table 1). Interestingly, there was a larger increase in the poorest quintile (1.56 per year) as compared to the richest quintile (1.38 per year).

In Ethiopia, a pronounced level of economic inequality was observed between all the rungs of economic status subgroups. There was a staggering 31 percentage points difference between the poorest and richest quintiles in 2016. The 2016 Population Attributable Factor (PAF) was 36.7%; this indicates that if the coverage of measles immunization in the four quintiles was the same as the richest wealth quintile, the national measles immunization coverage would be improved by 36.7% from the current state. The simple measures in that same year indicated that coverage of measles immunization in the richest subgroup was 1.7 times higher than the poorest subgroup. Household income can play a substantial role in access to care; numerous unintended expenses related to immunizations (i.e. transportation to clinics) may be tolerable in the richest families but not possible for the poorest ones [41].

This study showed absolute and relative educational status inequality in measles immunization in Ethiopia from 2000 to 2016. Based on the point estimates, in the most recent survey, the simple measures showed a Difference of 29.8 percentage points in coverage between the non-educated and most educated households while the Ratio showed the more educated mothers were 1.6 times more likely to have coverage for their children. The PAF value in the 2016 survey showed that if measles immunization coverage in all subgroups were at the same level as household with mothers who have secondary school education or higher, the overall country’s measles immunization coverage could be improved by a staggering 45%. Similar findings were reported by other studies in Ethiopia [42], Democratic Republic of Congo [43] and Nigeria [44]. Educated mothers are more likely to use health care services (i.e. childhood immunization) because of enhanced access and understanding to relevant healthcare information [45, 46]. Furthermore, educated mothers tend to have better communication skills which facilitate interactions with health workers, leading to a better understanding of immunization schedules and practices [45‐47].

Findings from this study showed no sex inequality in measles immunization coverage as there was no variation between male and female children. However, there was inequality when comparing the pattern of change in measles immunization coverage in rural and urban areas. In 2016, urban regions were 24.5 percentage points higher in measles coverage than rural areas. Similarly, in the earlier survey years, urban regions had more coverage, however over time, rural regions seemed to greatly increase their immunization coverage and urban areas slowly increased but then declined in 2016. Despite this, residence related inequities in coverage remained in Ethiopia, with the rural setting falling behind the urban area in all the studied years. The PAF value in 2016 indicated that without the existence of inequality between urban and rural, coverage of measles immunization in the country could be improved by close to 40%. The pro-urban finding around this vaccine is not surprising as utilization of health services in rural regions is inadequate because of availability, accessibility, quality of services and the characteristics of the users and the communities in which they live. Geographical access is a key factor in utilization of healthcare services as predominantly seen in rural areas with limited delivery of healthcare services [48, 49].

Another main finding from our study is presence of significant absolute and relative regional inequality in Ethiopia in all survey rounds. There was significantly lower measles immunization coverage in some regions as compared to other regions within the country. In 2016, the difference in immunization coverage between Addis Ababa (region with highest coverage) and Afar (region with lowest coverage) was 63 percentage points. To further illustrate the regional inequality, the ratio measure showed Addis Ababa had three times the immunization coverage compared to Afar. If all the other regions had the same measles immunization coverage level as that of Addis Ababa in 2016, the overall national coverage of measles immunization would be improved by 39 percentage points on average, as indicated by PAR. This result is consistent with a previous study in Ethiopia that showed measles immunization varied across different regions with in the country [50]. Culture, religion, economic status, vaccine supply, availability and accessibility of immunization health services could account for regional differences.

Our study has some strengths. This study made use of nationally-representative surveys to explore the nature and degree of inequality in Ethiopia from different perspectives. In the current literature, studies on this topic are rare and do not show significance. From this perspective, we believe the present findings make an important contribution to the knowledge base and provide a solid basis for further research. However, this paper has several limitations. First, the DHS data is cross-sectional and therefore no causality can be inferred from the associations. The authors have no influence over the selection and measurement of the variables when using secondary database analysis Second, we were not able to examine the effect of determinant factors on the observed measles inequality. Future studies might employ a decomposition technique to understand the individual contribution of different factors on measles inequality.

Several key findings from this study emerged. First, the progress in Ethiopia’s national coverage of measles immunizations increased until 2011, after which it plateaued though the absence of UIs around point estimates makes it difficult to comment on the progress. Second, significant inequality in childhood measles immunization was seen in subgroups based on economic status, educational status, place of residence and region in the country. Even though the patterns of inequality were different in the relative and absolute inequality measures used, inequalities were observed, leading to the need for multilevel interventions to reach missed children. Political determination is vital to progress and the reduction of such inequalities in Ethiopia. Where inequalities exist, the national, regional and district levels of government should make a pledge to reduce inequalities in coverage of measles immunization, formulating equity-sensitive strategies and assigning sufficient human and financial resources to implement them based on scope. International and national immunization experts and researchers need to continue investigating and monitoring inequalities related to measles immunization and propose possible solutions to achieve the related sustainable development goals.