Factors associated with measles vaccination status in children under the age of three years in a post-soviet context: a cross-sectional study using the DHS VII in Armenia

Based on the results from previous studies [1, 2, 4, 23] and the Armenian context [30, 34, 37], a variety of factors have been selected as independent variables. As general factors, we selected residence in urban or rural areas and the wealth index, which was a composition of household assets provided in quintiles. The chosen maternal factors were age in years, the marital status, measuring whether the women were either currently or formerly in union, and the level of education. For the maternal level of education, we used three levels of education: basic (grades 1–9), secondary (grades 10–12) and higher education (secondary special, which provides training for careers that require mid-level qualifications such as nurses or technicians, and university education) [36].

As paternal factors we used the age of the partner in years and the level of education [38]. The partners´ levels of education was divided in three categories: none or primary, secondary, and secondary special. Child related factors included in the study were the age of the child in months, gender, and the birth order, categorized into first-born, second-born and third-or higher born.

We also looked at factors representing health seeking behaviour, such as whether the distance to the health facility was perceived as a big problem or not, the number of antenatal care visits and attendance to a postnatal care visit two months after birth (yes, no).

To elaborate the influence of mass media, we combined three variables, including watching TV, listening to the radio, and reading the newspaper. The women were asked how often they have used the listed media and their responses ranged from “not at all=0”, “less than once a week=1” to “at least once a week=2”. The answers to all three questions were summarized and was then categorized into low exposure (total score 0–2) and high exposure (total score equal or above 3).

Information about the measles vaccination as the outcome variable was obtained based on mothers recall or from the child health card with or without a date. The variable was categorized as “vaccinated” if the child received MMR and “not vaccinated” otherwise.

Complex sample analyses were performed in order to account for cluster sampling design and sample weight in DHS data [39]. Descriptive characteristics of women and their children were presented with frequency (percentage) for categorical and mean (Standard Deviation (SD)) for continuous variables. We used chi-square tests for categorical and Student’s t-Test or one-way ANOVAs for continuous variables in order to evaluate bivariate associations between children’s background characteristics and their vaccination status. Factors associated with the outcome of interest with a p-value < 0.2 were included in the multivariate analyses. We used logistic regression models in order to identify factors associated with vaccination status. The results of this paper are presented with crude OR, where each exposure variable is analysed separately with the outcome of interest. In Model 1, factors associated with the outcome of interest (p < 0.2) in bivariate analyses were included. Model 2 additionally included wealth index and place of residence as they have been suggested as important predictors of child vaccination status in previous research [1, 4, 20, 23]. P values < 0.05 has been considered significant in the models. Statistical software package IBM SPSS Statistics version 24 (IBM, SPSS, Armonk,NY, USA) was used to perform the analysis.

The analysis of this study was based on existing survey data collected by the DHS (The DHS Programme, www.​dhsprogram.​com). The survey was approved by the ICF Institutional Review Board. The DHS has strict rules on providing informed consent to participants and ensure that the participants identity is protected. Before accessing the DHS data from the webpage an electronic registration form highlighting the desired data and plan for analysis was submitted and approved by the DHS.

A total of 525 women and their last-born child in the age of 12 to 35 months were included in the study (see Fig. 1). The background characteristics of the mothers and their children based on vaccination status are presented in Table 1. The mean age of the mothers was 28.35 years, 56.9% live in urban areas and 26.8% belonged to the richest wealth index quintile. More than half of the mothers had higher education (53.8%), around 97% were currently in a relationship and the mean age of their partners was 32.54 years. Among the partners, almost 10% had none or primary education, the rest had secondary or secondary special education.

The majority of the women (89.4%) did not perceive the distance to the health facility as a problem. The women received on average 7.79 antenatal care visits and almost 94% attended postnatal care visits two months after birth. Media exposure was low in approximately half of the women.

The mean age of the children was 22.6 months and 51.3% were boys. The children were mostly first and second born and 79.6% were vaccinated against measles. Among vaccinated children, vaccination status of 51% were recorded on the health cards and rest were based on mothers’ recall. The mean age of children who were vaccinated against measles was 23.15 months and only 2.9% (12 out of 418) children were vaccinated against measles at the age of 12 months (data not shown).

Children were more likely to be vaccinated against measles if they were older (p = 0.03), if they had a postnatal care visit within two months after birth (p = 0.01) and if their mothers had attended more antenatal care visits (p = 0.04) (Table 1).

In the crude logistic regressions, the age of the child, the maternal education level and attending postnatal care visits within two months after birth were significantly associated with the vaccination status of the child. The odds of being vaccinated were increasing with the increasing age of the child (OR 1.07; 95% CI: 1.02–1.12). Compared to children of mothers with basic education, children of mothers with secondary education were significantly more likely to be vaccinated (OR 2.86, 95% CI: 1.09–7.56). Also, children of mothers with higher education were more likely to be vaccinated, but the result was not significant (OR 1.87, 95% CI: 0.74–4.74). Furthermore, attending postnatal care visits within two months after birth increased the likelihood of the child being vaccinated more than twofold (OR 2.65, 95% CI: 1.29–5.45) (Table 2).

When adjusting for the potential confounders in Model 1, the odds of being vaccinated remained significantly associated with the children’s age, maternal education and attending a postnatal visit two months after birth. Children were more likely to be vaccinated if they were older (AOR 1.07, 95% CI: 1.03–1.12). Mothers who had secondary education were significantly more likely to vaccinate their children compared to those with basic education (AOR 2.76, 95% CI: 1.00–7.62). Attending postnatal care visits within two months after birth was also associated with higher odds of the child being vaccinated against measles (AOR 2.53, 95% CI: 1.14–5.63) (Table 2).

Additionally adjusting for place of residence and wealth index in Model 2 did not have an impact on the significance of the results. The current age of the child (AOR 1.07; 95% CI: 1.03–1.12), secondary level of education in mothers (AOR 3.38; 95% CI: 1.17–9.76) and attending postnatal visits within two months after birth (AOR 2.71; 95% CI: 1.17–6.30) stayed significantly associated with the vaccination status of the child. The latter two with even larger odds (Table 2).

Our study showed that the vaccination coverage in the studied population is far below the recommended 95% coverage and below the officially reported coverage [8]. The discrepancy between household surveys and reports from WHO/UNICEF has been recognized in previous studies [37] because they are largely based on administrative data from health-service-provider registries with potential weak administrative data systems [40]. There is a risk of over-reporting in these systems [40] and this may indicate that the actual coverage might be lower than official reports.

In the studied population, the mean age of the children in the vaccinated group was 23.15 months. Further a positive association between the age of the child and the probability of being vaccinated was observed. In both logistic regression models, the odds of being vaccinated increased with the increasing age of the child.

According to the Armenian vaccination schedule, measles vaccination should be given at the age of 12 months. Vaccinations given later than four weeks after the recommended age are considered to be not given on time [34] indicating that many children in the studied population may be vaccinated too late. Such delays in vaccination allow for longer periods of susceptibility [41] and could increase the possibility of measles transmissions in the society [26]. Previous studies have shown that incorrectly timed vaccinations were associated with outbreaks and re-emergence of a vaccine preventable disease [41].

The problem of delayed vaccinations is already known in Armenia. Schweizer et al. [34] showed that there have been improvements in the timeliness administration of the measles vaccine from year 2000 to 2010 in Armenia. However, our study indicated that there is still a potential delay. Agopian et al. [41] came to same conclusion and showed that the administration of Measles containing vaccines in Armenia was delayed with a median age of 61.1 weeks by analysing the ADHS 2015/2016.

The maternal level of education was associated with the vaccination status in this study. We found mothers with secondary level of education were more likely to vaccinate their child against measles than those with basic education. This is in line with the results from previous studies, where better education in mothers was associated with better immunization coverage [1, 20, 23, 24]. Often this is explained by the fact that educated mothers are more likely to remember dates, understand the importance of vaccination [22] or may have a higher income and thereby able to pay fees.

Children of mothers with higher education were not significantly more likely to be vaccinated compared to mothers with basic education. This might be surprising considering the benefit of maternal education on the child’s health. However, this finding might show a trend in Armenia, which has been shown in other, mainly high-income settings where clusters of vaccine hesitancy may appear in highly educated groups [42]. Other studies have shown there is a tendency among better educated mothers to skip vaccinations [30] and that the highest level of education is not associated with more positive views on vaccine importance and effectiveness in Armenia [18].

However, children to mothers with basic education were less likely to be vaccinated against measles than children to better educated mothers. The results suggested that interventions should focus on mothers with lower education in order to increase measles vaccination coverage.

Postnatal care visits two months after birth were significantly associated with measles vaccination in all regression models. This association has only been shown in other studies from the African context [25]. Attending postnatal care visits may represent health seeking behaviour and women who seek to see a health care professional for a check-up might be more likely to seek care for their child later. Furthermore, additional information about vaccination might have been provided during that visit and a positive relationship to the health care system might have been established [27].

These results show that attending postnatal care visits might be positively associated with the vaccination coverage and future interventions may benefit from encouraging mothers to attend postnatal care visits.

Measles is a disease with a high communicability, and therefore high immunization rates and vaccine efficacy is needed for reaching herd immunity [7]. In the primary studied population, 63 mothers (11%) were not aware of the vaccination status of their child and had to be excluded. There is a chance that some of these children were vaccinated and the mother only did not remember whether one of the injections, the child received, was a measles containing vaccine. However, data from WHO shows that a high percentage of measles cases in Armenia have an unknown or unclear vaccination status [43]. Thus, it is likely that children with an unknown vaccination status in this study were not vaccinated, which poses a risk for an outbreak. Furthermore, children who are not vaccinated against measles do not benefit from the potential immune training effects on the overall mortality by the measles vaccine, which has been shown in previous studies [5, 11].