Abstract
Background:
Delays in vaccination with the first/second dose of measles-containing vaccine (MCV1, MCV2) increases risk of measles transmission. MCV2 was added to Ethiopia’s routine immunization schedule in 2019, administered at age 15 months. We describe age at measles vaccination and characteristics associated with timely vaccination following MCV2 introduction.
Methods:
A secondary analysis from a 2021 cross-sectional household survey of children aged 12–35 months in Oromia, Ethiopia was conducted. MCV vaccination status and age at vaccination were determined from the child’s vaccination card. The cumulative card-confirmed vaccination coverage of MCV1 and MCV2 (among children 24–35 months) by age at vaccination was estimated. Among children who received MCV, vaccination timing was classified as early (MCV1: <9 months; MCV2: <15 months), timely (MCV1: 9 to <10 months; MCV2: 15 to <16 months), or delayed (MCV1: ≥10 months; MCV2: ≥16 months). Characteristics associated with timely MCV1 vaccination were analyzed using multivariate logistic regression, with urban/rural stratification.
Results:
Overall, 651 children aged 12–35 months (MCV1) and a subset of 270 children aged 24–35 months (MCV2) were analyzed. Cumulative MCV1 coverage was 59.9% by age 9 months and 76.3% by age 12 months, and MCV2 was 33.0% by age 15 months and 44.8% by age 24 months. MCV coverage was generally higher in urban compared to rural areas. Among vaccinated children, the proportion receiving early, timely, and delayed doses was 30.1%, 44.5%, and 25.3% for MCV1, and 28.6%, 41.2%, and 30.3% for MCV2, respectively. Characteristics associated with timely MCV1 vaccination differed in urban and rural areas, but common themes included increased caregiver immunization awareness and service delivery aspects.
Conclusions:
Although most children who received MCV1 were vaccinated by 12 months, many received delayed doses. MCV2 coverage was low, irrespective of vaccination timing. Findings highlight the importance of routine catch-up vaccination policies and guidance, and the need for tailored strategies in urban and rural settings to improve MCV coverage and timely vaccination.
Keywords: Measles, Measles-containing vaccine, Second year of life, Immunization coverage, Vaccination timing
Introduction
Measles vaccination coverage is a core indicator of the strength of immunization programs, as highlighted by the Immunization Agenda 2030 [1]. Although >95% vaccine coverage with two timely doses of measles-containing vaccine (MCV) is recommended, global coverage of the first dose of MCV (MCV1) only reached 84% in 2024 with the uptake of MCV2 lagging even further (76%) [2]. Strengthening the coverage of MCV2 is essential to boost protection and prevent outbreaks, but also because the second year of life contact point provides one of the few opportunities beyond infancy to provide catch-up doses of other recommended vaccines. Challenges with MCV2 uptake, in particular, include gaps in health worker training and knowledge of policies, low awareness among caregivers about the timing and need for a second dose, and errors made while recording doses [3–5].
In addition to shortfalls in overall MCV coverage, the timing of vaccination plays an important role in acute and long-term protection against measles and the effective population coverage of vaccination; early vaccination has been associated with attenuated immune responses to MCV [6,7] and delayed vaccination from the recommended routine immunization schedule could leave children susceptible to measles during a period of high exposure [8,9]. According to evidence synthesized in the World Health Organization’s position paper on measles vaccines, the recommended age for MCV1 vaccination must balance potential primary vaccine failure that occurs at younger ages with the risk of exposure to measles virus [8]. Although vaccine efficacy is estimated to be higher at 12 months (92.5%) compared to 9–11 months (84.0%), in countries such as Ethiopia, where incidence and mortality from measles in infants is high [10], routine vaccination with MCV1 is recommended at 9 months to protect against this heightened risk of infection [8]. A second dose of MCV (MCV2) is recommended irrespective of national MCV1 coverage to protect children who failed to seroconvert following MCV1 vaccination or did not receive MCV1 [8]. Collectively, these considerations highlight the importance of timely vaccination to prevent measles transmission, particularly in high-risk settings, and provide indicators for the strength of the second year of life platform for delivering immunization services, including MCV2 and catch-up vaccination.
In Ethiopia, MCV2 was introduced as part of the routine immunization schedule in 2019. The initial guidance for health care workers for MCV2 introduction rolled out in 2018 stated that MCV1 should be administered at 9 months and MCV2 should be administered between ages 15 to 18 months [11]. Subsequently, an updated Ethiopia Expanded Program on Immunizations (EPI) guide in 2021 stated that MCV2 should be given at 15 months of age, with a minimum interval between MCV1 and MCV2 doses of 4 weeks [12] and upper age limit of 24 months (at the time of the present assessment). Previous studies [4,5] have reported on factors affecting the uptake of MCV2, shortly after its introduction in the routine EPI schedule in Ethiopia; however, these analyses did not explore vaccination timing relative to the recommended schedule. In this study, we described the timing of MCV1 and MCV2 vaccination in urban and rural areas of Oromia based on the recommended EPI schedule in Ethiopia and we also conducted an exploratory analysis of household, caregiver, and child-level factors associated with timely compared to delayed receipt of MCV1.
Methods
This study reports on a secondary analysis of data collected as part of a larger assessment to understand immunization demand and service delivery barriers and facilitators for MCV2 vaccination two years after its introduction in Oromia Region, Ethiopia [4,5]. Detailed information on the study design and methodology are provided in the main publication [5] and also summarized below.
Study setting and design
This study was conducted in Oromia Region of Ethiopia, where an estimated 37% (40 million) of Ethiopia’s population lived in 2021 [13]. As part of the main study [5], a cross-sectional household survey with caregivers of children ages 12–35 months was carried out in 18 districts (woredas) of Oromia (stratified by urban versus rural settlement, administrative MCV1 baseline coverage ≥80% versus <80%, and recent measles outbreak status).
Data collection
This sub-study analyzed a subset of data collected during the household survey and included information on household (e.g. distance from nearest health facility and number of children <5 years of age in the household), caregiver (e.g. age, educational attainment, religion, and marital status), and child demographics (e.g. age based on date of birth recorded on the vaccination card, sex, and location of birth); caregiver’s knowledge, attitudes, and practices surrounding vaccination; and the child’s vaccination status and vaccination dates for MCV1 and MCV2, as recorded from home-based vaccination records (HBRs; e.g. vaccination card). In-person household survey interviews were conducted by trained surveyors using standardized and pre-tested questionnaires. Data were entered electronically on tablets with an electronic questionnaire pre-programmed with skip patterns and logic checks using Open Data Kit (ODK) software [14].
Data analyses
For this sub-study of timeliness of routine measles vaccination doses, the analytic dataset was restricted to age-eligible children with an available HBR and, among those who received MCV1 or MCV2, with valid recorded doses of MCV1 or MCV2. Considering age-eligibility for MCV doses in Ethiopia (recommended at 9 months for MCV1 and at 15 months for MCV2), we included children 12–35 months for MCV1 and 24–35 months for MCV2. The age range of 24–35 months was selected for MCV2 analysis to be consistent with standardized age-ranges used for second year of life vaccination coverage estimates and to reduce potential biases introduced by censoring vaccination status due to assumption that a child was never vaccinated if they did not receive MCV2 by their age of vaccination at the time of the survey. This assumption is weaker for “younger” children 15–24 months and, therefore, we restricted our analyses for MCV2 to children ages 24–35 months even though this reduces the sample size and precision of estimates for MCV2.
Valid MCV doses were defined as those with recorded date(s) of MCV1 or MCV2 administration, where MCV1 was given prior to MCV2 (as determined by the HBR vaccination dates), the interval between MCV1 and MCV2 dates was at least four weeks apart, and the minimum age at vaccination for MCV was at least five months. The age of vaccination was considered as the difference between the child’s date of birth and the date of MCV1 or MCV2 vaccination (calculated in days and converted to complete months), as recorded on the HBR. Reported vaccination coverage estimates reflect HBR confirmed coverage. The timing of MCV vaccination was described by plotting the cumulative proportion of children vaccinated for MCV1 or MCV2 against the age at vaccination. The analysis of the timing of vaccination was further restricted to children who were vaccinated for MCV1 or MCV2 (vaccinated subset) and the timeliness of doses was described by the proportion of children who were vaccinated early, at the recommended age (timely), or after the recommended age (delayed) based on the age of vaccination. Two different criteria were considered for analysis of timely and delayed doses, one defined by Ethiopia’s EPI schedule and the second to include a broader age range based on the World Health Organization’s EPI schedule; the criteria based on the routine immunization schedule were defined in close collaboration with Ethiopia’s EPI program to ensure that these definitions were programmatically relevant. For MCV1, early doses were defined as those received <9 months of age, timely doses were considered as either doses received between 9 to <10 months (e.g. Ethiopia’s EPI schedule) or 9 to <12 months (e.g. broader classification), and delayed doses were those received ≥10 months or ≥12 months. For MCV2, early doses were those received <15 months of age, timely doses were considered as either doses received between 15 to <16 months or 15 to <19 months, and delayed doses were those received ≥16 months or ≥19 months. No buffer period was considered for these criteria, as per consultation with Ethiopia’s EPI program. Conversions from days to months assumed an average of 30.4 days per month. Descriptive results were presented graphically and stratified by urban and rural areas, with results also included for all areas combined.
Among children 12–35 months who received MCV1 (vaccinated subset), characteristics or factors associated with timely receipt of MCV1 were examined in exploratory models using bivariate and multivariate logistic regression. Children receiving early doses of MCV1 were excluded from the association analyses since understanding factors associated with timely compared to delayed doses has greater programmatic relevance and comparisons using a multinomial logistic regression model were limited due to insufficient outcome sample size. The association analysis was only conducted for MCV1 as an outcome, since the sample size for MCV2 (among children ages 24–35 months) was insufficient to obtain stable regression estimates. Given the sampling design, models were run separately for urban and rural districts. The outcome of timely receipt of MCV1 was compared to delayed receipt of MCV1, considering timely as receipt of MCV1 between 9 to <10 months of age and delayed as receipt of MCV1 between 10 and 35 months of age. All possible combinations of explanatory variables were examined, and the final multivariate model was selected based on the lowest Akaike information criterion (AIC) score. Associations from regressions were described using odds ratios (OR) and adjusted ORs are presented with 95% confidence intervals (CIs). P-values <0.05 were considered statistically significant. Data were managed and analyzed in R [15]. Additionally, as noted as a limitation in the main study [5], sampling weights were not estimated or applied to the analyses due to the unavailability of reliable census data needed to calculate weights and the combination of sampling by convenience (due to insecurity in the study area) and randomized selection.
Results
Among 1185 children included in the household survey, 674 (57%) had available HBRs with ages between 12 and 35 months based on recorded date of birth (age-eligible for MCV1 analysis) and 651 (97%) had valid HBR records to be included in the analytic subset for MCV1. Of these, 279 (41%) were between 24 and 35 months (age-eligible for MCV2 analysis), and 270 (97%) had received MCV1 and had valid HBR records for MCV2 doses to be included in the analytic subset for MCV2. The HBR-confirmed vaccinated subset for timeliness analysis included 521 children 80% the analytic subset for MCV1) who were vaccinated for MCV1 and 119 children (44% of the analytic subset for MCV2) who were vaccinated for MCV2 (Fig. 1). The distribution of descriptive characteristics of children included in the household survey (n = 1185) and those included in the analytic subset for MCV1 (n = 651) is provided in Supplementary Table 1.
Fig. 1.
Flow diagram of data included in the analytic and vaccinated subsets to assess MCV1 and MCV2 timeliness, Oromia, Ethiopia – 2021. * Analyses carried out in analytic subset: cumulative proportion vaccinated by age of vaccination and HBR-confirmed coverage stratified by urban/rural setting. ** Analyses carried out in vaccinated subset: timeliness of vaccination and exploratory regression for MCV1 stratified by urban/rural setting. NOTES: HBR = home-based vaccination record, MCV = measles-containing vaccine.
Cumulative MCV1 coverage and timing of vaccination
Among children 12–35 months in all areas combined (n = 651; analytic subset), cumulative MCV1 coverage was 59.9% (rural 45.9%, urban 69.0%) by age 9 months and 76.3% (rural 61.9%, urban 85.8%) by age 12 months. At all ages, cumulative MCV1 coverage was the same as or higher in urban compared to rural areas (Fig. 2, Supplementary Table 3).
Fig. 2.
Cumulative home-based vaccination card coverage of the first dose of measles-containing vaccine (MCV1) by age at vaccination among children 12–35 months* (N = 651) in Oromia, Ethiopia – 2021. * MCV1 vaccination status based on home-based vaccination record.
Among children 12–35 months who received MCV1 (n = 521; vaccinated subset), 30.1% (rural 29.7%, urban 30.4%) received MCV1 before the recommended age (<9 months), 44.5% (rural 38.4%, urban 47.6%) received MCV1 at the recommended age (between 9 and < 10 months), and 25.3% (rural 32.0%, urban 22.1%) received MCV1 later than the recommended age (≥ 10 months) (Fig. 3, Supplementary Table 3). Of the 30.1% who received MCV1 prior to 9 months of age, many (71.1%) of these children were vaccinated at 8 months of age. MCV1 doses given prior to 8 months were more frequent in children living in rural compared to urban areas. Considering the broader definition of 9 to <12 months as a timely MCV1 dose, 63.5% (rural 58.7%, urban 65.9%) received MCV1 at the recommended age range, and 6.3% (rural 11.6%, urban 3.7%) received MCV1 later than recommended age (≥ 12 months) (Fig. 3, Supplementary Table 3).
Fig. 3.
Timing of the first dose of measles-containing vaccine (MCV1) by age at vaccination among children 12–35 months* who received MCV1 (N = 521) in Oromia, Ethiopia – 2021. * MCV1 vaccination status based on home-based vaccination record.
Cumulative MCV2 coverage and timing of vaccination
Among children 24–35 months who received MCV1 (n = 270; analytic subset), cumulative MCV2 coverage was 33.0% (rural 20.8%, urban 39.7%) by age 15 months and only increased to 44.8% (rural 36.5%, urban 49.4%) by age 24 months. For children vaccinated at 15 months of age and later, cumulative MCV2 coverage was higher in urban compared to rural areas (Fig. 4, Supplementary Table 4).
Fig. 4.
Cumulative home-based vaccination card coverage of the second dose of measles-containing vaccine (MCV2) by age at vaccination among children 24–35 months* (N = 270) in Oromia, Ethiopia – 2021. * MCV2 vaccination status based on home-based vaccination record.
Among children 24–35 months who received MCV2 (n = 119; vaccinated subset), 28.6% (rural 31.2%, urban 27.6%) received MCV2 before the recommended age (<15 months), 41.2% (rural 15.6%, urban 50.6%) received MCV2 at the recommended age (between 15 and < 16 months), and 30.3% (rural 53.1%, urban 21.8%) received MCV2 after the recommended age (≥ 16 months) (Fig. 5, Supplementary Table 4). Considering 15 to <19 months as a timely MCV2 dose, 64.7% (rural 56.2%, urban 67.8%) received MCV2 at the recommended age range, and 6.7% (rural 12.5%, urban 4.6%) received MCV2 after the recommended age (≥ 19 months) (Fig. 5, Supplementary Table 4).
Fig. 5.
Timing of the second dose of measles-containing vaccine (MCV2) by age at vaccination among children 24–35 months* who received MCV2 (N = 119) in Oromia, Ethiopia – 2021. * MCV2 vaccination status based on home-based vaccination record; please note the small sample size available in rural areas when interpreting these results.
Household, caregiver, and child characteristics associated with timely vaccination with MCV1
In exploratory analyses, associations between household, caregiver, and child demographic factors with timely compared to delayed MCV1 vaccination among children 12–35 months differed in urban and rural settings (vaccinated subset); bivariate regression results are described in Supplementary Table 2and results from multivariate regression with variables selected based on lowest AIC are provided in Table 1 for urban areas and Table 2 for rural areas.
Table 1.
Multivariate associations between household, caregiver, and child demographic factors and timely receipt of the first dose of measles-containing vaccine (MCV1) among children 12–35 months living in urban (N = 239) areas of Oromia, Ethiopia – 2021.
| Characteristic | Urban (N = 239) Timely (n = 164) compared to delayed (n = 75) receipt of MCV1 |
||
|---|---|---|---|
| n | aOR (95% CI) | p-value | |
| Number of vaccination visits child needs | 0.13 | ||
| 0 to 5 | 169 | ref | |
| 6 (correct as per EPI schedule) | 43 | 2.29 (0.99, 5.92) | |
| Don’t know | 27 | 0.84 (0.36, 2.05) | |
| Ever turned away from health center due to vaccine stock-out | 0.016* | ||
| No | 211 | ref | |
| Yes | 28 | 0.35 (0.15, 0.82) | |
| Travel time to vaccination center by walking | 0.36 | ||
| Between 1 and 6 h | 26 | ref | |
| >30 min to 1 h | 38 | 0.56 (0.15, 1.86) | |
| 30 min or less | 171 | 0.53 (0.17, 1.44) | |
| Don’t know | 4 | – | |
NOTES. aOR = adjusted odds ratio, CI = confidence interval, EPI = expanded program on immunizations, MCV1 = first dose of measles-containing vaccine, ref. = referent category; 4 observations in urban areas were dropped from multivariate analysis due to missing covariate data for the following variables: number of vaccination visits child needs (n = 2) and/or ever turned away from health center due to vaccine stock-out (n = 2).
p < 0.05.
Table 2.
Multivariate associations between household, caregiver, and child demographic factors and timely receipt of the first dose of measles-containing vaccine (MCV1) among children 12–35 months living in rural (N = 119) areas of Oromia, Ethiopia – 2021.
| Characteristic | Rural (N = 119) Timely (n = 64) compared to delayed (n = 55) receipt of MCV1 |
||
|---|---|---|---|
| n | aOR (95% CI) | p-value | |
| Caretaker’s age | 0.091 | ||
| 18–26 years | 78 | ref | |
| 27–80 years | 41 | 0.45 (0.17, 1.13) | |
| Sex of child | 0.34 | ||
| Male | 48 | ref | |
| Female | 71 | 0.62 (0.22, 1.65) | |
| Delivery location | 0.19 | ||
| Home | 21 | ref | |
| At HF or on the way to HF | 98 | 2.33 (0.66, 8.71) | |
| Number of vaccination visits child needs | |||
| 0 to 5 | 98 | ref | 0.39 |
| 6 (correct as per EPI schedule) | 11 | 0.43 (0.09, 2.03) | |
| Don’t know | 10 | 2.07 (0.31, 14.8) | |
| Number of times child is supposed to get measles vaccination | 0.032* | ||
| Don’t know about measles vaccine or number of doses | 62 | ref | |
| One dose | 21 | 1.62 (0.45, 5.94) | |
| Two doses | 36 | 4.56 (1.45, 16.1) | |
| Role of parents in immunization decision making for child | 0.038* | ||
| Mother or father only (one parent) | 29 | ref | |
| Both mother and father | 90 | 3.07 (1.06, 9.60) | |
| Frequency of vaccination availability | 0.21 | ||
| Every month | 75 | ref | |
| Every week | 30 | 0.73 (0.25, 2.10) | |
| Every day | 1 | – | |
| Don’t know | 13 | 3.13 (0.66, 16.9) | |
| Wait time at the vaccination center for vaccination | 0.002* | ||
| Between 1 and 6 h | 13 | ref | |
| >30 min to 1 h | 31 | 34.1 (2.55, 402)** | |
| 30 min or less | 72 | 27.7 (3.86, 592)** | |
| Don’t know | 3 | – | |
NOTES. aOR = adjusted odds ratio, CI = confidence interval, EPI = expanded program on immunizations, HF = health facility, MCV1 = first dose of measles-containing vaccine, ref. = referent category; 2 observations in rural areas were dropped from multivariate analysis due to missing covariate data for one or more of the following variables: caregiver believes that child has received all recommended vaccines (n = 1), number of vaccination visits child needs (n = 1), and/or role of parents in immunization decision making for child (n = 1).
p < 0.05.
Results should be interpreted with caution due to the small cell count for the number of observations in the referent category with the outcome (n = 2).
In urban areas, the final variables included in the exploratory model were the caregiver’s knowledge of the number of vaccination visits needed for the child, previous experiences with being turned away due to vaccine stock-out, and the travel time by walking to the vaccination center (Table 1). Of these, less timely MCV1 vaccination was significantly associated with previous experience with being turned away due to vaccine stock-out (ever being turned away versus not: aOR 0.35, 95% CI 0.15–0.82; Table 1).
In rural areas, the final variables included in the exploratory model were the caregiver’s age in years, child’s sex, birth location for the child, caregiver’s knowledge of the number of vaccination visits needed for the child, caregiver’s knowledge of the recommended number of doses of measles vaccine, the role of the parents in vaccine decision making, frequency of vaccination availability, and wait time at the vaccination center (Table 2). Of these variables, timely compared to delayed MCV1 vaccination was significantly associated with knowing that two doses of measles vaccine were needed (aOR 4.56, 95% CI 1.45–16.1; Table 2), the involvement of both parents (versus one, usually only the mother) in vaccine decision making for the child (aOR 3.07, 95% CI 1.06, 9.60; Table 2), and experiencing shorter wait times at the vaccination center (30 min or less versus 1–6 h: aOR 27.7, 95% CI 3.86–592; Table 2).
Discussion
Examination of vaccination timeliness provides context for population adherence to EPI schedules and insight into operational aspects of immunization service delivery for immunization programs. In this study, we found that most children who received MCV1 were vaccinated by 12 months of age, but a notable proportion also received delayed doses. In contrast, the overall coverage of MCV2 remained low, irrespective of age of vaccination. Measles vaccination coverage and timely receipt of vaccination could be improved through additional vaccination opportunities such as routine catch-up vaccination strategies and efforts to reduce missed opportunities for vaccination (MOV) [16].
Among children vaccinated for MCV1 in this study, nearly 25% received MCV1 after 10 months of age. The proportion of delayed MCV1 doses dropped to 6% after 12 months of age, suggesting that most children who were not vaccinated by the recommended age of 9 months were vaccinated by 12 months of age. However, any delays in vaccination could leave children vulnerable to VPDs and contribute to ongoing community transmission of VPDs. Additionally, some studies have demonstrated that delays in vaccination from earlier doses can lead to delayed schedule completion or even drop-out [17,18]. Service delivery (e.g. vaccine stock-outs, service provider availability) and caregiver access-related factors may affect vaccination completion and timeliness [19]. Vaccination coverage can be boosted for those children who have fallen behind on the EPI schedule through catch-up vaccination strategies such as integrated child health days, screening vaccination status and providing missed doses at non-immunization visits, catch-up campaigns, periodic intensification of routine immunization activities, etc. Catch-up vaccination strategies should be an ongoing component of routine immunization systems to ensure that children have timely and ample opportunities to receive their recommended vaccinations [20].
In addition to delayed doses, almost one-third of children in this study received MCV1 prior to the recommended age of 9 months. A similar proportion of early doses (27.6%) was noted in an analysis of 2016 Demographic Health Survey data in Ethiopia [21]. Early vaccination has the potential to reduce vaccine efficacy by interfering with protection provided by maternal antibodies [22] or attenuating the immune response to subsequent measles vaccination doses [23]. Programmatically, these doses are considered invalid and could reflect gaps in knowledge, training, or supportive supervision for health care workers who administer vaccines [24,25]. Additionally, these doses may overstate population coverage estimates, since early vaccination may provide suboptimal protection. Further examination of the timing of MCV1 administration in this study revealed that many of the early doses were given at 8 months of age, just prior to the recommended age, suggesting that health care workers may be weighing the potential for missed opportunities for vaccination, particularly as some facilities may not offer immunization services every day/week [26], with adherence to the national EPI schedule. Although the scope of this study did not allow for more detailed examination of the reasons for early vaccination, additional research could examine health care worker practices surrounding early vaccination, identify other reasons for early vaccination, and explore how immunization service availability may impact the timing of measles vaccination. Additionally, the recent roll out of the measles 5- dose vaccination vial in Ethiopia provides an opportunity to examine whether the new vial size improves health care worker adherence to vaccination schedules.
This study also examined urban and rural differences in vaccination timeliness and explored factors associated with timely compared to delayed vaccination. For both MCV1 and MCV2, vaccination coverage at all ages and the proportion of children receiving timely doses were higher in urban compared to rural settings. The gap in uptake of timely vaccination in rural settings and differences in contributing factors in urban-rural settings has been noted in other studies in Tanzania [27,28]. One study noted larger urban-rural differences in coverage for doses given later in the series [27]. While our study found a similar urban-rural gap for MCV1 and MCV2, we found that the factors associated with vaccination timeliness differ by setting; previous experience with being turned away due to vaccine stock-out was associated with delayed MCV1 in urban settings, and increased caregiver knowledge or awareness about measles vaccination, involvement of both parents for vaccine-decision making, and shorter wait times for vaccination were associated with timely MCV1 in rural settings. Similarly, lower maternal education was associated with delayed vaccination in both rural and urban settings of Tanzania, while lower access to vaccination (e.g. greater distance to nearest health facility) was associated with delayed vaccination in rural settings only [27]. These findings highlight that different strategies may be needed in urban and rural settings to improve MCV coverage and timeliness.
This study had several limitations. First, this analysis of vaccination timeliness was a secondary aim of the study and, therefore, the study design (including sample size) and data collection tools did not intentionally focus on variables and factors related to the timing of vaccination receipt. Second, there is variability in the published literature on how to define and assess vaccination timeliness; our study considered multiple definitions with a focus on programmatic relevance for the interpretation of results. Third, our analysis was restricted to children with a vaccination card to analyze the dates of vaccination for MCV, which may introduce selection bias and limits the generalizability of our findings, as nearly half of the study sample did not have vaccination cards at the time of the study. We examined the distribution of measured characteristics of children included in the household survey and those included in the analytic subset (Supplementary Table 1) and noted a higher proportion of children residing in rural areas and a higher proportion of home births in the household survey compared to the analytic subset. Vaccination card retention may be lower in rural areas, which may explain the lower proportion of rural residents in the analytic subset (restricted to children with vaccination cards). Additionally, home births are more common in rural areas, which may also explain the higher proportion noted in the household survey compared to the analytic subset. We attempted to minimize selection bias due to differences in these measured variables by stratifying our regression analyses by urban/rural setting. Finally, we excluded early doses from our exploratory analysis of characteristics associated with timely MCV1 vaccination since most of the early vaccinations in our study occurred within several weeks before the recommended age. Although early doses were not fully examined in our study, these doses warrant further examination to ensure that MCV coverage estimates, which include early doses, accurately reflect the population immunity against measles.
Conclusions
The findings from this study highlight the importance of considering vaccination timeliness alongside vaccination coverage for refining operational aspects of routine immunization programs. These challenges differed for MCV1 and MCV2; while MCV1 HBR-confirmed coverage was around 80% overall, many children received early or delayed doses, in contrast MCV2 uptake remained considerably lower, irrespective of vaccination timing. These results support the need for several programmatic responses. Vaccination coverage and the reduction in vaccination delays could be improved through enhancing focus on routine catch-up vaccination strategies beyond the recommended age, efforts to reduce MOVs, and identifying additional opportunities for children to become vaccinated. Defaulter tracing or reminder-recall systems could improve timely vaccination, while early doses could be prevented through strategies aimed at improving health worker adherence to vaccination schedules and exploring other reasons for early vaccination. Additionally, the urban-rural gaps noted in this study indicate that different or tailored strategies may be needed in urban and rural settings to improve MCV coverage and timely vaccination.
Supplementary Material
Acknowledgments
We thank the Oromia Health Bureau for providing measles surveillance data for the selection of assessment areas. This study was funded by Gavi, the Vaccine Alliance via the Ethiopia Targeted Country Assistance, Partner Engagement Framework, through the CDC-Foundation and from the U.S. Centers for Disease Control and Prevention.
Disclaimer
The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the U.S. Centers for Disease Control and Prevention or the CDC Foundation.
Abbreviations:
- AIC
Akaike information criterion
- CI
confidence interval
- EPI
expanded program on immunization
- HBR
home-based vaccination record
- HF
health facility
- MCV
measles-containing vaccine
- OR
odds ratio
Appendix A. Supplementary data and analyses
Supplementary data to this article can be found online at https://doi.org/10.1016/j.jvacx.2026.100892.
Footnotes
CRediT authorship contribution statement
Monica P. Shah: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Data curation. Abyot Bekele Woyessa: Writing – review & editing, Supervision, Methodology, Investigation, Formal analysis, Conceptualization. Almea M. Matanock: Writing – review & editing, Supervision, Methodology, Formal analysis. Binyam Moges Azmeraye: Supervision, Investigation. Leuel Lisanwork: Supervision, Investigation. Getnet Yimer: Supervision, Project administration. Shu-Hua Wang: Supervision, Project administration. Xueliang Pan: Data curation. Paulos Samuel: Conceptualization. Bekana Tolera: Conceptualization. Birhanu Kenate Sori: Conceptualization. Abebe Bekele: Conceptualization. Tesfaye Deti: Supervision, Methodology, Investigation, Formal analysis, Conceptualization. Getachew Wako: Conceptualization. Amsalu Shiferaw: Conceptualization. Yohannes Lakew Tefera: Conceptualization. Melkamu Ayalew Kokebie: Formal analysis, Conceptualization. Tatek Bogale Anbessie: Supervision, Project administration. Habtamu Teklie Wubie: Conceptualization. Qian An: Writing – review & editing, Methodology, Formal analysis. Aaron Wallace: Writing – review & editing, Supervision, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization. Ciara E. Sugerman: Writing – review & editing, Supervision, Project administration, Methodology, Funding acquisition, Formal analysis, Conceptualization.
Informed consent statement
Verbal informed consent was obtained from all participants prior to conducting interviews and participants had the opportunity to ask questions, decline participation, and withdraw from the interview at any time during the survey process.
Institutional review board statement
The protocol for this work underwent scientific and ethical review by the Oromia Health Bureau ethical review board (approval certificate letter number (BEFO/UBTFU/1-16/875). The protocol was also reviewed by the U.S. Centers for Disease Control and Prevention and received a non-research determination. Permissions were obtained at each level of the government in Ethiopia to conduct the assessment.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Implementation of this study reports financial support was provided by Gavi, the Vaccine Alliance via the Ethiopia Targeted Country Assistance, Partner Engagement Framework, through the CDC-Foundation and from the U.S. Centers for Disease Control and Prevention. CES has participated in several health and immunization system strengthening working groups led by Gavi (one of the funding entities) during the work and write up period. This role was purely technical and non-financial. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
The datasets used to prepare this report are available at the Oromia Regional Health Bureau and can be obtained from the corresponding author upon reasonable request.
References
- [1].Lindstrand A, Mast E, Churchill S, Rahimi N, Grevendork J, Brooks A, et al. Implementing the immunization agenda 2030: a framework for action through coordinated planning, monitoring & evaluation, ownership & accountability, and communications & advocacy. Vaccine 2024;42(Suppl. 1):S15–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Antoni S, Ferrari M, Wimmer A, Nedelec Y, Steulet C, Gacic-Dobo M, et al. Progress towards measles elimination – worldwide, 2000–2024. World Health Organ Week Epidemiol Rec 2025;100(48):591–604. [Google Scholar]
- [3].Masresha BG, Luce R, Okeibunor J, Shibeshi ME, Kamadjeu R, Fall A. Introduction of the second dose of measles containing vaccine in the childhood vaccination programs within the WHO Africa region - lessons learnt. J Immunol Sci 2018; Suppl:113–21. [PMC free article] [PubMed] [Google Scholar]
- [4].Solomon K, Aksnes BN, Woyessa AB, Geri C, Matanock AM, Shah MP, et al. Qualitative insights on barriers to receiving a second dose of measles-containing vaccine (MCV2), Oromia Region of Ethiopia. Vaccines (Basel) 2024;12(7). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Woyessa AB, Shah MP, Azmeraye BM, Pan J, Lisanwork L, Yimer G, et al. Factors associated with uptake of routine measles-containing vaccine doses among young children, Oromia Regional State, Ethiopia, 2021. Vaccines (Basel) 2024;12(7). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].van der Staak M, Ten Hulscher HI, Nicolaie AM, Smits GP, de Swart RL, de Wit J, et al. Long-term dynamics of measles virus-specific neutralizing antibodies in children vaccinated before 12 months of age. Clin Infect Dis 2025;80(4):904–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Gans HA, Arvin AM, Galinus J, Logan L, DeHovitz R, Maldonado Y. Deficiency of the humoral immune response to measles vaccine in infants immunized at age 6 months. JAMA 1998;280(6):527–32. [DOI] [PubMed] [Google Scholar]
- [8].Measles vaccines: WHO position paper - April 2017. Wkly Epidemiol Rec 2017;92(17):205–27. [PubMed] [Google Scholar]
- [9].Bielicki JA, Achermann R, Berger C. Timing of measles immunization and effective population vaccine coverage. Pediatrics 2012;130(3):e600–6. [DOI] [PubMed] [Google Scholar]
- [10].Eshetu D, Tosisa W, Regassa BT, Hundie GB, Mulu A. Epidemiology of measles outbreaks, incidence and associated risk factors in Ethiopia from 2000 to 2023: a systematic review and meta-analysis. BMC Infect Dis 2024;24(1):914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Federal Democratic Republic of Ethiopia Ministry of Health. Measles containing vaccine second dose introduction (MCV2) training manual for health workers. Addis Ababa, Ethiopia. 2018. March. [Google Scholar]
- [12].Ministry of Health Ethiopia. National implementation guideline for expanded program on immunization (revised edition). Addis Ababa, Ethiopia. 2021. June. [Google Scholar]
- [13].United Nations Office for the coordination of humanitarian affairs. Ethiopia - subnational population Statistics 2022. 2023. March. [Google Scholar]
- [14].Open Data Kit. San Diego, CA, USA:ODK Inc. URL https://getodk.org. [Google Scholar]
- [15].RStudio: integrated development for R. Boston, MA: RStudio. PBC. URL http://www.rstudio.com/: RStudio Team. [Google Scholar]
- [16].Intervention guidebook for implementing and monitoring activities to reduce missed opportunities for vaccination. Geneva: World Health Organization; 2019. [Google Scholar]
- [17].Janusz CB, Frye M, Mutua MK, Wagner AL, Banerjee M, Boulton ML. Vaccine delay and its association with undervaccination in children in sub-Saharan Africa. Am J Prev Med 2021;60(1 Suppl 1):S53–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Muhoza P, Shah MP, Amponsa-Achiano K, Gao H, Quaye P, Opare W, et al. Timeliness of childhood vaccinations following strengthening of the second year of life (2YL) immunization platform and introduction of catch-up vaccination policy in Ghana. Vaccines 2024;12(7). [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Bangura JB, Xiao S, Qiu D, Ouyang F, Chen L. Barriers to childhood immunization in sub-Saharan Africa: a systematic review. BMC Public Health 2020;20(1):1108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Leave no one behind: guidance for planning and implementing catch-up vaccination. Geneva: World Health Organization; 2021. [Google Scholar]
- [21].Desta T, Lemango E, Wayessa D, Wondowossen L, Kerie M, Masresha B. Invalid measles vaccine dose administration and vaccine effectiveness in Ethiopia. Pan Afr Med J 2021;40:229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Sato H, Albrecht P, Reynolds DW, Stagno S, Ennis FA. Transfer of measles, mumps, and rubella antibodies from mother to infant. Its effect on measles, mumps, and rubella immunization. Am J Dis Child 1979;133(12):1240–3. [DOI] [PubMed] [Google Scholar]
- [23].Nic Lochlainn LM, de Gier B, van der Maas N, van Binnendijk R, Strebel PM, Goodman T, et al. Effect of measles vaccination in infants younger than 9 months on the immune response to subsequent measles vaccine doses: a systematic review and meta-analysis. Lancet Infect Dis 2019;19(11):1246–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Kidanne L, Solomon M, Bisrat F, Asres M, Tadesse T, Asress A, et al. Child vaccination timing, intervals and missed opportunities in pastoral and semi-pastoral areas in Ethiopia. Ethiop J Health Dev 2019;33(Special issue). [Google Scholar]
- [25].Akmatov MK, Kimani-Murage E, Pessler F, Guzman CA, Krause G, Kreienbrock L, et al. Evaluation of invalid vaccine doses in 31 countries of the WHO African region. Vaccine 2015;33(7):892–901. [DOI] [PubMed] [Google Scholar]
- [26].Tsega A, Hausi H, Chirwa G, Steinglass R, Smith D, Valle M. Vaccination coverage and timely vaccination with valid doses in Malawi. Vaccine Rep 2016;6:8–12. [Google Scholar]
- [27].Yelverton V, Hair NL, Ghosh SH, Mfinanga SG, Ngadaya E, Baumgartner JN, et al. Beyond coverage: rural-urban disparities in the timeliness of childhood vaccinations in Tanzania. Vaccine 2022;40(37):5483–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Nadella P, Smith ER, Muhihi A, Noor RA, Masanja H, Fawzi WW, et al. Determinants of delayed or incomplete diphtheria-tetanus-pertussis vaccination in parallel urban and rural birth cohorts of 30,956 infants in Tanzania. BMC Infect Dis 2019;19(1):188. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets used to prepare this report are available at the Oromia Regional Health Bureau and can be obtained from the corresponding author upon reasonable request.