中国P站

Immunization; Poisson regression; MEDHS 2019; Prevalence rate

BCG: Bacille-Calmette-Guerin; DPT: Diphtheria, Pertussis and Tetanus; EMDHS: Ethiopia Mini Demographic and Health Survey; EPI: Expanded Program on Immunization; ICF: Inner City Fund; PCV: Pneumococcal Conjugate Vaccine; SSA: Sub-Saharan Africa; VPD: Vaccine-Preventable Diseases; CSA: Central Statistical Agency; VIF: Variance Inflation Factor

Vaccine-Preventable Childhood Diseases (VPDs), notably diphtheria, hepatitis B, Homophiles influenza type B (Hib), measles, mumps, pertussis, pneumonia, polio, rotavirus, diarrhea, rubella, cervical cancer and tetanus are critical public health burden pathogens in the globe. Potentially, those diseases are capable of causing morbidity, mortality and disabilities for millions across the globe. However, their effects can be minimized or prevented through the implementation and expansion of Cost-Effective Program on Immunization (EPI). Annually, VPDs are responsible for 17% of global under-five mortality.

According to WHO 2017 report, 116.2 million infants (85%) received the third dose of Diphtheria, Pertussis and Tetanus (DPT). Worldwide, 123 countries reached the third dose (DPT3) coverage to 90%. Despite the increasing uptake of available vaccines, an estimated 19.9 million children under 12 months left unprotected with the DTP3 vaccine. Further enhancement of global immunization coverage would prevent an additional 1.5 million deaths annually. Information obtained from case-based surveillance indicates that the annual new cases of measles were estimated to be 29.1 cases per 1 million people.

Although the globe has made fundamental progress in reducing under-five mortality from 12.6 million deaths in 1990 to 5.4 million in 2017, it remains a critical public health problem. In 2017, an estimated 5.4 million children under the age of five died worldwide. This interprets into 15,000 deaths per day. Sub-Saharan Africa (SSA) continues to be the region that accounts for the highest under-five mortality rate (76 deaths per 1000 live births in 2017) in the world.

Ethiopia's under-five mortality rate has decreased from 87 to 55 deaths per 1,000 live births from 2005-2019, according to the 2019 Ethiopian Mini Demographic and Health Survey, indicating a general decline in childhood mortality rates. The WHO's expanded programme of immunization and the global alliance for vaccination and immunization have significantly improved global vaccination coverage against childhood infectious diseases, including polio and measles, but 6.6 million children still die annually due to infection.

Immunization is a crucial public health intervention that enhances the immune system, preventing illness, disability and deaths from childhood infectious diseases. Vaccination currently prevents 2-3 million annual deaths, with an additional 1.5 million deaths potentially prevented through vaccination. In 2018, 19.4 million infants worldwide were left unprotected by immunization services. Since 2010, global average vaccination coverage has increased by only 1%; only 86% of the children received their full course of routine vaccinations. In resource-limited countries, about 16% of under-five deaths are attributed to vaccine-preventable diseases. In Ethiopia, vaccine-preventable diseases such as pneumonia and rotavirus are the principal causes of under-five mortality. Despite the benefits above mentioned, of the total number of unvaccinated children, 60% lived in ten developing countries in which Ethiopia is not exceptional.

Ethiopia has replaced previous DPT vaccines with a penta-valent vaccine, including Haemophilus influenza type B and hepatitis-B antigens. Ethiopia has introduced various vaccines, including PCV, Rota and HPV, into its national EPI service. Full immunization coverage varies from 36.6% in Somalia to 100% in Addis Ababa, with factors like socio-demographic characteristics and health service delivery influencing it. Ethiopia was launched EPI in 1980, planning to achieve 100% coverage in 1990. The government of the Federal Democratic Republic of Ethiopia organized all stakeholders to achieve universal immunization coverage [1]. Likewise, the trend of the national immunization coverage in Ethiopia show an inflation from 14.3% in 2000 to 39% in 2016. This study aims to improve vaccination coverage in Ethiopia by assessing immunization status of children aged 12-23 months and identifying associated factors, utilizing data from the 2019 Ethiopian Mini Demographic and Health Survey (EMDHS). This will aid planners in designing evidence-based resource allocation and public health responses [2].

Data source

The data used in this study was secondary and extracted from the “2019 Ethiopia Mini Demographic and Health Survey, for short 2019 EMDHS”. The 2019 EMDHS was the second MDHS in Ethiopia, implemented by the Ethiopian Public Health Institute (EPHI), in partnership with the Central Statistical Agency (CSA) and the Federal Ministry of Health (FMoH), under the overall guidance of the Technical Working Group (TWG). This cross-sectional data collected (available on DHS website: https://dhsprogram.com/data/) from March 21, 2019, to June 28, 2019 provided up-to-date data on key demographic and health indicators across nine regions and two administrative cities, aimed at providing valuable insights into the health and demographic status of the population at that specific point in time.

Sampling design

Participants of the study were selected using a two-stage stratified cluster sampling method. Each region was divided into urban and rural areas, creating 21 sampling strata. Enumeration Areas (EAs) were independently selected within each stratum using Probability Proportional to Size (PPS) in the first stage. A total of 305 EAs (93 urban, 212 rural) were chosen based on the 2019 population and housing census. Household listings were conducted in these EAs and for large EAs, segments were created and one segment was selected using PPS. In the second stage, 30 households per cluster were systematically selected. All women aged 15-49 in these households, whether permanent residents or visitors who stayed the night before, were eligible for interviews. Additionally, height and weight measurements were collected from children aged 0-59 months. This method ensured precise and representative data collection across regions.

Eligibility criteria

The study extracted data on vaccination records for children aged 12-23 months from the EMDHS 2019 database, obtaining permission from Inner City Fund International, the owner of the raw data. The dataset included 5753 mothers with live children, with 1094 mothers with children aged 12-23 included in the final analysis [3].

Study variables

In order to address objectives mentioned, the study utilized two variable cohorts, dependent variable and independent variables.

Child immunization coverage/Outcome variable: As WHO recommended, basic childhood vaccines consists of polio, pentavalent (diphtheria, tetanus, pertussis, haemophiles influenza and hepatitis B vaccine), measles and Bacillus Calmette Guerin (BCG) that can prevent common childhood infections. Child immunization coverage refers to the proportion of children in a population who have received those recommended vaccines within a specified timeframe. It is a critical indicator of public health effectiveness in preventing vaccinepreventable diseases and ensuring community protection through herd immunity.

Child immunization coverage can be assessed in multiple ways. It can be binary, indicating whether a child has received all recommended vaccines (fully immunized) or not (partially or not immunized); alternatively, it can be measured as the total number of vaccines received by each child.

Thus, the dependent variable/outcome variable of this study is the total number of immunizations (total count of immunizations) received by children across 18 different vaccines (BCG, DPT1, DPT3, Polio 0-3, Measles 1-2, Pentavalent 1-3, Pneumococcal 1-3, Rotavirus 1-2). This involves summing up the counts for all vaccine types received by each child or group of children. This approach enables you to model and explore factors influencing the total number of immunizations received by children based on various predictors captured in the dataset.

The 2019 EMDHS report on vaccination coverage was gathered from immunization cards and mothers' verbal responses, with vaccination dates copied onto questionnaires when cards were available and recalls when not recorded.

Predictors: Though there are myriad of individual, household and community-level characteristics that are supposed to have an influence on the children’s vaccination coverage, this study, however, concentrates only on the most important and common sociodemographic, maternal and child-related characteristics, especially analyzed in most prior works and significant from theoretical ground, such as:

• Socio-demographic characteristics: Those refer to the social and demographic factors that describe the study population and are paramount important to understand patterns in health, education, economic status and other aspects of human behavior and wellbeing. Those include maternal age, mothers’ level of education, place of residence, family wealth index, religion, marital status, educational attainment of husband, etc.

• Maternal and children characteristics: Maternal characteristics include determinants, such as maternal ANC visit, utilization of family planning, parity and reproductive health history, like the place of delivery, received counselling from professionals during pregnancy and availability of vaccination and health cards; while children-related characteristics involves factors, such as children’s sex, children’s primary caregiver and children’s vital status.

Statistical model and analysis

Poisson regression model: The Poisson regression model stands out as a robust statistical approach for analyzing count data in epidemiological research, particularly in studying factors influencing child immunization coverage in Ethiopia, measured as the number of immunizations. This model is chosen for its suitability in handling count data and addressing issues like overdispersion, providing robust insights into factors influencing public health outcomes. Unlike linear regression, which assumes normality and constant variance, Poisson regression accommodates the inherently non-negative, discrete and often skewed nature of count data. This makes it an appropriate choice for modeling outcomes like immunization counts, where values are strictly non-negative integers.

The whole analysis was done using R version 4.4.1. To adjust the non-proportional allocation of the Sample to different regions and their rural and urban areas, weighting was applied. So, the representativeness of the survey results both at the national and regional levels is guaranteed. Both exploratory (descriptive statistics) and confirmatory (inferential statistics) data were adopted to visualize, characterize and infer the immunization data of children aged 12 to 23 months. Frequencies and percentages were used to describe the categorical variables, mean and standard deviation for quantitative data; and graphs such as geospatial maps, heatmap, histogram, QQplot, etc. to visualize the data. Finally, we presented results using tables. Predictors with a p-value less than 0.05 in the bivariate analysis were included in the final regression analysis.

Socio-demographic characteristics

Table 1 reveals that immunization coverage varies significantly across a number of demographic information in the study. The majority of respondents are young parents, with the 25-29 age group being the largest (1858, 32.3%), followed by the 20-24 (1143, 19.9%) and 30-34 (1239, 21.5%) age groups. The 35-39 age group accounts for 765 respondents (13.3%), while the 15-19, 40-44, and 45-49 age groups represent smaller proportions at 296 (5.1%), 339 (5.9%) and 113 (2%) respectively. Oromia has the highest population of respondents (719, 12.5%), followed by Afar (652, 11.3%), SNNPR (660, 11.5%) and Somali (637, 11.1%). Regions like Benishangul (530, 9.2%), Amhara (511, 8.9%) and Tigray (454, 7.9%) have moderate populations. Lower populations are seen in Gambela (450, 7.8%), Harari (447, 7.8%) and Dire Dawa (402, 7%), with Addis Ababa having the smallest (291, 5.1%). The majority of respondents are Muslim (2974, 51.7%), followed by Orthodox Christians (1612, 28%) and Protestants (1053, 18.3%). A small proportion (114, 2%) adheres to other religions. Most respondents live in rural areas (4425, 76.9%), with a smaller proportion residing in urban areas (1328, 23.1%). A significant portion of respondents have no education (3149, 54.7%), while 1823 (31.7%) have primary education, 480 (8.3%) have secondary education and 301 (5.2%) have higher education. The wealth distribution shows that 1964 respondents (34.1%) are in the poorest category, followed by poorer (994, 17.3%), middle (805, 14%), richer (738, 12.8%) and the richest (1252, 21.8%). Most households are headed by males (4598, 79.9%), with 1155 (20.1%) headed by females. Slightly more respondents do not have access to clean water (2923, 50.8%) compared to those who do (2830, 49.2%) [4].

Table 1: Socio-demographic and other characteristics of children’s immunization status, mini EDHS.

Family sizes are predominantly 4-6 children (2871, 49.9%), followed by 7-9 children (1842, 32%), 0-3 children (638, 11.1%) and more than 9 children (402, 7%). Most households have 1-2 children under five (2079, 36.1% and 2504, 43.5% respectively), with smaller proportions having 3 (794, 13.8%), 4 (126, 2.2%) or 5 children (32, 0.6%). A small number of households (218, 3.8%) have no children under five. The vast majority of respondents (5599, 97.3%) do not have access to a landline telephone, with only 154 (2.7%) having access. The sex distribution of children is fairly even, with slightly more males (2969, 51.6%) than females (2784, 48.4%). Most children are alive (5414, 94.1%), with a small proportion deceased (339, 5.9%). The majority of children live with the respondent (5293, 92%), while a small number live elsewhere (121, 2.1%). Half of the deliveries occur at home (2881, 50.1%), with significant numbers also in public health facilities (2563, 44.6%). Deliveries in private health facilities (157, 2.7%) and NGO and other health facilities (152, 2.6%) are less common [5,6].

Regarding Antenatal Care (ANC) visits, 35% had seven or more visits (2011), 14.7% had four visits (847), 13.3% had three visits (768) and 6.1% had two visits (353). A smaller number had one visit (141, 2.5%) or no visits at all (1044, 18.1%). A significant portion of respondents (2098, 36.5%) received counselling during pregnancy, while 837 (14.5%) did not.

Family planning usage is relatively low, with 67.3% (3874) not using any method and 32.7% (1879) using family planning. Most respondents have 0-3 children (3046, 52.9%), followed by 4-6 children (2032, 35.3%) and 7 or more children (675, 11.7%).

Prevalence and immunization coverage counts by region

The following analysis examines the distribution of total immunization coverage counts across various regions in Ethiopia. The data captures the number of immunizations received by children, ranging from zero to eighteen doses, as per the Ethiopian Expanded Program on Immunization (EPI) guidelines. This analysis will explore patterns and disparities in immunization coverage by region.

The analysis of immunization coverage across various regions in Ethiopia in Table 2 reveals significant disparities, with some regions showing high non-immunization rates and others demonstrating substantial full immunization achievements. For instance, Afar and Somali have high numbers of children receiving no immunizations (489 each), while regions like Addis Ababa and Dire Dawa have more children achieving full immunization (90 and 64, respectively). This disparity indicates a need for targeted interventions in regions with low coverage to improve immunization rates and ensure equitable health outcomes for all children.

Table 2: Immunization coverage counts by region.

Table 3 shows that the prevalence rates for each vaccination type, reflecting the extent to which children in the surveyed population have been immunized. Among 5752 children analyzed, significant proportions received key vaccinations: DPT vaccines, critical for protection against diphtheria, pertussis and tetanus, had a prevalence rate of 35.2%. Similarly, polio vaccination, essential in preventing paralysis from poliovirus, achieved a prevalence rate of 38.3%, with 2205 children immunized. Measles vaccination, crucial for averting outbreaks of this highly contagious disease with potential severe complications, reached a prevalence rate of 22.3% among 1282 children.

Table 3: Prevalence of immunization.

Pentavalent vaccines, safeguarding against five diseases including diphtheria, tetanus, pertussis, hepatitis B and Haemophiles influenzae type b (Hib), achieved a prevalence rate of 35.2% among 2026 children. Pneumococcal vaccines, targeting Streptococcus pneumoniae bacteria and preventing pneumonia and related illnesses, were administered to 1944 children, resulting in a prevalence rate of 33.8%. Meanwhile, rotavirus vaccines, crucial in preventing severe diarrhea and a leading cause of death among young children worldwide, had a prevalence rate of 33.3% with 1916 children immunized. BCG vaccines, protecting against Tuberculosis (TB) and particularly severe forms in children, achieved a prevalence rate of 34.4% with 1980 children vaccinated. These figures underscore both achievements and areas for further enhancement in childhood immunization coverage across different regions and demographics.

The pie chart in Figure 1 illustrates the prevalence of various vaccinations among children in Ethiopia based on the Mini-EDHS 2019 data. The highest coverage is observed for BCG vaccination at 34.42%. Pneumococcal and DPT series show substantial coverage with Pneumococcal 1, 2 and 3 at 33.80%, 30.02% and 31.24% respectively and DPT 1, 2 and 3 at 35.17%, 31.52% and 25.90%. Polio vaccinations also have notable prevalence with Polio 0, 1, 2, and 3 at 19.63%, 32.28%, 32.58% and 35.80%. The Pentavalent series has good coverage with Pentavalent 1, 2 and 3 at 35.17%, 31.52% and 25.70%. Rotavirus vaccinations are lower with Rotavirus 1 at 33.31% and Rotavirus 2 at 28.69%, while Measles vaccination shows the lowest coverage with Measles 1 at 22.29% and Measles 2 at 3.74%. Overall, the chart reveals high initial vaccination rates but a drop in coverage for subsequent doses, indicating challenges in completing vaccination series.

Figure 1: Prevalence of various vaccinations, Mini-EDHS-2019.

Figure 2 is a heatmap showing the immunization coverage across various regions in Ethiopia. The plot employs a color gradient from light blue to dark blue to denote different levels of coverage, with light blue indicating lower coverage and dark blue indicating higher coverage. Each region is labeled at its centroid, facilitating the identification of corresponding coverage values. The legend on the right provides a gradient scale ranging from approximately 300 to 700, enabling quick interpretation of coverage levels across the regions.

Figure 2: Immunization coverage heatmap in Ethiopia.

In interpreting the plot, regions like Oromia and SNNP are observed to have higher immunization coverage, as reflected by the darker blue shades and coverage values near 700. Regions such as Amhara and Afar exhibit medium coverage, indicated by intermediate shades of blue. In contrast, regions like Addis Ababa and Gambella display lower coverage, with lighter blue shades and coverage values closer to 300. This visualization effectively highlights areas with varying immunization coverage, indicating where additional resources or efforts might be needed to improve vaccination rates.

Parameter estimates from poisson regression

Sociodemographic characteristics: The finding from the Poisson regression model analysis is recapitulated in Table 4. In this model, regions such as Afar, Amhara, Oromia, Somali, Benishangul, SNNPR, Gambela, Harari, Addis Ababa and Dire Dawa show negative coefficients, indicating lower immunization counts compared to an unspecified reference category. These coefficients are statistically significant with p-values suggesting significant impacts on immunization coverage. Particularly, Afar stands out with a coefficient of -0.536 (SE=0.041, p<0.001), indicating significantly lower child immunization coverage compared to other regions in the study. In contrary, religion such as Muslim exhibits a coefficient of 0.0597 (SE=0.0195, p=0.002), indicating followers of the Muslim religion tend to have slightly higher immunization counts compared to an unspecified reference category (Orthodox). Conversely, protestant and other region follower do not show statistical significance (p>0.1), suggesting no significant association with immunization counts in this study. The analysis also reveals, type of residence such as rural has a coefficient of -0.0611 (SE=0.023, p=0.008), indicating that individuals residing in rural areas tend to have slightly lower immunization counts compared to those in urban areas, with statistical significance.

Table 4 also indicates that the wealth index categories are significantly associated with immunization coverage counts. Specifically, poorer household (0.1653, SE=0.026, p<0.001), households with middle income (0.1538, SE=0.028, p<0.001), those who are rich (0.2293, SE=0.029, p<0.001) and richest (0.2265, SE=0.034, p<0.001) all show higher immunization coverage counts compared to the reference category (those who are poorest economically). Additionally, households headed by females have higher immunization coverage counts (0.0496, SE=0.018, p=0.005), highlighting the positive impact of female household heads on immunization rates. Unlike to income, household head age has a coefficient of -0.0010 (SE=0.0006, p=0.077) and birth order has a coefficient of -0.0119 (SE=0.0072, p=0.101), neither of which show strong significance [7].

Table 4: Poisson regression analysis of socio-demographic, maternal and child determinants of child vaccination in Ethiopia, 2019.

Maternal and child characteristics: The result also indicates mothers with secondary level shows a coefficient of 0.0629 (SE=0.024, p=0.010), suggesting higher immunization counts compared to the reference category (no education). However, those with primary and higher are not statistically significant (p>0.1), indicating no significant impact on immunization counts. It also shows that families with 4-6 children have a coefficient of 0.0593 (SE=0.021, p=0.004), indicating higher immunization coverage counts compared to the reference category. Families with 7-9 children and those with more than 9 children do not show statistical significance (p>0.1), suggesting no significant impact on immunization coverage counts.

It is also stated that female children have lower immunization coverage counts with a coefficient of -0.0446 (SE=0.0133, p<0.001). Place of delivery at public health facilities (0.1383, SE=0.0185, p<0.001) and NGO and other health facilities (0.2109, SE=0.0432, p<0.001) are associated with higher counts, whereas private facilities are not significant (p>0.1). Increased maternal ANC visits positively correlate with higher immunization coverage, with coefficients rising from 0.1395 (SE=0.0449, p=0.002) for 2 visits to 0.3529 (SE=0.0438, p<0.001) for 5 visits, and remaining significant for 6 or more visits. Child age shows a strong negative association (-0.0262, SE=0.0005, p<0.001). Pregnancy counselling (0.0713, SE=0.0162, p<0.001) and family planning (0.1971, SE=0.0143, *p<0.001) are positively associated with higher immunization counts. Lastly, maternal parity for 4-6 children (0.0934, SE=0.0266, p<0.001) and 7 or more children (0.1337, SE=0.0512, p=0.009) also shows a positive correlation. The analysis of deviance table (Table 5) for the Poisson regression model with a log link function and response variable, “immunization counts” sequentially adds predictor terms to assess their impact on model fit. Initially, the null model with no predictors has a deviance of 25,713 on 2,774 degrees of freedom. As seen in Table 5, there is a significant improvement in fit from the null model, with a decrease in deviance from 25,713 to 20,409 on degrees of freedom of 2,774 and 2,730, respectively. This indicates that including predictors enhances the model's explanatory power. The AIC value of 27,960 suggests reasonable fit, balancing model complexity and goodness of fit. Region, residence type, education category, clean water access, sanitation facilities, wealth index, place of delivery, maternal ANC visits, child age, pregnancy counselling, family planning and maternal parity all exhibit highly significant effects (p<0.001). Specifically, child age has the largest deviance reduction, indicating its strong impact on immunization counts. In contrast, variables such as household head sex and child sex do not show significant contributions to the model (p>0.1). Overall, the model provides insights into how these predictors, key socio-demographic, maternal and child-related contribute to understanding immunization coverage [8].

Table 5: Analysis of deviance table model: Poisson, link: log.

Model adequacy assessment

The overdispersion statistic for the Poisson regression model was determined and found to be 0.950, indicating that the model does not exhibit overdispersion. This means that the observed variance of the response variable "Immunizations" is close to what is expected under the Poisson distribution assumption. The model's fit is adequate in explaining the variability in the data without significant excess variability beyond what the Poisson model accounts for.

The model adequacy assessment reveals robust findings, suggesting that the model adequately explains the variability in the data without significant overdispersion: Metrics, deviance and pearson Chi-square ratios, confirm a good fit. A deviance ratio (0.997) close to 1 indicates a good fit (model fits well). Similarly, a ratio close to 1 (0.950) also suggests a good fit. The Variance Inflation Factor (VIF) values indicate no multi-collinearity issues among predictors, with values generally close to 1. This suggests that multi-collinearity is not a significant issue among the predictors. Specifically, variables such as region, religion, education level, household wealth and child caregiver have VIFs around 1.05, indicating low to moderate multi-collinearity; and the remaining variables, namely age of mother, ANC utilization, place of delivery, pregnancy counselling, birth order and sex of the child have VIFs around 1.02, indicating minimal multi-collinearity. These low VIF values suggest that each predictor is largely independent of the others in the model.

As depicted in the Figures 3 and 4, residuals analysis shows no abnormal patterns. The Pearson and deviance residual plots should exhibit randomness without discernible patterns, while the Q-Q plot should show residuals aligning closely with a straight line if the assumption of normality is met. There is also no significant zeroinflation detected, collectively supporting the model's appropriateness for the data. Given these diagnostics, the Poisson regression model appears to be an appropriate choice for analyzing the immunization counts data compared to Negative Binomial or Quasi-Poisson regression since no overdispersion were detected [9].

Figure 3: Histogram of residuals.

Figure 4: Plot of standardized residual vs. fitted model.

The plots of Pearson and deviance residuals, as well as a Q-Q plot are visualized in Figures 5-7. These plots are essential for assessing model adequacy: The Pearson and deviance residual plots should exhibit randomness without discernible patterns, while the Q-Q plot should show residuals aligning closely with a straight line if the assumption of normality is met. These diagnostics help ensure that the regression model assumptions hold and provide insights into potential outliers or influential observations.

Figure 5: Plot of Pearson residuals. Figure 6: Q-Q plot of Pearson residuals. Figure 7: Plot of deviance residuals.

Figure 6: Q-Q plot of Pearson residuals.

Figure 7: Plot of deviance residuals.

This detailed demographic information highlights the diverse backgrounds of the respondents, indicating varying needs and challenges in achieving high immunization coverage across different regions, religions, education levels and socio-economic statuses. Tailored strategies are essential to address these diverse needs and ensure effective immunization cover.

The overall finding indicates that immunization coverage in Ethiopia was about 40%, which includes various vaccines like BCG, DPT, Polio and Measles. The highest coverage is observed for BCG vaccination at 34.42%. Pneumococcal and DPT series show substantial coverage with Pneumococcal 1, 2 and 3 at 33.80%, 30.02% and 31.24% respectively, and DPT 1, 2 and 3 at 35.17%, 31.52% and 25.90%. Polio vaccinations also have notable prevalence with Polio 0, 1, 2 and 3 at 19.63%, 32.28%, 32.58% and 35.80%. The Pentavalent series has good coverage with Pentavalent 1, 2 and 3 at 35.17%, 31.52% and 25.70%. Rotavirus vaccinations are lower with Rotavirus 1 at 33.31% and Rotavirus 2 at 28.69%, while Measles vaccination shows the lowest coverage with Measles 1 at 22.29% and Measles 2 at 3.74%. Thus, though there were high initial vaccination rates, it had shown to drop in coverage for subsequent doses, indicating challenges in completing vaccination series.

According to the EDHS 2016, the coverage for the third dose of the pentavalent vaccine (DPT-HepB-Hib3) was about 53% and the coverage for the first dose of the measles vaccine was 54%. The 2020 WHO/UNICEF Estimates of National Immunization Coverage (WUENIC) data estimated that the coverage for the third dose of DPT was around 63% and for measles, it was around 61%. In the African region, the WHO reported that the average DPT3 coverage was around 76% in 2020, with measles coverage (first dose) around 70%. Specifically, in 2020, Nigeria had a DPT3 coverage of around 57% and measles coverage of 54%; Kenya had a DPT3 coverage of about 86% and measles coverage of 83%; Uganda had a DPT3 coverage of around 89% and measles coverage of 95%.

A study conducted in 2020 in the Oromia region of Ethiopia reported DPT3 coverage of 62% and measles coverage of 57%. The African Union’s Africa CDC reported in 2022 that the average vaccination coverage across Africa for DPT3 was around 72% and for measles was around 69%. According to the Global Vaccine Action Plan (GVAP) 2020 progress report, Africa has been lagging behind global immunization targets, with several countries struggling to reach 80% coverage for basic vaccines.

In summary, immunization coverage in Ethiopia has been reported at about 40% in the Mini-EDHS 2019. Other sources, like EDHS 2016 and WUENIC 2020, show a slightly higher coverage rates, indicating some improvement but still below the African average. The overall immunization coverage in Africa is higher compared to Ethiopia, with an average of around 75-76% for DPT3 and 70% for measles. However, there is significant variability among different countries within Africa. In conclusion, while Ethiopia's immunization coverage has seen improvements, it remains lower compared to the average coverage in the African region. Efforts to increase immunization rates in Ethiopia are crucial to meet global health targets and ensure better health outcomes for children.

The Poisson regression model analysis reveals several significant factors influencing immunization coverage among children in Ethiopia. Regional effects indicate that children in Afar have immunization counts that decrease by approximately 53.6% compared to the reference category. Other regions also show decreases in immunization counts: Amhara by 33.8%, Oromia by 30.2%, Somali by 34.2%, Benishangul by 32.7%, SNNPR by 35.1%, Gambela by 36.3%, Harari by 31.4%, Addis Ababa by 37.2% and Dire Dawa by 35.5%.

In terms of religion, Muslim children have immunization counts that increase by approximately 6% compared to Orthodox children, while Protestant and other religious affiliations show no significant changes. Children living in rural areas experience a decrease in immunization counts by about 6.1% compared to those in urban areas.

The wealth index reveals that children from poorer households see an increase in immunization counts by 16.5%, middle-income households by 15.4%, rich households by 22.9% and the richest households by 22.7%, all compared to the poorest households. Female-headed households have higher immunization counts, increasing by approximately 5% compared to male-headed households. Maternal education shows that mothers with secondary education have an increase in immunization counts by 6.3% compared to mothers with no education, while primary and higher education show no significant changes.

Family size impacts immunization coverage, with families having 4-6 children seeing an increase by 5.9% compared to smaller families, whereas larger family sizes show no significant changes. Female children have immunization counts that decrease by about 4.5% compared to male children. The place of delivery also matters, as deliveries at public health facilities lead to an increase in immunization counts by 13.8%, and deliveries at NGO and other health facilities result in a 21.1% increase, while private facilities show no significant impact.

Maternal ANC visits positively correlate with immunization coverage, with immunization counts increasing by 14% for two visits and 35.3% for five visits, with significant effects for six or more visits as well. Child age is negatively associated with immunization coverage, with each unit increase in age leading to a 2.6% decrease in immunization counts. Pregnancy counseling and family planning are both positively associated with higher immunization counts, with increases of 7.1% and 19.7%, respectively. Lastly, maternal parity shows that families with 4-6 children experience a 9.3% increase in immunization counts, and those with seven or more children see a 13.4% increase.

Full immunization coverage is below national targets, negatively correlated with early maternal age, female household head, low maternal education and rural housing, while positively associated with prenatal counselling and delivery at public health facilities highlighting the need for targeted strategies to address regional disparities and socio-economic barriers.

Ethical clearance and DHS data set authorization letter of approval with reference number (155878) were obtained from the DHS program data set admin and the Inner City Fund (ICF) international. This was done after being registered and sending the concept note of this study through the website in order to get permission from Inner City Fund (ICF) International to access and use the dataset. After obtaining permission, the dataset was downloaded from the website at http://www.DHS program.com. Data confidentiality was maintained throughout our analysis and then after.

The authors would like to acknowledge the DHS program data set admin and the Inner City Fund (ICF) International for providing us with all the relevant secondary data used in this study. Finally, we would like to thank all who directly or indirectly supported us.

GG1 conceptualized and designed the study, performed the analysis and drafted the manuscript. MT1, HG, SH, GA1 and GA2 validated the analysis and provided significant input in reviewing the study's design and the draft manuscript. MK, MA, GY, GG2 and HK conducted a thorough review of the manuscript for important intellectual content and contributed to the final approval of the version intended for submission.

This study has no received any funds.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

1. Truche P, Botelho F, Bowder AN, Levis AW, Greenberg SL, et al. (2021) . World J Surg 45: 2643-2652.

   [] [] []

2. Guthold R, Baltag V, Katwan E, Lopez G, Diaz T, et al. (2021) Easy access to the latest global, regional and national adolescent health data: the world health organization maternal, newborn, child and adolescent health and ageing data portal. J Adolescent Health 68: 243-245.

   []

3. Teferi E (2018) Factors influencing coverage and key challenges to achieving targets of routine immunization in Africa: A systematic review. Ethiop J Pediatr Child Health 13.

   []

4. Masresha B, Luce R, Shibeshi M, Katsande R, Fall A, et al. (2018) Status of measles elimination in eleven countries with high routine immunisation coverage in the WHO African region. J Immunol Sci 28: 140-144.

   [] []

5. Rutherford ME, Mulholland K, Hill PC (2010) . Trop Med Int Health 15: 508-519.

   [] [] []

6. Faruk MO, Arafat ME, Shanta SH (2023) . PLoS One 18: e0288022.

   [] [] []

7. Greenwood B (2014) . Philos Trans R Soc Lond B Biol Sci 369: 20130433.

   [] [] []

8. Laupeze B, Del Giudice G, Doherty MT, van der Most R (2021) . NPJ Vaccines 6: 93.

   [] [] []

9. Remy V, Zollner Y, Heckmann U (2015) J Mark Access Health Policy 3: 27041.

   [] [] []