Non-uptake of dual protective polio vaccine and its determinants among children in Ethiopia using Ethiopian Demographic Health Survey 2019: a mixed-effect model

Abstract

Introduction

The polio vaccine is the live-attenuated antigen that prevents poliomyelitis. According to a report by the WHO, about 1 million less than 5-year-old children missed the polio vaccination from 2018 to 2021. Even though Ethiopia is the most prioritised country for polio eradication, there is not enough evidence about the combined oral and inactivated vaccine in Ethiopia.

Objective

To assess the non-uptake of the dual protective polio vaccine and its determinants among children in Ethiopia using the Ethiopian Demographic Health Survey (EDHS) 2019.

Methods

The secondary data analysis of a community-based cross-sectional study was conducted using EDHS 2019 data among 3094 participants. Mixed-effects binary logistic regression was used for descriptive analysis and identifying the predictors using a p value of <0.05. Intraclass correlation was used to assess the clustering effect.

Results

The prevalence of non-uptake of the dual protective polio vaccine in Ethiopia was 44% (95% CI 42.2% to 45.8%). Predictors like women with low proportions of community media exposure (adjusted OR (AOR)=2.3, 95% CI 1.8 to 2.8) and no history of antenatal care visits (AOR=2.3, 95% CI 3.89 to 6.35) were significantly associated with non-uptake of the dual protective polio vaccine.

Conclusion

The burden of non-uptake of the dual protective polio vaccine in Ethiopia was still high. Low community media exposure and no antenatal care utilisation were the predictors of non-uptake of the dual protective polio vaccine. We recommend improving media access and antenatal care services to expand polio vaccination.

Correspondence to Muluken Chanie Agimas; [email protected]

Strengths and limitations of this study

• The strength of the study is that we use national-level data to generate conclusive findings and empower the power of the findings.

• But because of recall bias, there may be a risk of non-differential misclassification.

• The current study shares all the limitations of a cross-sectional study.

• Additionally, because of secondary data, other potentially important variables were not considered for analysis.

Introduction

The polio vaccine is the live-attenuated antigen that prevents poliomyelitis. Administering both oral poli vaccine (OPV) and inactivated polio vaccine (IPV) can protect the poliovirus more than a single vaccine.¹ In the absence of the polio vaccine, poliomyelitis is the major threat for under-5 children, causing irreversible paralysis of the lower extremities.² According to a report by the WHO, about 1 million less than 5-year-old children missed polio vaccination from 2018 to 2021.³ Globally, there are 52 poliovirus-epidemic countries and 185 non-endemic countries.⁴ To reverse the poliomyelitis-related disability, immunisation with the polio vaccine is the best strategy; otherwise, if one child is infected with the poliovirus, children around the globe are at risk of poliomyelitis.⁵

To achieve the global strategic goal of eradicating the poliovirus as a programmatic emergency for global public health, the WHO recommended the combined OPV and IPV programme. IPV is a systemic immunity booster that avoids the paralytic and circulating vaccine-derived polioviruses that exist with OPV.⁶ The WHO reported that the number of polio cases had decreased from 256 in 2013 to 171 in 2014 to 33 in 2018.⁷

The distribution of wild poliovirus has been highly reduced by the Global Polio Eradication Initiative.⁸ Nowadays, types 2 and 3 wild poliovirus strains have been zero globally. However, the type 1 wild poliovirus vaccine is endemic in 33 African countries, including Ethiopia.⁹ In Africa, the wild poliovirus has been prevented, but because of low vaccine coverage, circulating vaccine-derived poliovirus is the major challenge.¹⁰ To reduce this problem, a dual, or both, OPV and IPV is the pillar for eradicating the polio virus disease, or the combined vaccine has profound effectiveness in protecting against poliovirus.^{11 12} In 2015, Ethiopia introduced the IPV, and it was administered along with OPV-3 at 14 weeks.¹³

As the 2019 Ethiopian Demographic Health Survey (EDHS) reported, the vaccine coverage of polio vaccine among Ethiopian children was 32% OPV-0, 60% OPV-3 and 55% IPV, respectively.¹⁴ This suggested that the coverage of polio vaccination is low compared with the global target (85%).¹ Because of low access to health services and low socioeconomic and health-seeking behaviour, vaccination practices vary across the country.^15–19 Because the wild poliovirus is imported from Kenya, Sudan, Somalia and Djibouti, Ethiopia is the prioritised African country in the global polio eradication initiative.²⁰ The living style of Ethiopia is highly dispersed with a large population density, and in the low population density area, people are highly mobile. This causes low vaccination coverage in Ethiopia. Even though Ethiopia is the most prioritised country for polio eradication using the dual protection polio vaccine, there is not enough evidence or information about the combined OPV and IPV.¹ Therefore, assessing the prevalence of non-uptake of the dual protective polio vaccine and identifying the possible predictors is very important to improve the dual protective polio vaccination coverage in Ethiopia. However, there is not enough conclusive evidence about the dual protective polio vaccine in Ethiopia. Therefore, the current study aimed to assess the prevalence and its determinants of non-uptake of the dual protective polio vaccine among children in Ethiopia using the EDHS 2019 data set to provide an important evidence-based practice of polio eradication initiatives.

Objective

General objective

• To assess the prevalence and its determinants of non-uptake of the dual protective polio vaccine among children in Ethiopia using the EDHS 2019 data set.

Specific objectives

• To determine the prevalence of non-uptake of the dual protective polio vaccine among children in Ethiopia using the EDHS 2019 data set.

• To identify the determinants of non-uptake of the dual protective polio vaccine among children in Ethiopia using the EDHS 2019 data set

Methods

Study area, design and period

This country-level community-based cross-sectional study was used in Ethiopia using the secondary data set of mini-EDHS 2019 from 21 March to 28 June 2019. Ethiopia comprises a total of 12 regions. These are Afar, Oromia, Amhara, south Ethiopia, south-west Ethiopia, central Ethiopia, Harari, Tigray, Somalia, Benishangul Gumuz, Gambela and Sidama, and it has two city administrations, namely Addis Ababa and Dire Dawa. Geographically, the regions of Ethiopia are subdivided into 68 zones, 817 Woreda and about 16 253 kebele.²¹ In 2020, Ethiopia will have a total population of 114 963 588.²² The mini-EDHS 2019 data was aimed at evaluating the Growth and Transformation Plan of Ethiopia.

Population and eligibility criteria

All women who had a child were the source population, and women who had a child aged 12–35 months in the enumeration area were the study population. All women who had a child aged 12–35 months earlier in the survey were included in the study.

Patient and public involvement

We used the secondary analysis of the EDHS 2019 data set, and hence patient and public involvement were not applicable.

Sample size and techniques

First, the mini-EDHS 2019 was applied using the sampling weight method to attain a representative sample.²³ Again for the purpose of the current study sample size estimation, we used the sampling weight technique, and thus a total sample of 3094 was used for the analysis. The reason for applying the sampling weight technique was because the size of the illegible population varies across regions and residences. The stratified method was applied to stratify the regions into urban and rural, with a total of 21 strata. Then two-stage sampling was applied, such as first samples of enumerations being selected from urban and rural areas. Second, from the selected enumeration areas, households were selected by multistage sampling technique. From these households, women having a child aged 12–35 months during the survey were selected and interviewed.²³

Sampling weight techniques of EDHS

The mini-EDHS 2019 used the following sampling weighting techniques to get a representative sample. These are:

• Weighting for households (hv005)=the reciprocal of a households*the rate of the response in the household.²³

• Weighting for the men’s subsample=the reciprocal of a household in the subsample*the rate of the response in the subsample household.²³

• Individual weight for women=the household*the rate of the individual response for the women in the stratum.²³

• Men’s individual sampling weight=the men’s subsample weight*the reciprocal of men’s response rate.²³

Study variable

Dependent variable

Non-uptake of the dual protective polio vaccine (yes, no).

Independent variables

Residence, place of delivery, antenatal care visit (yes/no), frequency of antenatal care (ANC), wealth index, educational status, media exposure, women’s age, community educational status, community media exposure, region and numbers of children. Such independent variables were selected based on their availability in the data set, plausibility and review of the previous evidence/articles.

Operational definition

Non-uptake of the dual protective polio vaccine

The dependent variable was classified as ‘yes’ or ‘no’. Children who did not uptake both OPV-3 and IPV were classified as having no dual protection; otherwise, they were categorised as having uptake of the dual protective polio vaccine.¹

Media exposure

Media exposure was measured based on whether women had access to read newsletters, listen to the radio and watch television. Accordingly, if they have access to all three media (newsletter, radio and television) at least once a week, we categorised them as ‘yes’, otherwise, ‘no’.²⁴

Community educational status

Community educational status is computed as the proportion of educated community achievement in the community, as a high proportion and a low proportion.²⁵

Community media exposure

Community media exposure is computed based on the individual’s reply to experience, with radio and TV in each cluster as a high proportion and a low proportion of media exposure.²⁵

Data source, collection and tool

Initially, the data set was requested online from the official demographic health survey website at www.measuredhs.com, and approval to access the data was obtained after 2 working days. The EDHS used the interview-administered questionnaire among women who had children aged 12–35 months. The data quality was assured by pretesting, training of data collectors and supervision.²³

Data processing and analysis

After accessing the data from DHS International, cleaning, recoding, sampling weight and missing data checking were conducted using STATA software V.14. After this, mixed-effects binary logistic regression was used to identify the possible determinants for non-uptake of the dual protective polio vaccine. The reason we used such a model was because of the hierarchical nature of the EDHS data and the possibility of considering a natural nesting of data. We built models like the null model (a model with no predictors), model I (level one predictors), model II (a model with level two predictors) and model III (mixed-effect model). The mixed-effect model is:

Logit (Yij) = β0j + ∑βXi + ϒZj + εj, where β0j = β0 + μj, μj∼N (0, σ2 u).²⁶

where

εj = ε0 + εj, εj∼N (0, σ2 ε)

Logit (Yij) = ln(Yij/(1 - Yij)) or log odds of non-uptake of dual protective polio vaccine.

(Yij) = the probability of non-uptake of dual protective polio vaccine for women ‘i’

in the enumeration area and residence of rural and urban region ‘j’.

β0j = random intercept of the cluster ‘j’.

‘εj’ = residual for each cluster ‘j’.

‘β’ = the fixed-effect regression coefficient.

‘Xi’ = level one predictor.

‘ϒZj’ = level II predictor for clusters.

To test the clustering effect, the intraclass correlation coefficient (ICC) was used with a cut-off of >0.05 (>5%). For each model, ICC ((ρ) = σ² ε/σ² ε +σ²µ),²⁷ proportional change in variance (=variance of the null model minus variance of the next model/variance in the null model*100)²⁷ and Akaike information criteria (AIC=2k-2lnL, where k is the number of parameters, and L is the maximum value of the likelihood function of the model) were calculated. Then the best model was selected based on the lowest AIC value (table 1). The significant predictors were selected using a p value of <0.05. To estimate the effect measure of each predictor, an adjusted OR with a 95% confidence level was used.

Table 1

Mixed-effect binary logistic regression model of individual- and community-level factors predicting non-uptake of dual protective polio vaccine in Ethiopia using the Ethiopian Demographic Health Survey 2019 data

Random effect	Null model	Model I	Model II	Model III
Variance	0.78	0.3	0.47	0.19
Intraclass coefficient	19.1%	15.1%	13.3%	3.8%
Proportional change in variance (%)	Reference	61.5%	40%	76%
Log likelihood	−542.3	−428.1	−512	−402
Akaike information criteria	1088.7	872.2	1035	827

Results

In this study, 3094 women who had children aged 12–35 months prior to the survey were included and interviewed about the non-uptake of the dual protective polio vaccine. About 1669 (54%) women were delivered to their homes. Additionally, 2381 (84%) women had media exposure, and 2629 (85%) of the women had a history of ANC visits. Furthermore, the majority of the participants, 2299 (74.3%), were from rural residences (table 2).

Table 2

Characteristics of the women who had children aged 12–35 months in Ethiopia using 2019 Ethiopian Demographic Health Survey data

Variables	Category	Weighted frequency	Per cent (%)
Women age	15–24	857	27.7
	25–34	1636	52.9
	≥35	601	19.4
Media exposure	Yes	443	16
	No	2381	84
Education status	No education	1532	49.5
	Primary school	1161	37.5
	Secondary	268	8.7
	Higher education	132	4.3
	Diredawa	17	0.5
Region	Tigray	213	6.9
	Afar	49	1.6
	Amhara	614	19.8
	Oromia	1236	40
	Somalia	201	6.5
	Benishangul	37	1.2
	SNNPR	609	19.7
	Gambela	14	0.4
	Tigre	60	11.1
	Others*	30	5.50
Place of delivery	Health facility	1425	46
	Home	1669	54
Community media exposure	High proportion exposure	1962	63.4
	Low proportion exposure	1132	36.6
Community education	High proportion educated	1188	38.4
	Low proportion educated	1916	61.6
Residence	Urban	795	25.7
	Rural	2299	74.3
ANC utilisation	Yes	2629	85
	No	465	15
Frequency of ANC visit	<4	2925	94.5
	≥4	169	5.5
	Poorest	693	22.4
Wealth index	Poorer	662	21.4
	Middle	598	19.3
	Richer	513	16.6
Numbers of live birth	Richest	628	20.3
	1–2	1317	42.6
	3–4	859	27.8
	>4	918	29.6

*Statistically significant variable.

ANC, antenatal care.

Prevalence of non-uptake of dual protective polio vaccine

According to EDHS 2019, the prevalence of non-uptake of the dual protective polio vaccine was 44% (95% CI 42.2% to 45.8%).

Determinants of non-uptake of dual protective polio vaccine

Models like the null model, a model with a level one predictor (model I), a level two/community-level model (model II) and a mixed-effect model (model III) were built. To select the best model from those models, the AIC for each model was calculated, and the model with the lowest AIC was the best model (table 1). Hence, the best model was the mixed-effect (model III) model. Bivariable mixed-effect binary logistic regression was used to select the possible candidate variable for multivariable mixed-effect binary logistic regression with a p value of less than 0.25. Based on this, ANC visit, place of delivery, community education, community media exposure, residence and wealth index were the candidate predictors for multivariable mixed-effect binary logistic regression. Finally, in multivariable mixed-effect binary logistic regression analysis, ANC follow-up and community media exposure were the predictors of non-uptake of the dual protective polio vaccine (table 3). Thus, women from low proportions of community media exposure were 2.3 (adjusted OR (AOR)=2.3, 95% CI 1.8 to 2.8) times more likely to non-uptake of the dual protective polio vaccine than those from high proportions of community media exposure. Additionally, women who had no history of ANC visits were 5.12 (AOR=5.12, 95% CI 3.89 to 6.35) times more likely to non-uptake the dual protective polio vaccine than their counterparts (table 3).

Table 3

Determinants of non-uptake of dual protective polio vaccine among Ethiopian women who had child aged 12–35 month using the Ethiopian Demographic Health Survey 2019 data

Variables	Null model	Model I	Model II	Model III (mixed model)
Individual-level factors
ANC utilisation
No		5.83 (4.43–7.23)		5.12 (3.89–6.35)*
Yes		Reference		Reference
Place of delivery
Home		1.6 (1.04–2.3)		1.3 (0.8–2)
Facility delivery		Reference		Reference
Community-level factors
Wealth index
Poorest		1.7 (0.9–3.3)		1.33 (0.59–2.96)
Poorer		1.1 (0.5–1.9)		0.85 (0.37–1.95)
Middle		1.09 (0.6– 2.1)		1.1 (0.5–2.5)
Richer		0.9 (0.4–1.7)		0.96 (0.4–2.2)
Richest		Reference		Reference
Community educational status
High-proportion educated			1.4 (0.9–1.9)	1.35 (0.9–2.02)
Low-proportion educated			Reference	Reference
Residence
Rural			1.4 (0.8–2.3)	1.32 (0.64–2.7)
Urban			Reference	Reference
Community-media exposure
Low-proportion of media exposure			2.9 (2.67–3.13)*	2.3 (1.8–2.8)*
High-proportion of media exposure			Reference	Reference

*Statistically significant variable.

ANC, antenatal care.

Discussion

In this study, an attempt has been made to assess the prevalence of non-uptake of the dual protective polio vaccine in Ethiopia using the EDHS 2019 data set. Hence, the prevalence of non-uptake of the dual protective polio vaccine in Ethiopia was 44% (95% CI 42.2% to 45.8%). This finding is lower than the study conducted among the most at-risk populations in Ethiopia (68.5%).¹ This may be because of the time variation between the study and the previous study, which was conducted among pastoralists, remote or conflict-affected areas and urban slums. This area is a highly disadvantaged setting for health service utilisation, including the vaccination service, which makes a difference in the coverage of non-uptake of the dual protective polio vaccine. But globally, it was targeted to achieve 85%²⁸ to uptake the dual-protective polio vaccine. But in our study, about 56% of children uptake the dual protective polio vaccine (OPV-3 and IPV), and the rest, 44%, did not uptake the dual protective polio vaccine (OPV-3 and IPV). The possible reason for the high non-uptake of the dual protective polio vaccine in Ethiopia might be because of the high level of insecurity, internal instability and displacement in some parts of the country, which makes it difficult to access health services, including the vaccination programme. This implies it needs to expand the vaccination programme more.

Regarding the contributing factors for non-uptake of the dual protective polio vaccine, women who had no history of ANC utilisation were more likely to non-uptake the protective polio vaccine. This finding was supported by a study conducted in India.^{29 30} This might be because the ANC is a conventional platform for counselling and educating women about the advantages and preparedness of child vaccination. The ANC utilisation provides a good opportunity for exposure to vital information and subsequent healthcare utilisation, like delivery service, postnatal care and vaccination. However, women do not access the ANC service.

Additionally, women from low proportions of community media exposure were more likely to non-uptake the dual polio protective polio vaccine than those from high proportions of community media exposure. This finding is supported by a study done in sub-Saharan African countries.³¹ This is because media exposure is a very important key to promoting knowledge, health-seeking and preventive behaviour for women.³² In developing countries like Ethiopia, the health system is very weak, and because of this, the media is the most important source of health-related information. However, women with a low proportion of community media exposure lack the opportunity to get important health-related information like vaccination. This could influence the uptake of the dual protective polio vaccine. According to the previous evidence, residents who were rural were more likely to not uptake the dual protective polio vaccine. This is because rural women have low health service uptake behaviour, low access to health services, a lack of knowledge and low media exposure. However, it was not a significant variable in this study. This may be associated with the variability of the sample size or the power of the study.

Thus, the findings of this study have implications and provide input for designing different strategies and policies to expand the vaccination practice in Ethiopia. It is also an important input to the global polio eradication initiative. The strength of the study is that we use national-level data to generate conclusive findings and empower the power of the findings. However, because of recall bias, there may be a risk of non-differential misclassification, and it shares all the limitations of a cross-sectional study. Additionally, because of secondary data, other potentially important variables were not considered for analysis. Therefore, we recommend that future studies use other additional variables, such as mixed qualitative and quantitative studies, to overcome the limitations of this study.

Conclusion

The burden of non-uptake of the dual protective polio vaccine in Ethiopia was still high. Low community media exposure and no antenatal care utilisation were the predictors of non-uptake of the dual protective polio vaccine. We recommend improving media access and antenatal care services to expand polio vaccination.

The authors would like to give thanks to demographic health international.

Data availability statement

All data relevant to the study are included in the article.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

Because it was secondary data, ethical consent was not applicable; rather, data were requested and authorised to be accessed by DHS International. The EDHS 2019 data were collected in accordance with national and international ethical standards. The authorisation letter was obtained to use data.

Footnote

X @biruk

Contributors Conceptualisation: MCA. Data curation: EAM, MAB and TKT. Formal analysis: MCA, MAB, TKT and BBA. Investigation: MCA and AK. Methodology: MCA, MAB, TKT, EAM and BBA. Software: MCA, TKT and AKB. Supervision: MCA and AK. Visualisation: MAB, TKT, AKB, EAM and BBA. Writing the original draft: MCA, TKT and EAM. The guarantor of the study is MCA.

Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests None declared.

Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Provenance and peer review Not commissioned; externally peer reviewed.

References

¹ Gebremedhin S, Shiferie F, Tsegaye DA, et al. Oral and Inactivated Polio Vaccine Coverage and Determinants of Coverage Inequality Among the Most At-Risk Populations in Ethiopia. Am J Trop Med Hyg 2023; 109: 1148–56. doi:10.4269/ajtmh.23-0319

² Faith MR, Juliet B, Tumuhamye N, et al. Factors associated with the utilization of inactivated polio vaccine among children aged 12 to 23 months in Kalungu District, Uganda. Health Policy Plan 2020; 35: i30–7. doi:10.1093/heapol/czaa099

³ WHO. Polio eradication initiative. 2018. Available: http://www.emro.who.int/afg/programmes/polio-eradication-initiative.html [Accessed 29 Jan 2021 ].

⁴ World Health Organization. Wild poliovirus weekly update. 2010. Available: http://www/flutrackers.com/forum/showthread.php

⁵ Lucas AO, Gilles HM. Short textbook of public health medicine for the tropics. 2003 CRC press.

⁶ Tevi-Benissan C, Okeibunor J, du Châtellier GM, et al. Introduction of Inactivated Poliovirus Vaccine and Trivalent Oral Polio Vaccine/Bivalent Oral Polio Vaccine Switch in the African Region. J Infect Dis 2017; 216: S66–75. doi:10.1093/infdis/jiw616

⁷ World Health Organization. World health organization vaccination coverage cluster surveys: reference manual. 2018.

⁸ Organization, W H. Poliomilitis. 2022. Available: https://www.who.int/health-topics/poliomyelitis#tab=tab_1 [Accessed 20 Nov 2022 ].

⁹ Galea S. Macrosocial Determinants of Population Health. Springer, 2007.

¹⁰ Paul P, Chouhan P, Health G. Socio-demographic factors influencing utilization of maternal health care services in India. Clin Epidemiol Glob Health 2020; 8: 666–70. doi:10.1016/j.cegh.2019.12.023

¹¹ Parker EP, Molodecky NA, Pons-Salort M, et al. Impact of inactivated poliovirus vaccine on mucosal immunity: implications for the polio eradication endgame. Expert Rev Vaccines 2015; 14: 1113–23. doi:10.1586/14760584.2015.1052800

¹² Alfaro-Murillo JA, Ávila-Agüero ML, Fitzpatrick MC, et al. The case for replacing live oral polio vaccine with inactivated vaccine in the Americas. Lancet 2020; 395: 1163–6. doi:10.1016/S0140-6736(20)30213-0

¹³ WHO A. IPV introduced in Ethiopia. 2015. Available: https://www.afro.who.int/news/inactivated-polio virus-vaccine-ipv-introduced-ethiopia [Accessed 9 Apr 2023 ].

¹⁴ Ethiopian Public Health Institute and ICF. Ethiopia mini demographic and health survey 2019: final report. Rockville, MD EPHI and ICF; 2021.

¹⁵ Hajizadeh M. Socioeconomic inequalities in child vaccination in low/middle-income countries: what accounts for the differences? J Epidemiol Community Health 2018; 72: 719–25. doi:10.1136/jech-2017-210296

¹⁶ Dimitrova A, Carrasco-Escobar G, Richardson R, et al. Essential childhood immunization in 43 low- and middle-income countries: Analysis of spatial trends and socioeconomic inequalities in vaccine coverage. PLoS Med 2023; 20: e1004166. doi:10.1371/journal.pmed.1004166

¹⁷ Srivastava S, Fledderjohann J, Upadhyay AK. Explaining socioeconomic inequalities in immunisation coverage in India: new insights from the fourth National Family Health Survey (2015-16). BMC Pediatr 2020; 20: 295. doi:10.1186/s12887-020-02196-5

¹⁸ Ataguba JE, Ojo KO, Ichoku HE. Explaining socio-economic inequalities in immunization coverage in Nigeria. Health Policy Plan 2016; 31: 1212–24. doi:10.1093/heapol/czw053

¹⁹ Wondimu A, Cao Q, Asuman D, et al. Understanding the Improvement in Full Childhood Vaccination Coverage in Ethiopia Using Oaxaca-Blinder Decomposition Analysis. Vaccines (Basel) 2020; 8: 505. doi:10.3390/vaccines8030505

²⁰ Losey L, Ogden E, Bisrat F, et al. The CORE Group Polio Project: An Overview of Its History and Its Contributions to the Global Polio Eradication Initiative. Am J Trop Med Hyg 2019; 101: 4–14. doi:10.4269/ajtmh.18-0916

²¹ Csa, I J C S A. Central statistical agency (CSA)[Ethiopia] and ICF. Ethiopia demographic and health survey. Addis Ababa, 2016.

²² Economic, UNDoI. World population prospects. UN; 1978.

²³ Indicators, KJE and ICF. Mini Demographic and Health Survey. 2019.

²⁴ Fetene SM, Gebremedhin T. Uptake of postnatal care and its determinants in Ethiopia: a positive deviance approach. BMC Pregnancy Childbirth 2022; 22: 601. doi:10.1186/s12884-022-04933-3

²⁵ Birhanu BE, Kebede DL, Kahsay AB, et al. Predictors of teenage pregnancy in Ethiopia: a multilevel analysis. BMC Public Health 2019; 19: 601. doi:10.1186/s12889-019-6845-7

²⁶ Derseh N, Gelaye K, Muluneh A. Spatiotemporal patterns and determinants of undernutrition among late adolescent girls in Ethiopia using Ethiopian demographic and health surveys 2000 to 2016: spatiotemporal and multilevel approach. 2021.

²⁷ Merlo J, Chaix B, Ohlsson H, et al. A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena. J Epidemiol Community Health 2006; 60: 290–7. doi:10.1136/jech.2004.029454

²⁸ Rachlin A, Patel JC, Burns CC, et al. Progress Toward Polio Eradication - Worldwide, January 2020-April 2022. MMWR Morb Mortal Wkly Rep 2022; 71: 650–5. doi:10.15585/mmwr.mm7119a2

²⁹ Dixit P, Dwivedi LK, Ram F. Strategies to Improve Child Immunization via Antenatal Care Visits in India: A Propensity Score Matching Analysis. PLoS One 2013; 8: e66175. doi:10.1371/journal.pone.0066175

³⁰ Budu E, Seidu A-A, Agbaglo E, et al. Maternal healthcare utilization and full immunization coverage among 12-23 months children in Benin: a cross sectional study using population-based data. Arch Public Health 2021; 79: 34. doi:10.1186/s13690-021-00554-y

³¹ Jung M, Lin L, Viswanath K. Effect of media use on mothers’ vaccination of their children in sub-Saharan Africa. Vaccine (Auckl) 2015; 33: 2551–7. doi:10.1016/j.vaccine.2015.04.021

³² Randolph W, Viswanath K. Lessons learned from public health mass media campaigns: marketing health in a crowded media world. Annu Rev Public Health 2004; 25: 419–37. doi:10.1146/annurev.publhealth.25.101802.123046