Childhood immunization Childhood INTERACTIVE VISUALIZATION OF HEALTH DATA HEALTH OF VISUALIZATION INTERACTIVE EXPLORATIONS OF INEQUALITY EXPLORATIONS
EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION WHO SWITZERLAND 20, AVENUE APPIA 20, AVENUE CH-1211 GENEVA 27 CH-1211 GENEVA WORLD HEALTH ORGANIZATION HEALTH WORLD HEALTH METRICS AND MEASUREMENT CLUSTER METRICS AND MEASUREMENT HEALTH DEPARTMENT OF INFORMATION, EVIDENCE AND RESEARCH OF INFORMATION, DEPARTMENT ISBN 978 92 4 156561 5 http://www.who.int/gho/health_equity/report_2018_immunization/en/
EXPLORATIONS OF INEQUALITY Childhood immunization
INTERACTIVE VISUALIZATION OF HEALTH DATA Explorations of inequality: childhood immunization
ISBN 978-92-4-156561-5
© World Health Organization 2018
Some rights reserved. This work is available under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/igo).
Under the terms of this licence, you may copy, redistribute and adapt the work for non-commercial purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there should be no suggestion that WHO endorses any specific organization, products or services. The use of the WHO logo is not permitted. If you adapt the work, then you must license your work under the same or equivalent Creative Commons licence. If you create a translation of this work, you should add the following disclaimer along with the suggested citation: “This translation was not created by the World Health Organization (WHO). WHO is not responsible for the content or accuracy of this translation. The original English edition shall be the binding and authentic edition”.
Any mediation relating to disputes arising under the licence shall be conducted in accordance with the mediation rules of the World Intellectual Property Organization (http://www.wipo.int/amc/en/mediation/rules).
Suggested citation. Explorations of inequality: childhood immunization. Geneva: World Health Organization; 2018. Licence: CC BY-NC-SA 3.0 IGO.
Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris.
Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders. To submit requests for commercial use and queries on rights and licensing, see http://www.who.int/about/licensing.
Third-party materials. If you wish to reuse material from this work that is attributed to a third party, such as tables, figures or images, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright holder. The risk of claims resulting from infringement of any third-party-owned component in the work rests solely with the user.
General disclaimers. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of WHO concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by WHO in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters.
All reasonable precautions have been taken by WHO to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The responsibility for the interpretation and use of the material lies with the reader. In no event shall WHO be liable for damages arising from its use.
The statistics presented in this publication are based on the re-analysis of Demographic and Health Surveys micro-data and are not necessarily the official statistics of WHO.
Design and layout by L’IV Com Sàrl, Villars-sous-Yens, Switzerland. Printed by the WHO Document Production Services, Geneva, Switzerland. Contents
Forewords...... viii
Acknowledgements...... x
Frequent abbreviations and acronyms...... xi
Executive summary...... xii
1. Introduction...... 1
2. Methods...... 4 Data sources...... 4 Priority countries...... 4 Childhood immunization indicator...... 4 Analysis approaches...... 5 Interactive visuals and tables...... 8
3. Afghanistan...... 10 Country context...... 10 Descriptive overview...... 10 Multiple regression analysis...... 12
4. Chad...... 14 Country context...... 14 Descriptive overview...... 14 Multiple regression analysis...... 16
5. Democratic Republic of the Congo...... 18 Country context...... 18 Descriptive overview...... 18 Multiple regression analysis...... 20
6. Ethiopia...... 22 Country context...... 22 Descriptive overview...... 22 Multiple regression analysis...... 24
7. India...... 26 Country context...... 26 Descriptive overview...... 26 Multiple regression analysis...... 28
iii 8. Indonesia...... 30 Country context...... 30 Descriptive overview...... 30 Multiple regression analysis...... 32
9. Kenya...... 34 Country context...... 34 Descriptive overview...... 34 Multiple regression analysis...... 36
10. Nigeria...... 38 Country context...... 38 Descriptive overview...... 38 Multiple regression analysis...... 40
11. Pakistan...... 42 Country context...... 42 Descriptive overview...... 42 Multiple regression analysis...... 44
12. Uganda...... 46 Country context...... 46 Descriptive overview...... 46 Multiple regression analysis...... 48
13. Multicountry assessment...... 50 Descriptive overview...... 50 Characteristics associated with DTP3 immunization coverage across countries ...... 58
14. Conclusion...... 62 Taking stock...... 62 Understanding the analysis...... 64 Moving forward...... 64
Appendix 1...... 67
Index...... 68
iv Contents
Figures Figure 3.1. DTP3 immunization coverage among one-year-olds in Afghanistan, disaggregated by background characteristics (DHS 2015)...... 11 Figure 3.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Afghanistan, calculated as odds ratio (DHS 2015)...... 12 Figure 4.1. DTP3 immunization coverage among one-year-olds in Chad, disaggregated by background characteristics (DHS 2014)...... 15 Figure 4.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Chad, calculated as odds ratio (DHS 2014)...... 16 Figure 5.1. DTP3 immunization coverage among one-year-olds in the Democratic Republic of the Congo, disaggregated by background characteristics (DHS 2013)...... 19 Figure 5.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in the Democratic Republic of the Congo, calculated as odds ratio (DHS 2013)...... 20 Figure 6.1. DTP3 immunization coverage among one-year-olds in Ethiopia, disaggregated by background characteristics (DHS 2016)...... 23 Figure 6.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Ethiopia, calculated as odds ratio (DHS 2016)...... 24 Figure 7.1. DTP3 immunization coverage among one-year-olds in India, disaggregated by background characteristics (NFHS 2015)...... 27 Figure 7.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in India, calculated as odds ratio (NFHS 2015)...... 28 Figure 8.1. DTP3 immunization coverage among one-year-olds in Indonesia, disaggregated by background characteristics (DHS 2012)...... 31 Figure 8.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Indonesia, calculated as odds ratio (DHS 2012)...... 32 Figure 9.1. DTP3 immunization coverage among one-year-olds in Kenya, disaggregated by background characteristics (DHS 2014)...... 35 Figure 9.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Kenya, calculated as odds ratio (DHS 2014)...... 36 Figure 10.1. DTP3 immunization coverage among one-year-olds in Nigeria, disaggregated by background characteristics (DHS 2013)...... 39 Figure 10.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Nigeria, calculated as odds ratio (DHS 2013)...... 40
v Figure 11.1. DTP3 immunization coverage among one-year-olds in Pakistan, disaggregated by background characteristics (DHS 2012)...... 43 Figure 11.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Pakistan, calculated as odds ratio (DHS 2012)...... 44 Figure 12.1. DTP3 immunization coverage among one-year-olds in Uganda, disaggregated by background characteristics (DHS 2016)...... 47 Figure 12.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Uganda, calculated as odds ratio (DHS 2016)...... 48 Figure 13.1. National DTP3 immunization coverage among one-year-olds in 10 priority countries (DHS 2012–2016 and NFHS 2015)...... 50 Figure 13.2. DTP3 immunization coverage among one-year-olds by child’s sex in 10 priority countries (DHS 2012–2016 and NFHS 2015)...... 51 Figure 13.3. DTP3 immunization coverage among one-year-olds by birth order in 10 priority countries (DHS 2012–2016 and NFHS 2015)...... 52 Figure 13.4. DTP3 immunization coverage among one-year-olds by mother’s age at birth in 10 priority countries (DHS 2012–2016 and NFHS 2015)...... 52 Figure 13.5. DTP3 immunization coverage among one-year-olds by mother’s education in 10 priority countries (DHS 2012–2016 and NFHS 2015)...... 53 Figure 13.6. DTP3 immunization coverage among one-year-olds, by household economic status in 10 priority countries (DHS 2012–2016 and NFHS 2015)...... 54 Figure 13.7. DTP3 immunization coverage among one-year-olds by place of residence in 10 priority countries (DHS 2012–2016 and NFHS 2015)...... 55 Figure 13.8. Wealth-related Erreygers concentration index for DTP3 immunization coverage among one-year-olds in 10 priority countries (DHS 2012–2016 and NFHS 2015)... 56 Figure 13.9. Education-related Erreygers concentration index for DTP3 immunization coverage among one-year-olds in 10 priority countries (DHS 2012–2016 and NFHS 2015)... 56 Figure 13.10. Concentration curves for DTP3 immunization coverage by household economic status in 10 priority countries (DHS 2012–2016 and NFHS 2015)...... 57 Figure 13.11. Adjusted associations between household economic status and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for the richest quintile compared with the poorest quintile (DHS 2012–2016 and NFHS 2015)...... 58 Figure 13.12. Adjusted associations between mother’s education and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for the most educated compared with the least educated (DHS 2012–2016 and NFHS 2015)...... 59 Figure 13.13. Adjusted associations between mother’s age at birth and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for the 20–34 years subgroup compared with the 15–19 years subgroup (DHS 2012–2016 and NFHS 2015)...... 59 vi Contents
Figure 13.14. Adjusted associations between birth order and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for the 1st born compared with the 6th born or higher (DHS 2012–2016 and NFHS 2015)...... 60 Figure 13.15. Adjusted associations between sex of household head and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for females compared with males (DHS 2012–2016 and NFHS 2015)...... 61 Figure 13.16. Adjusted associations between place of residence and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for urban compared with rural (DHS 2012–2016 and NFHS 2015)...... 61
Tables Table 2.1. Priority countries included in this report...... 4 Table 2.2. Background characteristics and associated subgroups applied to describe within- country inequality...... 6
Table 2.3. Background characteristics included in the multiple regression analyses...... 7
vii Foreword
he United Nations 2030 Agenda for Sustainable Development resolves to combat inequalities within and among countries. It endeavours to reach the furthest behind first, envisioning a world with universal and equitable access to education, health care and social protection, where no one is left behind. Health equity is a cross-cutting priority for the World Health Organization (WHO). Equity-oriented policies, programmes and practices help to ensure that health and the opportunities to attain health can be realized by everyone, everywhere and always.
I am now pleased to welcome the report Explorations of inequality: childhood immunization as an important extension of this work. This is a landmark report that goes beyond single stratification, illustrating how vulnerability and advantage can compound across multiple dimensions of inequality. The report offers a detailed analysis of national coverage and inequalities in childhood immunization across 10 priority countries, probing the factors that contribute to national coverage levels, as well as the inequalities that impede progress. Building on disaggregated data – the starting point of inequality measurements – the analyses employed in this report deepen our understanding of childhood immunization coverage. The report demonstrates how sophisticated approaches to health monitoring are relevant to decision-makers in health and other sectors.
Achieving the equity mandate of the Agenda for Sustainable Development requires concerted efforts, underpinned by reliable information systems that measure and monitor progress. This report is an important contribution to establishing that evidence base in childhood immunization. The type of exploration presented here serves as an example of what can be done across other health topics and settings.
Dr Lubna A Al-Ansary Assistant Director-General Health Metrics and Measurement Cluster WHO headquarters Geneva, Switzerland
viii Foreword
onitoring inequalities in health remains central to advancing health equity, signalling where gaps in health exist and encouraging innovative approaches to address them. To this end, M national health information systems with the capability to collect, analyse and report on health inequality are a necessary investment for health inequality monitoring. WHO promotes the monitoring of health inequalities, globally and nationally. In recent years, publications such as the Handbook on health inequality monitoring: with a special focus on low- and middle-income countries, and the National health inequality monitoring: a step-by-step manual have outlined systematic approaches for monitoring; the State of inequality report series has shared the results of cross-country and national monitoring, enabled by tools such as the Health Equity Assessment Toolkit and the Health Equity Monitor database and theme page.
In this report, the 10 priority countries present unique case studies for childhood immunization, illustrating the value of health inequality monitoring in different contexts. In some countries, the rigorous analysis presented in the report was previously lacking; the findings in these settings will be particularly valuable as an evidence basis for childhood immunization initiatives to reach underserved populations, and to prompt health information system strengthening. In countries that reported low levels of inequality, ongoing assessment is vital to ensure that progress in childhood immunization continues to benefit all population groups. Within all 10 countries, the report called for heightened action to address geographical inequalities between subnational regions.
Dr John T Grove Director Department of Information, Evidence and Research Health Metrics and Measurement Cluster WHO headquarters Geneva, Switzerland
ix Acknowledgements
Ahmad Reza Hosseinpoor (Health Equity Monitoring Lead, Department of Information, Evidence and Research, Health Metrics and Measurement Cluster, World Health Organization, Geneva, Switzerland) led the conceptualization and development of the report in close collaboration with Nicole Bergen (Faculty of Health Sciences, University of Ottawa, Ottawa, Canada), Anne Schlotheuber (Technical Officer, Department of Information, Evidence and Research, Health Metrics and Measurement Cluster, World Health Organization, Geneva, Switzerland) and Cecilia Vidal Fuertes (consultant). Ahmad Reza Hosseinpoor and Cecilia Vidal Fuertes defined the analytical framework. Cecilia Vidal Fuertes prepared the data and did the analyses; Nicole Bergen drafted the report text and provided technical editing support; and Anne Schlotheuber developed the graphics and interactive visuals.
The WHO Collaborating Center for Health Equity Monitoring (International Center for Equity in Health, Federal University of Pelotas, Brazil) provided guidance on the computation of the childhood immunization indicator used in this report. The report benefited from the contributions of numerous subject matter experts, who supported the development of the report across all stages of development. We acknowledge Owen O’Donnell (Erasmus University Rotterdam, Rotterdam, Netherlands), who made contributions to refining the analysis approach and reviewed the report. Aluisio JD Barros (Federal University of Pelotas, Pelotas, Brazil) and Adama Diop (consultant) reviewed the report. From Gavi, the Vaccine Alliance, Emmanuel Bor, Gustavo Correa, Samuel Muller, Antara Sinha and Carol Szeto reviewed early drafts of the report, and provided valuable comments.
Reviewers from the World Health Organization included: Jhilmil Bahl, Vinoid Bura, Somnath Chatterji, Kamal Fahmy, Tracey S Goodman, Quamrul Hasan, Annet Kisaye, Theadora Swift Koller, Wahyu Retno Mahanani, Helen Matzger, Alejandro Ramirez Gonzalez, Stephanie Shendale and Nadia Abd El-aziz Teleb Badr.
AvisAnne Julien provided copy-editing support and proofread the document, and Christine Boylan prepared the index.
Funding for the project was provided by Gavi, the Vaccine Alliance.
x Frequent abbreviations and acronyms
DHS Demographic and Health Surveys DTP combined diphtheria, tetanus toxoid and pertussis vaccine ECI Erreygers normalized concentration index ICT Islamabad: Islamabad Capital Territory NFHS National Family Health Survey WHO World Health Organization
xi Executive summary
Childhood immunization is a key intervention to Vaccine Alliance identified as the highest priority promote the health, well-being and survival of for childhood immunization. These countries face children. Since the introduction of the Expanded the most severe immunization challenges, and Programme on Immunization in 1974 by the World together account for more than 70% of children Health Organization (WHO), global efforts have who do not get a full course of basic vaccines. been in place to expand the benefits of vaccines Within each country, the report assessed: childhood across all countries and population groups. Despite immunization data disaggregated by child, mother, marked success in many aspects of immunization household and geographic characteristics; wealth- programmes, inequalities in childhood and education-related inequalities in childhood immunization remain a challenge. Gavi, the Vaccine immunization, calculated as concentration curve Alliance, established in 2000, is committed to and concentration index; and adjusted associations accelerating the equitable uptake and coverage of between immunization coverage and selected vaccines, especially within poor countries and for socioeconomic, demographic and geographic populations that are underserved. factors, through multiple regression analysis. The findings of the report are presented alongside This report provides a closer look at the factors details about the country context, followed by a associated with childhood immunization. The multicountry assessment and discussion of the report focuses on the 10 countries that Gavi, the major implications of the findings.
Methods at a glance
Gavi, the Vaccine Alliance priority countries: Afghanistan, Chad, Democratic Republic of the Congo, Ethiopia, India, Indonesia, Kenya, Nigeria, Pakistan and Uganda. Data sources: For each priority country, data were sourced from the latest publicly available Demographic and Health Surveys (DHS), ranging from 2012 to 2016. Surveys were conducted with a representative sample of women aged 15–49 years and included questions about childhood immunization and background characteristics. Childhood immunization indicator: The childhood immunization indicator featured in this report is DTP3 immunization: the percentage of children aged 12–23 months who had received three doses of the combined diphtheria, tetanus toxoid and pertussis vaccine (DTP3) (or any DTP-containing vaccine, including pentavalent and tetravalent). Descriptive analysis: Disaggregated data show childhood immunization coverage broken down by: child’s sex, birth order, mother’s age at birth, mother’s education, mother’s ethnicity or caste/tribe, sex of household head, household economic status, place of residence and subnational regions. Concentration curve and concentration index were used to measure how DTP3 coverage varied by socioeconomic status. Multiple regression analysis: A logistic regression model was used to assess the associations between DTP3 immunization coverage and selected child, mother, household and geographic characteristics. Estimated associations are presented as odds ratios. For more details about the methods, please refer to Chapter 2.
xii Executive summary
Major findings, by country sharply in quintile 5. When adjusted for other characteristics, the odds of immunization were 2.5 Afghanistan: Nationally, only three in five one-year- times higher in the richest than the poorest quintile olds in Afghanistan received three doses of the DTP and twice as high in mothers with secondary vaccine. Childhood immunization demonstrated no education or more as in mothers with no education. inequality by child sex, and low levels of inequality by The regions of Mandoul and Mayo Kebbi Ouest had birth order; however, coverage varied substantially 27 and 19 times higher odds of coverage than Chari according to mother’s characteristics (age at birth, Baguirmi, respectively. education and ethnicity), household economic status, place of residence and subnational region. In Chad in 2014–2015, a child of a mother aged 20– Certain ethnic groups (especially the Nuristanis) 34 years with secondary education or higher and and regions (including Nooristan and Urozgan) belonging to the richest 20% had up to 7.2 times registered very low levels of coverage. When higher chance of receiving DTP3 immunization controlling for other background characteristics, compared with a child of a teenaged mother with immunization coverage was significantly associated no education, from the poorest 20% household. with mother’s age at birth, mother’s education, household economic status and subnational region. Democratic Republic of the Congo: Overall, the Other things equal, the odds of coverage were 2.7 Democratic Republic of the Congo had national times larger for children in the richest quintile than DTP3 immunization covering three out of five in the poorest quintile, and almost twice as large one-year-olds. Although there was no or little among mothers aged 35–49 years (compared to inequality by child’s sex, birth order or mother’s age those aged 15–19 years) and among mothers with at birth, coverage varied by other factors: mother’s secondary school or more (compared to those with education, mother’s ethnicity, household economic no education). status, place of residence and subnational region. For instance, coverage across ethnic groups In Afghanistan in 2015, a child of a teenaged mother ranged from a minimum of 43% in the Uele Lac with no education had one third the chance of being Albert and Cuvette Central subgroups to 83% vaccinated as a child of a mother 20–49 years of in the Bakongo Nord & Sud subgroup. While age with secondary education or higher; if the child coverage demonstrated a steep gradient across of the uneducated, teenaged mother belonged to wealth quintiles, the urban–rural split spanned 20 the poorest 20%, this chance dropped to one ninth percentage points, with higher coverage in urban (compared to a child of a highly educated mother areas. Controlling for other factors, immunization aged 20–49 years in the richest 20%). coverage was associated with mother’s ethnicity, household economic status and subnational region. Chad: Childhood immunization coverage in Chad The odds of receiving the third dose of the DTP was very low, with just one third of children vaccine were 9 times higher for children in the reporting DTP3 immunization. DTP3 coverage Nord-Kivu province than children in the Katanga showed no difference between boys and girls, but province, other things equal. there was some inequality by birth order, mother’s age at birth, sex of household head and place of Ethiopia: About half of one-year-olds in Ethiopia residence. Coverage varied substantially according were covered by DTP3 immunization. With the to mother’s education level and subnational region. exception of child’s sex, where coverage was Coverage across wealth quintiles demonstrated the same in boys and girls, Ethiopia reported a mass deprivation pattern, where coverage was inequalities according to all other background uniformly low in quintiles 1 to 4, and increased characteristics featured in the report. Most notably,
xiii EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
coverage was higher among children of mothers In India in 2015–2016, children with highly educated with secondary school or more education, richer mothers aged 20–49 years who belonged to the households, urban areas and certain subnational richest 20% of the population had a 5.3 times regions (especially Addis Ababa, Dire Dawa and higher chance of being vaccinated, compared Tigray, where coverage was over 80%). Ethnicity- with children born to teenaged mothers with no related inequality demonstrated a substantial education, in the poorest 20% of the population. difference of 60 percentage points, from 22% in the Affar population to 82% in the Tigray population. Indonesia: Indonesia reported 72% DTP3 Controlling for other factors, immunization immunization coverage among one-year-olds. coverage was associated with mother’s age at Male and female children had similar levels of birth, mother’s education, sex of household head, coverage, as did children of male- and female- household economic status and subnational region. headed households. Coverage dropped sharply with increasing birth order, and pronounced In Ethiopia in 2016, the chance of receiving the inequalities were observed by mother’s education, third dose of DTP vaccine was 6.7 times higher household economic status and subnational region. among a child whose mother is 20–49 years of age Some inequality was observed by mother’s age and primary school educated, and who lived in a at birth (maximal coverage in the 20–34 years male-headed household, compared with a child of group) and place of residence (favouring urban a teenaged mother with no education in a female- areas). Controlling for other factors, immunization headed household. coverage was strongly associated with birth order, mother’s education, household economic status India: DTP3 immunization coverage in India and subnational region. Across provinces, the odds reached nearly four out of five one-year-olds. of childhood immunization demonstrated large Coverage was equal in boys and girls and female- variation: children in DI Yogyakarta, Bali, East Nusa and male-headed households, and there was little Tenggara and North Sulawesi were more than 8 difference between coverage in urban and rural times as likely to be covered as those in Banten. areas. Gradients in coverage were observed across birth order (higher order births showing lower In Indonesia in 2012, a child who was part of a coverage), mother’s education level (less-educated household in the richest 20%, and whose mother subgroups showing lower coverage) and wealth was aged 35–49 years, had a 6.4 times greater quintiles (poorer quintiles showing lower coverage). chance of being vaccinated compared to a child DTP3 immunization coverage tended to be lower living in a household of the poorest 20%, and among children of mothers aged 35–49 years, whose mother was a teenager. children belonging to scheduled tribes and children in certain subnational regions (e.g. Nagaland and Kenya: Nine out of 10 one-year-olds in Kenya Arunachal Pradesh, where coverage was just over were covered by DTP3 immunization. There was 50%). Adjusting for other factors, immunization no inequality by child’s sex, sex of household coverage was associated with birth order, mother’s head or place of residence; some inequality was education and economic status, and there was large reported by mother’s age at birth and household variation in the odds of coverage across regions. economic status. Coverage was lower for children There was a weak, although significant, association born 6th or higher, for children of mothers with no for mother’s age at birth and mother’s caste/tribe education, and for children of Maasai and Somali as well as for place of residence (demonstrating ethnic groups. Across the eight subnational regions, higher likelihood of coverage in rural areas). the lowest coverage was reported by the North Eastern region (78%) and the highest coverage xiv Executive summary
was reported by the Central region (96%). After In Nigeria in 2013, children of mothers aged controlling for background characteristics, DTP3 20–34 years who were highly educated, living immunization coverage was positively associated in a rich household in the South South region with mother’s education and household economic were among the most advantaged in terms of status. The odds of reporting DTP3 coverage were childhood immunization: their chance of being over 2 times higher for 1st to 5th born children, vaccinated was 300 times higher than children with compared to those 6th born or higher. Statistically teenaged mothers with no education, living in poor significant associations with DTP3 immunization households in the North West region. were also reported across subnational regions in Kenya. Pakistan: Among one-year-olds in Pakistan, national DTP3 immunization coverage was 65%. In Kenya in 2014, children had a higher chance of Variable levels of inequality were reported across being vaccinated if they belonged to the richest background characteristics, most prominently by 40% of households and their mother had at least mother’s education, household economic status primary school education: compared to those who and subnational region. For both education and were part of the poorest 20% and whose mother had wealth, coverage was markedly lower in the most no education, their chances were 6.3 times higher. disadvantaged subgroups (the least-educated and the poorest). Across subnational regions, Nigeria: In Nigeria, two in five one-year-olds coverage spanned from 28% in Balochistan received the third dose of DTP-containing vaccine. to 92% in ICT Islamabad. After controlling for While females and males reported similar other characteristics, immunization coverage levels of coverage, substantial inequalities were was strongly associated with mother’s education observed by all other studied characteristics. and especially household economic status, and Coverage among 1st born children was about subnational region. The odds of coverage differed twice as high as coverage among children 6th greatly across regions, with ICT Islamabad having born or higher; coverage was 24 percentage points 7.8 times higher odds of coverage than Sindh. higher among female-headed households; and coverage by mother’s age at birth spanned 20 In Pakistan in 2012–2013, a child of a mother aged percentage points, from 22% in 15–19-year-olds 20–34 years with higher than secondary education to 42% in 20–34-year-olds. DTP3 immunization and from the richest 20% of the population had demonstrated steep gradients by both mother’s a 28 times higher chance of being vaccinated, education and household economic status, and compared with a child of a teenaged mother with was about 2.5 times higher in urban than rural no education and from the poorest 20% of the areas. Across the six subnational regions, coverage population. ranged from 14% in the North West region to 81% in the South East region. Overall, immunization Uganda: Four out of five one-year-olds in was highly concentrated among children of richer Uganda were covered by DTP3 immunization. households and more-educated mothers. After Little variation in coverage was reported across adjusting for other factors, immunization coverage several studied background characteristics, except was significantly associated with most background for mother’s education and subnational region. characteristics (birth order and place of residence Coverage was similar in boys and girls and differed were nearly significant and child’s sex was not by less than 4 percentage points across subgroups significant). Strong associations were evident for for mother’s age at birth, sex of household head, mother’s education, household economic status household economic status and place of residence. and subnational regions.
xv EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
According to birth order, coverage was lower in 6th and inequality by subnational region tended to or higher order births, but similar across children be substantial. All countries reported variation by of 1st to 5th birth order. By mother’s education, mother’s education and subnational region and coverage was higher among children of the most- all (except Uganda) demonstrated inequality on educated subgroup (with more than secondary the basis of household economic status. All 10 school) than other education subgroups, with a priority countries showed a positive association difference spanning 11 percentage points. Across between mother’s education level and childhood subnational regions, coverage varied from 71% in immunization coverage. Countries that reported low Busoga to 91% in Teso. After adjusting for other national coverage (e.g. Chad, Ethiopia and Nigeria) characteristics, childhood immunization was tended to demonstrate steep gradients and/or significantly associated with birth order, mother’s mass deprivation patterns across socioeconomic age at birth and subnational region. For instance, subgroups; the odds of immunization tended to be children of mothers with more than secondary significantly higher in more advantaged subgroups school education had twice as high odds of in these countries. Countries with higher national receiving the DTP3 vaccine than children whose coverage (e.g. India, Indonesia, Kenya and Uganda), mothers had no education. more often demonstrated marginal exclusion or universal patterns across socioeconomic subgroups, Key messages and tended to have lower urban–rural inequality. Across the 10 priority countries, the national DTP3 immunization coverages ranged from 34% in Chad When considered alongside knowledge of the to 90% in Kenya. Evaluating performance based country context, the results of this report can be on national averages alone, however, masks the used to inform equity-oriented policies, programmes situation in population subgroups. The countries and practices to promote universal childhood faced distinct patterns of inequality, from Uganda, immunization coverage. This report serves as a where inequality tended to be very small for most basis for more detailed explorations at the national of the featured characteristics, to Nigeria, where and subnational levels, and a baseline for future inequality was pronounced for most characteristics. health inequality monitoring efforts. Monitoring and exploring inequalities in health is essential as Despite the uniqueness of each country situation, countries strive to “leave no one behind” on the path some commonalities emerged. Inequalities by towards sustainable development. child’s sex tended to be minimal or non-existent,
xvi 1. Introduction
Immunization is a safe, effective and cost-effective Still, not all children have equal opportunity to health intervention across different life stages, benefit from vaccines. While the success of including childhood (1–4). Childhood immunization immunization programmes is linked to aspects is a key strategy for the reduction of child morbidity of health systems – such as accessibility of health and mortality: over the period from 2000 to facilities, health worker knowledge, vaccine supply 2015, the greatest decline in child mortality was and cost for vaccinations (12) – wider determinants attributable to reductions in vaccine-preventable of health also impact childhood immunization. In diseases (5). Immunization is considered an essential low-resource settings, childhood immunization health service, and its coverage is monitored as part programmes are affected by political stability, of Sustainable Development Goal 3: to ensure war and unrest, gender equality, living conditions, healthy lives and promote well-being for all at all traditional or cultural practices, geographic ages (6). characteristics, government spending on health, and receipt of external support for health (13,14). In 1974, the World Health Organization (WHO) established the Expanded Programme on Within countries, gaps in immunization coverage Immunization to develop and promote routine exist between population subgroups. The report immunization programmes in countries. Since this State of inequality: childhood immunization presents time, national immunization programmes have been within-country inequalities in 69 countries, introduced around the world, and subsequently demonstrating large and persistent gaps on the basis shaped by new technologies, evolving vaccine of economic status and mother’s education level in delivery practices and innovative approaches to most countries (15). Countries have demonstrated financing. Global campaigns and initiatives, past different patterns of inequality in childhood and present, have contributed to the expanded immunization, with variable levels of within-country reach of vaccines and their benefits to diverse inequality by wealth, education, gender, place of geographic settings and population groups (7–11). residence and other characteristics (15–18).
Gavi, the Vaccine Alliance
Established in 2000, Gavi, the Vaccine Alliance was founded to address inequities in access to childhood vaccines, especially in poorer countries and among disadvantaged populations. Increasingly, guided by the framework of the WHO Global Vaccine Action Plan (9), Gavi has adopted a more explicit focus on addressing within-country inequalities (19). Through the 2016–2020 Gavi Strategic Action Plan, Gavi committed to the goal of accelerating equitable uptake and coverage of vaccines, and will monitor equity of coverage and barriers based on geography, wealth quintiles, mother’s (or female caretaker’s) education and fragile state status (20).
By 2020, Gavi aspires to reach over 300 million children, and see a 10% reduction in under-5 mortality in Gavi-supported countries. Gavi currently engages with 72 countries, including 10 high-priority, tier-1 countries. These 10 countries face severe immunization challenges, and together account for more than 70% of children who do not get a full course of basic vaccines. Gavi’s engagement with these countries is guided by Gavi’s Partners Engagement Framework, based on four key principles: country-focused approach; differentiation; transparency; and accountability (21).
1 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Countries have employed numerous strategies to health, policy-makers, researchers and public health increase the coverage of childhood immunization professionals). among underserved populations, with varying success. Common ways of promoting childhood The report is organized in 14 chapters. Chapter immunization include: providing education 2 outlines the methods, including data sources, to parents and community members; issuing priority country selection, the immunization reminder cards; giving conditional cash transfers indicator and analysis approaches. Chapters 3–12 for parents who vaccinate their children; delivering present findings across the 10 featured countries. regular vaccination outreach activities in villages; These chapters provide a brief overview of the identifying unvaccinated children at home visits and country context, followed by the findings from referring them to clinics; and integrating vaccination the analyses. Chapter 13 presents a multicountry services with other health services (22). Further assessment of the findings across all 10 countries. research is needed to determine how policies, Finally, Chapter 14 concludes with an overview of programmes and practices can be designed for the major findings and suggested directions for moving highest impact in target populations. forward towards universal childhood immunization.
This report aims to facilitate a better understanding References of factors associated with childhood immunization 1. Ozawa S, Clark S, Portnoy A, Grewal S, Brenzel L, Walker coverage. Focusing on each of Gavi’s 10 tier-1 DG. Return on investment from childhood immunization countries, the in-depth analyses characterize the in low- and middle-income countries, 2011–2020. magnitude of inequality in childhood immunization Health Aff (Millwood). 2016;35(2):199–207. across multiple dimensions of inequality and identify 2. Andre F, Booy R, Bock H, Clemens J, Datta S, John T factors that are associated with immunization et al. Vaccination greatly reduces disease, disability, death and inequity worldwide. Bull World Health. 2008 coverage. The main objectives of the report are: February 1;86(2):140–6. 3. Bloom DE. The value of vaccination. In: Curtis N, Finn A, • to present disaggregated childhood immunization Pollard AJ, editors. Hot topics in infection and immunity data by child, mother, household and geographic in children VII. New York: Springer; 2011. characteristics in Gavi’s 10 tier-1 countries 4. Clements CJ, Nshimirimanda D, Gasasira A. Using (Afghanistan, Chad, Democratic Republic of the immunization delivery strategies to accelerate progress Congo, Ethiopia, India, Indonesia, Kenya, Nigeria, in Africa towards achieving the Millennium Development Pakistan and Uganda); Goals. Vaccine. 2008 April;26(16):1926–33. 5. Countdown to 2030 Maternal, Newborn & Child • to measure wealth- and education-related Survival. Tracking progress towards universal coverage for reproductive, newborn and child health: the 2017 inequalities in those countries; and report. Report No.: Licence: CC BY-NC-SA 3.0 IGO. Washington (DC): United Nations Children’s Fund • to assess the factors that underlie childhood (UNICEF) and the World Health Organization (WHO); immunization coverage in those countries. 2017. 6. Resolution A/RES/70/1. Transforming our world: the The findings of this report serve to inform evidence- 2030 agenda for sustainable development. Adopted based and equity-oriented policies, programmes by the United Nations General Assembly, New York, 25 September 2015. and practices; they also indicate areas for further research. The report was developed for 7. WHO Department of Immunization, Vaccines and Biologicals, UNICEF Programme Division. Global an audience with some knowledge of statistics Immunization Vision and Strategy 2006–2015. Geneva: as well as health inequality measurements and World Health Organization; 2005. their applications (especially staff of ministries of
2 1. Introduction
8. Vandelaer J, Bilous J, Nshimirimana D. Reading Every 17. Arsenault C, Harper S, Nandi A, Mendoza Rodríguez JM, District (RED) approach: a way to improve immunization Hansen PM, Johri M. Monitoring equity in vaccination performance. Bull World Health. 2008;86(3):A–B. coverage: a systematic analysis of demographic and 9. Global Vaccine Action Plan 2011–2020. Geneva: World health surveys from 45 Gavi-supported countries. Health Organization; 2013. Vaccine. 2017 February;35(6):951–9. 10. Immunization [Internet]. New York: UNICEF; 2018 18. Restrepo-Méndez MC, Barros AJ, Wong KL, Johnson (https://www.unicef.org/immunization/, accessed 20 HL, Pariyo G, França GV et al. Inequalities in full April 2018). immunization coverage: trends in low- and middle- income countries. Bull World Health. 2016 November 11. Immunization policy and strategies [Internet]. 1;94(11):794–805A. Immunization, vaccines and biologicals. Geneva: World Health Organization (http://www.who.int/ 19. Gandhi G. Charting the evolution of approaches immunization/programmes_systems/policies_ employed by the Global Alliance for Vaccines and strategies/en/, accessed 20 April 2018). Immunizations (GAVI) to address inequities in access to immunization: a systematic qualitative review of GAVI 12. Rainey JJ, Watkins M, Ryman TK, Sandhu P, Bo A, policies, strategies and resource allocation mechanisms Banerjee K. Reasons related to non-vaccination and through an equity lens (1999–2014). BMC Public Health under-vaccination of children in low- and middle- [Internet]. 2015 December (http://bmcpublichealth. income countries: findings from a systematic review biomedcentral.com/articles/10.1186/s12889-015-2521- of the published literature, 1999–2009. Vaccine. 2011 8, accessed 16 April 2015). October;29(46):8215–21. 20. Gavi, the Vaccine Alliance 2016–2020 Strategy 13. Arsenault C, Johri M, Nandi A, Mendoza Rodríguez [Internet]. Geneva: Gavi, the Vaccine Alliance; 2014 JM, Hansen PM, Harper S. Country-level predictors of (https://www.gavi.org/about/strategy/, accessed 18 vaccination coverage and inequalities in Gavi-supported April 2018). countries. Vaccine. 2017 April;35(18):2479–88. 21. Engaging partners for success [Internet]. Geneva: Gavi, 14. Glatman-Freedman A, Nichols K. The effect of social the Vaccine Alliance; 2018 (https://www.gavi.org/ determinants on immunization programs. Hum Vaccines library/news/gavi-features/2017/engaging-partners- Immunother. 2012 March 13;8(3):293–301. for-success/, accessed 24 April 2018). 15. State of inequality: childhood immunization. Geneva: 22. Oyo-Ita A, Wiysonge CS, Oringanje C, Nwachukwu World Health Organization; 2016. CE, Oduwole O, Meremikwu MM; Cochrane Effective 16. Hosseinpoor AR, Bergen N, Schlotheuber A, Gacic-Dobo Practice and Organisation of Care Group, editor. M, Hansen PM, Senouci K et al. State of inequality in Interventions for improving coverage of childhood diphtheria-tetanus-pertussis immunisation coverage immunisation in low- and middle-income countries. in low-income and middle-income countries: a Cochrane Database Syst Rev. 2016 July 10. multicountry study of household health surveys. Lancet Glob Health. 2016 September;4(9):e617–e626.
3 2. Methods
Data sources of basic vaccines (2). The countries featured in this report span three WHO regions: the African This report used data from the Demographic and Region; the South-East Asia Region; and the Eastern Health Surveys (DHS). The DHS are large-scale Mediterranean Region. For each priority country, nationally representative household surveys that the latest publicly available DHS was used. Table collect comparable information about population 2.1 presents the list of countries included in this health indicators and their socioeconomic context report, the year of DHS or NFHS fieldwork and the (1). The DHS use standardized questionnaires WHO region. administered to a representative sample of women aged 15–49 years, focusing primarily on maternal Childhood immunization indicator and child health. The DHS use a multistage cluster design. In India, the DHS is referred to The childhood immunization indicator featured as the National Family Health Survey (NFHS), in this report is receipt of the third dose of the which follows the same general protocol and combined diphtheria, tetanus toxoid and pertussis methodology. vaccine (DTP3). Coverage of DTP3 immunization was defined as: the percentage of children aged Priority countries 12–23 months who had received three doses of the combined DTP vaccine at the time of the survey. This report provides an in-depth analysis for a DTP3 immunization coverage was calculated set of 10 priority countries identified by Gavi, considering all DTP-containing vaccines, including the Vaccine Alliance as tier-1 countries. These pentavalent, tetravalent and/or DTP only. The lower countries together account for more than 70% of limit of this age range is the youngest age by which children worldwide who do not get a full course children should have received three doses, and
TABLE 2.1. Priority countries included in this report
No. Country DHS/NFHS fieldwork yeara WHO region 1 Afghanistan 2015 Eastern Mediterranean 2 Chad 2014 African 3 Democratic Republic of the Congo 2013 African 4 Ethiopia 2016 African 5 India 2015 South-East Asia 6 Indonesia 2012 South-East Asia 7 Kenya 2014 African 8 Nigeria 2013 African 9 Pakistan 2012 Eastern Mediterranean 10 Uganda 2016 African a In countries where the DHS or NFHS fieldwork spanned two years, this indicates the first year of fieldwork. These countries are: Chad (2014–2015), Democratic Republic of the Congo (2013–2014), India (2015–2016) and Pakistan (2012–2013).
4 2. Methods
the upper limit is the oldest age by which they Data were disaggregated by characteristics should have received three doses. DTP3 coverage of the child, the mother and the household, as is considered a key indicator of immunization well as geographic characteristics (Table 2.2). programme performance, and has been widely For most factors, subgroups were constructed applied for national and global monitoring (3). using comparable definitions and categorizations across countries. For example, subgroups based The DTP3 coverage indicator was constructed on mother’s age at birth were grouped as 15–19 based on the information on immunization collected years, 20–34 years and 35–49 years in all countries. from health cards as well as from mothers’ reports. Inequality according to household economic status According to DHS protocols, if the immunization was evaluated by comparing the level of coverage card was available, the interviewer copied the across five quintiles of a wealth index. (The wealth vaccination dates directly to the questionnaire. index was constructed using information on assets When there was no immunization card, or if ownership, housing characteristics, and access to the vaccine had not been recorded on the card, services (4)). For mother’s education, subgroups the respondent was asked to verbally report the reflected the highest level of education attended vaccines administered to the child. (a standard DHS variable). Due to small sample sizes in some countries, the two highest education Analysis approaches categories were sometimes combined as “secondary school or more”. The report presents the results of two types of analyses: descriptive analysis of inequalities Country-specific subgroup categorizations were by background characteristics (including data applied for mother’s ethnicity, caste/tribe and disaggregation and calculation of socioeconomic subnational region. More information about these inequality using summary measures of inequality); variables are included in the country context and multiple regression analysis to examine section of each chapter, and the DHS or NFHS associations of DTP3 with each characteristic final country reports (5). In countries where controlling for the others. For all analyses, survey information on mother’s ethnicity or caste/tribe sampling design, including sample weights and was available, the categorization defined by the clustering effects, were taken into account when DHS or NFHS survey was used. In the Democratic calculating point estimates and their confidence Republic of the Congo, Ethiopia and Kenya, specific intervals. categories were combined in one category defined as “other” and only larger ethnic groups were For more information about the analysis presented individually. For subnational regions, approaches, please see the accompanying unless otherwise noted, the data presented in technical note and glossary: this report reflect the subnational administrative http://www.who.int/gho/health_equity/ divisions that were current at the time of the report_2018_immunization/technical_notes.pdf survey and for which data were collected. In the Democratic Republic of the Congo and Indonesia, the subnational divisions have changed since the Descriptive overview time of the survey. In Chad, Kenya and Uganda, Disaggregation by background characteristics some of the divisions were combined. In Pakistan, To assess inequality in DTP3 immunization coverage the DHS was not administered in two subnational within each country, the report first presents the level regions, and therefore they were also not included of coverage among subgroups of the population. in this report.
5 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
TABLE 2.2. Background characteristics and associated subgroups applied to describe within-country inequality
Characteristics Subgroups Child characteristics Child’s sex Female Male Birth order 1st born 2nd–3rd born 4th–5th born 6th born or higher Mother characteristics Mother’s age at birth 15–19 years 20–34 years 35–49 years Mother’s educationa No education Primary school Secondary school More than secondary school Mother’s ethnicity or caste/tribeb Country-specific – see each country chapter Household characteristics Sex of household head Female Male Household economic status Quintile 1 (poorest) Quintile 2 Quintile 3 Quintile 4 Quintile 5 (richest) Geographic characteristics Place of residence Rural Urban Subnational region Country-specific – see each country chapter a In Afghanistan, Chad, the Democratic Republic of the Congo, Ethiopia and Kenya, the combined category of “secondary school or more” was applied due to small sample sizes of women with higher levels of education. b Mother’s ethnicity was applied in four countries: Afghanistan, Democratic Republic of the Congo, Ethiopia and Kenya; caste/tribe was applied in India.
Wealth- and education-related inequalities were calculated.) The ECI is a summary measure To further assess within-country inequality in DTP3 of absolute inequality that indicates the extent to immunization coverage, this report focused on two which DTP3 immunization coverage is concentrated background characteristics: household economic among the rich (or the poor) or among the more status and mother’s education. The report used educated (or less educated). The ECI is a variant the concentration index and concentration curve to of the generalized concentration index that shows provide a more complete analysis of how coverage a specific adjustment applicable for binary health varied across the full distribution of household indicators whose average is bounded between zero wealth and mother’s education (6,7). and one (8–10). The ECI has a negative value when the health indicator is concentrated among the poor To quantify the magnitude of inequality in DTP3 (or less educated); and it has a positive value when immunization coverage in a single numerical the health indicator is concentrated among the rich value, the Erreygers normalized concentration (or more educated). When there is no inequality, index (ECI) was calculated. (In this report, both the ECI value is 0. ECI values have a possible range wealth-related and education-related ECI values from -1 to +1.
6 2. Methods
The concentration curve is a graphical represen- Multiple regression analysis tation of how DTP3 coverage is distributed A logistic regression model was used to assess across subgroups with an inherent ordering, such the adjusted associations between DTP3 as household economic status. In the case of immunization coverage and selected demographic, household economic status, the concentration socioeconomic and geographic characteristics. curve plots the cumulative share of DTP3 coverage The multiple regression framework establishes against the cumulative share of children ranked the association of each characteristic with DTP3 from poorest to richest. In a situation of perfect coverage, accounting for other characteristics. The equality, the concentration curve will be a 45-degree development of the framework was guided by a diagonal line (line of equality); that is, the share literature review and subject to data availability. of immunization coverage exactly matches the share of children belonging to each wealth rank. Results of the regression analysis were reported If DTP3 coverage is lower among poorer children, using adjusted odds ratios. Odds ratios were used the concentration curve will lie below the line of to compare the relative odds (that is, the chance) equality; if higher among poorer children, the curve of DTP3 immunization of one subgroup against a will lie above the line of equality. The further the reference subgroup (Table 2.3). An odds ratio equal concentration curve lies from the equality line, the to 1 means that the factor does not affect the odds greater the inequality. of DTP3 immunization. An odds ratio greater than 1 indicates that the characteristic is associated with higher odds of immunization compared to
TABLE 2.3. Background characteristics included in the multiple regression analyses
Characteristics Reference and other subgroups Child’s sex Reference subgroup: male Birth order Reference subgroup: 6th born or higher Other subgroups included as dummy variables: 1st born; 2nd–3rd born; 4th–5th born Mother's age at birth Reference subgroup: 15–19 years Other subgroups included as dummy variables: 20–34 years; 35–49 years Mother's educationa Reference subgroup: no education Other subgroups included as dummy variables: primary school; secondary school; more than secondary school Mother’s ethnicity or caste/tribeb Reference subgroup: country-specific – see each country chapter Other subgroups included as dummy variables: country-specific – see each country chapter Sex of household headc Reference subgroup: male Household economic status Reference subgroup: quintile 1 Other subgroups included as dummy variables: quintiles 2–5 Place of residence Reference subgroup: rural Subnational region Reference subgroup: country-specific – see each country chapter Other subgroups included as dummy variables: country-specific – see each country chapter a In Afghanistan, Chad, the Democratic Republic of the Congo, Ethiopia and Kenya, the combined category of “secondary school or more” was applied due to small sample sizes of women with higher levels of education. b Mother’s ethnicity was applied in two countries: Democratic Republic of the Congo and Kenya; caste/tribe was applied in India. c Sex of household head was not included for Afghanistan, because nearly all households (99%) were headed by males.
7 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
the reference subgroup; an odds ratio less than 1 through the regional variation in different ethnic indicates lower odds of immunization compared to subgroups. So, ethnicity was dropped from the the reference subgroup. For example, for place of logistic regressions in these countries, although residence, the reference subgroup was rural: if the ethnicity was used in the descriptive analysis. odds ratio was more than 1, the chance of coverage was higher in urban areas, and if the odds ratio was Interactive visuals and tables less than 1, the chance of coverage was less likely in urban areas. Interactive visuals and tables accompany this report, enabling further exploration of patterns in the data, The statistical inference of the odds ratio was and customized visual outputs. The interactive determined by the 95% confidence intervals and visuals and tables convey additional information the p-value of the test of statistical significance. about the underlying data to contextualize the The significance level was considered as 0.05: interpretation of the report findings and make the if the p-value was below 0.05, the association results more transparent. Information included in was considered statistically significant. P-values the interactive visuals covers: national averages, between 0.05 and 0.10 were considered to be disaggregated data, ECI values and odds ratios. nearly significant. Joint hypothesis testing was 95% confidence intervals are reported for all also used to test whether the characteristics with estimates. For disaggregated data, the population multiple subgroups (e.g. regions, etc.), taken as share for each population subgroup and sample a whole, were significantly associated with DTP3 size limitations are presented. For odds ratios, the immunization coverage. p-values as well as information about the logistic regression model (number of observations used in In addition to exploring the adjusted associations the model, pseudo R2 and Wald Chi2) are listed. between background characteristics and DTP3 To access the interactive visuals referenced in this coverage, the report analysed how the relationship report, see the QR codes and URLs that appear between these variables and DTP3 coverage at the end of each country chapter. To access the changed after accounting for interaction effects. interactive tables, see Appendix 1. For each country, interactions were individually tested between: mother’s education and household References wealth; mother’s age at birth and household wealth; 1. The DHS Program. Demographic and Health Surveys mother’s age at birth and education; and place [Internet]. Washington (DC): US Agency for of residence and household wealth. For almost International Development; 2018 (dhsprogram.com, all combination cases, interactions were not accessed 21 May 2018). statistically significant on whole. 2. Targeted country assistance. [Internet]. Geneva: Gavi, the Vaccine Alliance; 2018 (https://www.gavi.org/ support/pef/targeted-country-assistance/, accessed Multicollinearity was also tested. This is the 21 May 2018). condition where a very high correlation exists 3. Wallace AS, Ryman TK, Dietz V. Overview of between two or more variables included in the global, regional, and national routine vaccination multiple regression model, resulting in unstable coverage trends and growth patterns from 1980 to estimates (in our case, odds ratios). In Afghanistan 2009: implications for vaccine-preventable disease and Ethiopia, ethnicity showed collinearity (with eradication and elimination initiatives. J Infect Dis. 2014 subnational region). The model including ethnicity November;210(Suppl. 1):S514–S22. in both countries demonstrated that the variable 4. Rutstein SO, Johnson K. The DHS wealth index. was not statistically significant and the associations Calverton: ORC Macro; 2004. between DTP3 and ethnicity were mainly explained
8 2. Methods
5. The DHS Program. DHS Final Reports [Internet]. 8. Erreygers G. Correcting the concentration index. J Health Washington (DC): US Agency for International Econ. 2009 March;28(2):504–15. Development; 2018 (https://dhsprogram.com/ 9. Erreygers G, Van Ourti T. Measuring socioeconomic publications/publication-search.cfm?type=5, accessed inequality in health, health care and health financing 21 May 2018). by means of rank-dependent indices: a recipe for good 6. O’Donnell O, van Doorslaer E, Wagstaff A, Lindelow practice. J Health Econ. 2011 July;30(4):685–94. M. Analyzing health equity using household survey 10. O’Donnell O, O’Neill S, Van Ourti T, Walsh B. data. A guide to techniques and their implementation. conindex: estimation of concentration indices. Stata J. Washington (DC): World Bank; 2008. 2016;16(1):112. 7. Handbook on health inequality monitoring: with a special focus on low- and middle-income countries. Geneva: World Health Organization; 2013:126.
9 3. Afghanistan
Country context The improvement of immunization services, including revising the Expanded Programme on Afghanistan has a population of 34 million, with Immunization policy and strengthened monitoring 15.5% under 5 years of age (1). The country consists mechanisms, is a strategic area of action for child of 34 provinces, and major ethnic groups include health in Afghanistan (4). Expanding coverage and Nuristani, Pashtun, Tajik, Hazara, Uzbek, Turkmen, addressing inequity are addressed in immunization Baloch and Pashai. The under-5 mortality rate in forums and country plans. Afghanistan has improved over the past decades, from 177.3 deaths per 1000 live births in 1990 Descriptive overview to 70.4 deaths per 1000 live births in 2016 (2); however, despite progress, the country has not yet Disaggregation by background achieved the Millennium Development Goal target characteristics of 59 (3). Key risk factors for child health include Nationally, the coverage of DTP3 immunization was poverty, underemployment, natural disasters, 58% among one-year-olds. Figure 3.1 shows the conflict migration, compromised security situation, coverage of DTP3 immunization in Afghanistan by urbanization and certain social norms (4). population subgroups.
The Expanded Programme on Immunization began There was no variation in coverage according to operating in Afghanistan in 1978 (5). Expanded child sex. Coverage varied somewhat according Programme on Immunization activities have to birth order, ranging from 54% among 2nd–3rd been implemented in the majority of districts, born children to a maximum of 63% among 1st and are managed through a three-tier system born children. with responsibilities coordinated across national, provincial and district levels (6). The National DTP3 coverage showed a substantial gradient Immunization Program in Afghanistan delivers by mothers’ characteristics. Compared to the services through facility-based, outreach and mobile 15–19 years subgroup, coverage increased by 9 approaches (6). Routine immunization is a core percentage points (from 50% to 59%) in the 20–34 component of the Basic Package of Health Services, years subgroup, and an additional 5 percentage a strategy to bolster access to health services in points in the 35–49 years subgroup (64%). DTP3 rural populations (7). While the introduction of this coverage also increased consistently with mother’s package has benefited vaccine delivery, challenges education, whereby coverage was notably lower and service delivery gaps persist (8). for the no education subgroup (54%) compared to the secondary school or more subgroup (80%). Childhood immunization in the country remains low There was a substantial difference in immunization overall, with national estimates falling well below coverage across ethnic subgroups: coverage ranged coverage goals of 90% (6). Inequality is a concern in from 1% in Nuristani populations to 71% in Uzbek Afghanistan, as childhood immunization is variable populations. across geographic and demographic subgroups. For instance, immunization coverage is lower in areas Female-headed households reported higher with security concerns (5), and among the poorest (9) immunization coverage (69%) than male-headed and those living farther from health facilities (4,10). households (58%). Afghanistan demonstrated a
10 3. Afghanistan
FIGURE 3.1. DTP3 immunization coverage among one-year-olds in Afghanistan, disaggregated by background characteristics (DHS 2015)
Characteristic Subgroup Characteristic Subgroup Child's sex Female 58.2 Subnational Nooristan 0.7 58.3 region 2.1 Male 58.4 Urozgan Kandahar 24.8 Birth order 1st born 63.1 Zabul* 26.8 2nd-3rd born 54.4 Ghor 33.0 4th-5th born 58.9 Farah 33.3 6th born or higher 59.3 Samangan 38.5 Mother's age 15-19 years 49.7 Ghazni 39.1 at birth 20-34 years 59.0 Logar 41.5
35-49 years 64.0 Kunarha 42.7 44.5 Mother's No education 54.1 Daykundi education Sar-e-pul 45.0 Primary school 70.4 Paktya 45.2 Secondary school or more 79.8 Khost 49.3 Mother's Baloch 44.4 Helmand 49.9 ethnicity 60.4 Hazara Baghlan 52.7 Nuristani 0.7 Balkh 53.0
Pashai 43.5 Kunduz 57.9
Pashtun 51.7 Panjsher 59.0 Jawzjan 61.3 Tajik 64.6 Badghis 61.4 Turkmen 57.2 Herat 62.2 Uzbek 70.8 Laghman 65.5 Other 48.6 Takhar 66.5 Sex of Female 68.7 Nimroz 69.2 household head Male 58.2 Kapisa 71.6
Household Quintile 1 (poorest) 48.9 Kabul 73.1 economic 76.1 status Quintile 2 52.1 Parwan Paktika 77.1 Quintile 3 54.4 Bamyan 78.3 Quintile 4 64.7 Nangarhar 78.6 Quintile 5 (richest) 70.4 Wardak 80.0 55.2 Place of Rural Faryab 81.9 residence Urban 68.2 Badakhshan 82.4
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average. *Estimate was based on 25–49 cases. gradient of increasing coverage across household The immunization coverage of children from urban wealth quintiles. Coverage increased moderately in areas (68%) was 13 percentage points higher the three poorest quintiles (from 49% in quintile than children from rural areas (55%). Coverage 1 to 54% in quintile 3) and demonstrated larger across regions varied substantially. Notably, the increases in the two richest quintiles (65% in gap between the regions with lowest coverage quintile 4 and 70% in quintile 5). (Nooristan and Urozgan, 1% and 2%, respectively) and those with the highest coverage (Badakhshan and Faryab, 82%) was over 80 percentage points.
11 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Wealth- and education-related inequalities related concentration curve for DTP3 immunization In Afghanistan, the wealth-related ECI for DTP3 coverage in Afghanistan is available in Chapter 13. immunization coverage had a positive value of 0.191, indicating that coverage was more concentrated Multiple regression analysis among richer households. The education-related ECI was 0.135, suggesting that coverage was higher Figure 3.2 presents the adjusted associations among those with more education. In both cases, between DTP3 immunization coverage and selected the ECIs were statistically significant. The wealth- socioeconomic, demographic and geographic
FIGURE 3.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Afghanistan, calculated as odds ratio (DHS 2015)
Characteristic Subgroup Characteristic Subgroup Subnational Nooristan (reference) Child's sex Male (reference) region Urozgan 3.26
Female 0.99 Zabul 31.21 Kandahar 39.45 Birth order 6th born or higher (reference) Logar 74.98 Ghazni 77.70 1.21 1st born Farah 94.68 Kunarha 105.23 2nd-3rd born 0.75 Ghor 110.32 Samangan 114.05 4th-5th born 0.92 Paktya 125.16
Mother's age 15-19 years (reference) Khost 130.06 at birth Sar-e-pul 135.89 20-34 years 1.72 Balkh 150.61 Helmand 157.58 35-49 years 1.91 Daykundi 165.72 Jawzjan 176.11 Mother's No education (reference) education Baghlan 186.96 Kunduz 209.86 Primary school 1.62 Kabul 216.00
Secondary school or more 1.87 Panjsher 238.28 Herat 243.18 Household Quintile 1 (poorest) (reference) Nimroz 251.98 economic status Laghman 265.41 Quintile 2 1.40 Takhar 341.18 Badghis 358.97 Quintile 3 1.76 Kapisa 376.89
Quintile 4 2.48 Nangarhar 410.28 Parwan 461.39 Quintile 5 (richest) 2.68 Faryab 559.56 Paktika 564.02 Place of Rural (reference) Wardak 702.37 residence Bamyan 808.77 Urban 1.21 Badakhshan 897.18
1 2 1 4 16 64 256 1,024 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
Note: Sex of household head was not included in this analysis because nearly all households (99%) were headed by males. 12 3. Afghanistan
4. Children and women in Afghanistan: a situation analysis factors. There was large variation in the odds of 2014. Kabul: UNICEF; 2014. childhood immunization across regions, adjusting 5. Mashal T, Nakamura K, Kizuki M, Seino K, Takano T. for other characteristics. For example, compared Impact of conflict on infant immunisation coverage in to Nooristan, the region with the lowest odds of Afghanistan: a countrywide study 2000–2003. Int J coverage, children living in Wardak, Bamyan or Health Geogr. 2007;6(1):23. 6. Islamic Republic of Afghanistan Ministry of Public Badakhshan were more than 700 times as likely Health. Comprehensive Multi-Year Plan for National to be covered. Immunization Program 2011–2015. Kabul: Government of Afghanistan; 2012. Immunization coverage was significantly associated 7. Islamic Republic of Afghanistan Ministry of Public with household economic status. For instance, the Health. A basic package of health services for Afghanistan, 2005/1384. Kabul: USAID Afghanistan; odds of DTP3 immunization for a one-year-old child 2005. in either of the two richest quintiles were around 8. Mbaeyi C, Kamawal NS, Porter KA, Azizi AK, Sadaat I, 2.5 times higher than those of a child in the poorest Hadler S et al. Routine immunization service delivery quintile. Mother’s age at birth was associated with through the basic package of health services program in Afghanistan: gaps, challenges, and opportunities. J childhood immunization: children of mothers aged Infect Dis. 2017 July 1;216(Suppl. 1):S273–S279. 35–49 years and 20–34 years had a probability 1.9 9. Akseer N, Bhatti Z, Rizvi A, Salehi AS, Mashal T, and 1.7 times higher, respectively, of being covered Bhutta ZA. Coverage and inequalities in maternal and than children of mothers aged 15–19 years. Taken child health interventions in Afghanistan. BMC Public Health [Internet]. 2016 September;16(S2) (http:// as a whole, the education variable was statistically bmcpublichealth.biomedcentral.com/articles/10.1186/ significant, indicating variation in DTP3 coverage s12889-016-3406-1, accessed 4 April 2018). across mother’s education: children whose mothers 10. Hemat S, Takano T, Kizuki M, Mashal T. Health-care had secondary school or more education were 1.9 provision factors associated with child immunization times more likely to be covered than those in the coverage in a city centre and a rural area in Kabul, Afghanistan. Vaccine. 2009 May;27(21):2823–9. no education subgroup. On whole, the birth order variable was also statistically significant.
The remaining factors presented non-significant associations with DTP3 coverage. For instance, other things equal, child’s sex and place of residence were not associated with immunization coverage.
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Interactive visuals and tables Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). Interactive visuals and tables allow for further exploration 2. Levels and trends in child mortality report 2017. of inequality in childhood immunization in Afghanistan. Estimates developed by the UN Inter-agency Group for Child Mortality Estimation. United Nations Children’s To access interactive visuals: Fund, World Health Organization, World Bank and United Nations [Internet]. New York: United Nations SCAN HERE: or VISIT: Children’s Fund; 2017 (http://www.childmortality.org/ files_v21/download/IGME%20report%202017%20 http://apps.who.int/gho/ data/view.wrapper.HE- child%20mortality%20final.pdf, accessed 21 February VIZ21?lang=en&menu=hide 2018). 3. Akseer N, Salehi AS, Hossain SM, Mashal MT, Rasooly MH, Bhatti Z et al. Achieving maternal and child health gains in Afghanistan: a Countdown to 2015 country case To access interactive tables, see Appendix 1. study. Lancet Glob Health. 2016;4(6):e395–e413.
13 4. Chad
Country context Descriptive overview
Chad has a population of 14 million, with 18.6% Disaggregation by background under 5 years of age (1). The country consists of characteristics 23 subnational regions. In this report, as per the National DTP3 immunization coverage was 34% protocol followed in the DHS, the regions of Borkou among one-year-olds in Chad. Figure 4.1 presents and Tibesti were grouped as one, as were the coverage of DTP3 immunization disaggregated by regions of Ennedi-Est and Ennedi-Ouest; thus, data child, mother and household characteristics, as well are presented for 21 subnational regions. In 1990, as geographical characteristics. the under-5 mortality rate was 210.8 deaths per 1000 live births, which dropped to 127.3 deaths per Immunization coverage was not different between 1000 live births in 2016 (2). The poor state of child boys and girls, and varied only slightly by birth health in Chad has been attributed to widespread order, with 6th or higher order births reporting 3 poverty, multifaceted inequalities, geopolitical percentage points higher coverage than 1st order instability and weak institutional capabilities (3). births.
Chad’s national Inter-Agency Coordinating The level of coverage across subgroups of mother’s Committee (ICC), chaired by the Minister of Health age at birth demonstrated a reverse U-shape, where and supported by other government ministries and coverage peaked in the 20–34 years subgroup United Nations partner organizations, is the high- (36%) and was lower for the 15–19 and 35–49 level decision-making body for matters concerning years subgroups (7 and 4 percentage points, childhood immunization (4). While the country has respectively). Chad presented a linear gradient a technical advisory group for polio, an independent in DTP3 coverage across education subgroups, national advisory committee on immunization has with a pattern of equivalent increases in coverage not yet been established (4). between the three education subgroups. Coverage in the secondary education or higher subgroup was The immunization coverage among one-year- double that of the no education subgroup (54% olds in Chad is very low (5). The expansion of the versus 27%, respectively). Reaching Every District (RED) strategy to 54 out of 90 health districts in recent years has made some Looking at household characteristics, coverage in headway in improving immunization coverage; the subgroup with female household heads was intensified efforts to reach nomadic populations, 8 percentage points lower than those with male and the movement of people to more secure areas household heads. Inequality according to economic such as camps, have also contributed to coverage status showed little variation in coverage between increases (4). Operational difficulties in the health quintiles 1 to 4, and a sharp increase of more than sector, as well as programme implementation 10 percentage points between quintiles 4 and 5. challenges and security issues have impeded progress in childhood immunization (4). Factors Children in urban areas had 8 percentage points that predict non-use of maternal health services higher coverage than children in rural areas (32% or childhood immunization include: rural residence; versus 40%). Regions also showed substantial low level of parental education; and having multiple variation in coverage, which fall into four groupings: children under 5 years of age in the family (6).
14 4. Chad
FIGURE 4.1. DTP3 immunization coverage among one-year-olds in Chad, disaggregated by background characteristics (DHS 2014)
Characteristic Subgroup Characteristic Subgroup 33.8 Child's sex Female 34.5 Subnational Chari Baguirmi 5.5 region
Male 33.1 Batha 7.9
Birth order 1st born 33.4 Borkou / Tibesti 8.1
2nd-3rd born 31.8 Wadi Fira 8.9
4th-5th born 32.4 Ouaddai 10.1
6th born or higher 36.5 Lac 17.8
Mother's age 15-19 years 28.7 Sila 18.7 at birth 20-34 years 35.5 Barh El Gazal 20.5
35-49 years 32.0 Ennedi 20.6
Mother's No education 27.3 Salamat 23.7 education
Primary school 41.0 Guera 25.6
Secondary school or more 54.0 Kanem 25.9
Sex of Female 27.2 Logone Occidental 28.3 household head Male 34.9 Hadjer-Lamis 31.0
Household Quintile 1 (poorest) 28.5 Logone Oriental 40.4 economic status Quintile 2 32.6 N'Djamena 41.4
Quintile 3 32.0 Moyen Chari 41.7
Quintile 4 33.6 Mayo Kebbi Est 42.6
Quintile 5 (richest) 45.0 Tandjile 45.1
Place of Rural 32.4 Mayo Kebbi Ouest 58.4 residence
Urban 40.1 Mandoul 62.8
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average. the first group contains five regions with coverage Wealth- and education-related inequalities levels up to 10%; the second group contains nine The wealth-related inequality of DTP3 immunization regions with coverage between 18% (Lac) and coverage, measured using the ECI, had a positive 31% (Hadjer-Lamis); the third group contains five value of 0.100, indicating that coverage was regions with coverage between 40% (Logone concentrated among wealthier households, but Oriental) and 45% (Tandjile); and, finally, the only to a limited extent. The education-related ECI fourth group contains the two regions with levels in Chad was 0.184, demonstrating a concentration of coverage around 60% (Mayo Kebbi Ouest and of coverage among the children of mothers with Mandoul). higher education. Both ECI values were statistically
15 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
significant. The wealth-related concentration socioeconomic, demographic and geographic curve for DTP3 immunization coverage in Chad is factors. The odds of DTP3 coverage differed available in Chapter 13. greatly across regions in Chad, with most regions presenting substantial and statistically significant Multiple regression analysis higher odds of coverage than the reference region of Chari Baguirmi. Notably, Mandoul and Mayo Kebbi Figure 4.2 presents estimated odds ratios and Ouest had 27 and 19 times higher odds of coverage confidence intervals for the adjusted associations than Chari Baguirmi, respectively. between DTP3 immunization coverage and selected
FIGURE 4.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Chad, calculated as odds ratio (DHS 2014)
Characteristic Subgroup Characteristic Subgroup
Child's sex Male (reference) Subnational Chari Baguirmi (reference) region Female 1.09 Borkou / Tibesti 1.16
Birth order 6th born or higher (reference) Batha 1.38
1st born 0.91 Wadi Fira 1.69
2nd-3rd born 0.74 Ouaddai 1.84
4th-5th born 0.83 Lac 3.51
Mother's age 15-19 years (reference) Ennedi 3.95 at birth 20-34 years 1.48 Barh El Gazal 4.13
35-49 years 1.15 Sila 4.16
Mother's No education (reference) Salamat 4.90 education Primary school 1.14 Guera 5.03
Secondary school or more 1.95 Kanem 5.25
Sex of Male (reference) N'Djamena 5.49 household head Female 1.00 Logone Occidental 5.71
Household Quintile 1 (poorest) (reference) Hadjer-Lamis 7.20 economic status Quintile 2 1.11 Moyen Chari 9.69
Quintile 3 1.12 Logone Oriental 10.24
Quintile 4 1.26 Mayo Kebbi Est 11.33
Quintile 5 (richest) 2.48 Tandjile 11.84
Place of Rural (reference) Mayo Kebbi Ouest 19.28 residence Urban 0.84 Mandoul 27.32
1 2 1 2 4 8 16 32 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
16 4. Chad
After controlling for other characteristics, the odds of immunization for children in the richest quintile was 2.5 times higher than for children in the poorest quintile. There was no evidence, however, of significant variation between coverage in quintile 1 and coverage in quintiles 2 to 4. Mother’s education was associated with DTP3 coverage, as the subgroup with secondary education or more was twice as likely to be covered as the no education subgroup. Compared to mothers aged 15–19 years, the odds of coverage were 1.5 times higher in the 20–34 years subgroup.
Other factors including child sex, sex of household head, place of residence and birth order demonstrated no associations with DTP3 immunization coverage, after controlling for background characteristics.
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). 2. Levels and trends in child mortality report 2017. Estimates developed by the UN Inter-agency Group for Child Mortality Estimation. United Nations Children’s Fund, World Health Organization, World Bank and United Nations [Internet]. New York: United Nations Children’s Fund; 2017 (http://www.childmortality.org/ files_v21/download/IGME%20report%202017%20 child%20mortality%20final.pdf, accessed 21 February 2018). 3. Chad country programme document 2012–2016 [Internet]. Report No.: E/ICEF/2011/P/L.22. New York: UNICEF; 2011 (https://www.unicef.org/about/ execboard/files/Chad_final_approved_2012-2016_20_ Interactive visuals and tables Oct_2011.pdf, accessed 6 April 2018). 4. Global Vaccine Action Plan: priority country reports Interactive visuals and tables allow for further exploration on progress towards GVAP-RVAP goals. Annex to the of inequality in childhood immunization in Chad. GVAP Secretariat Annual Report 2016. Geneva; World Health Organization; 2016. To access interactive visuals: 5. Global Health Observatory. Health Equity Monitor SCAN HERE: or VISIT: [Internet]. Geneva: World Health Organization; 2018 (http://www.who.int/gho/health_equity/en/, accessed http://apps.who.int/gho/ 4 April 2018). data/view.wrapper.HE- 6. Carlson M, Paintain LS, Bruce J, Webster J, Lines J. Who VIZ21?lang=en&menu=hide attends antenatal care and expanded programme on immunization services in Chad, Mali and Niger? The implications for insecticide-treated net delivery. Malar J. 2011;10(1):341. To access interactive tables, see Appendix 1.
17 5. Democratic Republic of the Congo
Country context with more-educated mothers and children from wealthier families (6,7). The likelihood of incomplete The Democratic Republic of the Congo has a or untimely immunization has been associated population of 76 million, with 18.5% under 5 years with clinic characteristics, which may be linked to of age (1). Prior to the adoption of their new whether user fees are charged and, if so, how they Constitution in 2006, the country consisted of 11 are structured (8). provinces; in 2015, the country was further divided into 26 subnational regions – 25 provinces and 1 Descriptive overview capital city region (2). (Note that this report contains data from the 2013 DHS, covering 11 provinces.) The Disaggregation by background under-5 mortality rate in the Democratic Republic characteristics of the Congo has improved from 183.6 deaths per National coverage of DTP3 immunization among 1000 live births in 1990 to 94.3 deaths per 1000 one-year-olds was 61% in the Democratic live births in 2016 (3). Republic of the Congo. Figure 5.1 presents DTP3 immunization coverage disaggregated by The Comprehensive Multi-Year Plan 2015–2019 background characteristics. outlines the current 5-year strategy for the Expanded Programme on Immunization in the Low levels of inequality in coverage were reported Democratic Republic of the Congo, including by child sex, birth order or mother’s age at birth; benchmarks and priorities to enhance vaccine however, coverage did vary by mother’s education, delivery throughout the country. The Expanded mother’s ethnicity and wealth quintile. For education Programme on Immunization is supported by an subgroups, those in the secondary education or extensive and well-trained team of staff, and most higher subgroup had substantially higher coverage vaccines are delivered by nurses at government (72%) than those in the less-educated subgroups, hospitals or health centres (4). Health workers where coverage was 56% and 52% in the no work with community volunteers, who inform education and primary education subgroups, the community about immunization services, respectively. DTP3 immunization coverage varied by and contribute to planning and implementing mother’s ethnicity, with coverage ranging from 43% immunization-related activities (4). in the Uele Lac Albert and Cuvette Central subgroups to 83% in the Bakongo Nord & Sud subgroup. While the country has sufficient numbers of health facilities, the delivery of immunization services Across wealth quintiles, there was a sharp gradient suffers from a widespread lack of refrigerators in DTP3 immunization coverage: coverage showed at health centres, frequent vaccine shortages gradually larger step-wise increases across and stockouts, and financial and administrative quintiles, spanning from 48% in the poorest to 83% delays (4,5). Additionally, low population demand in the richest. The difference in coverage based on for childhood immunization, inadequate health the sex of the household head was low. worker remuneration and a lack of supervision have contributed to many missed opportunities for Coverage among children in urban areas (74%) was immunization (4). Coverage has been reported to 20 percentage points higher than for children in be higher among children in urban areas, children rural areas (54%). There was variation in coverage
18 5. Democratic Republic of the Congo
FIGURE 5.1. DTP3 immunization coverage among one-year-olds in the Democratic Republic of the Congo, disaggregated by background characteristics (DHS 2013)
Characteristic Subgroup Characteristic Subgroup 60.5 Child's sex Female 60.9
Male 60.1 Subnational Equator 43.6 region Birth order 1st born 62.7
2nd-3rd born 59.2 Orientale 45.9 4th-5th born 62.1
6th born or higher 58.8
Mother's age 15-19 years 58.8 Maniema 47.6 at birth 20-34 years 61.1
35-49 years 59.3 Katanga 51.5 Mother's No education 56.2 education Primary school 52.0
Secondary school or more 71.6 Kasai-Oriental 53.3 Mother's Bakongo Nord & Sud 83.3 ethnicity Bas-Kasaï et Kwilu-Kwango 67.0
Basele-K, Man. et Kivu 74.9 Kasai-Occidental 54.1
Cuvette Central 43.1
Kasai, Katanga, Tanganyika 55.6 Bandundu 62.2 Ubangi et Itimbiri 46.3
Uele Lac Albert 42.6
Other 70.9 Sud-Kivu 75.3 Sex of Female 63.4 household head Male 59.7
Household Quintile 1 (poorest) 47.6 Bas-Congo 82.8 economic status Quintile 2 51.1
Quintile 3 56.2 Kinshasa 84.5 Quintile 4 69.3
Quintile 5 (richest) 83.0
53.9 Place of Rural Nord-Kivu 87.0 residence Urban 74.3
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average. across provinces, spanning 43 percentages points better-off. The education-related ECI summarized from Equator (44%) to Nord-Kivu (87%). inequality across mother’s education and demonstrated a value of 0.200. This result indicates Wealth- and education-related inequalities that DTP3 coverage was higher among children The wealth-related inequality of DTP3 immunization with more-educated mothers. Wealth-related and coverage in the Democratic Republic of the Congo education-related ECIs in the Democratic Republic showed a positive ECI value of 0.287, indicating of the Congo were statistically significant. The that coverage was more concentrated among the wealth-related concentration curve for DTP3
19 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
immunization coverage in the Democratic Republic showing significantly higher odds of DTP3 of the Congo is available in Chapter 13. immunization than Katanga. Notably, children in Nord-Kivu and Bas-Congo were more than 9 and Multiple regression analysis 7 times more likely to be covered than children in Katanga, respectively. Figure 5.2 shows the adjusted associations between DTP3 immunization coverage among one-year-olds Controlling for other factors, immunization coverage and selected socioeconomic, demographic and was strongly associated with economic status and geographic factors. Across provinces, the odds of mother’s ethnicity. The odds of receiving the third coverage differed, with the majority of provinces dose of the DTP vaccine were nearly 3 times higher
FIGURE 5.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in the Democratic Republic of the Congo, calculated as odds ratio (DHS 2013)
Characteristic Subgroup Characteristic Subgroup Child's sex Male (reference)
Female 0.99 Subnational Katanga (reference) region Birth order 6th born or higher (reference)
1st born 0.88 Maniema 1.31 2nd-3rd born 0.76
4th-5th born 0.90
Mother's age 15-19 years (reference) Kasai-Oriental 1.47 at birth 20-34 years 0.99
35-49 years 0.85 Kasai-Occidental 1.66 Mother's No education (reference) education Primary school 0.85
Secondary school or more 1.44 Equator 2.88 Mother's Uele Lac Albert (reference) ethnicity Bakongo Nord & Sud 2.87
Bas-Kasaï et Kwilu-Kwango 2.94 Kinshasa 3.09
Basele-K, Man. et Kivu 3.30
Cuvette Central 1.31 Bandundu 3.22 Kasai, Katanga, Tanganyika 3.53
Ubangi et Itimbiri 1.43
Other 3.86 Orientale 3.28 Sex of Male (reference) household head Female 1.21
Household Quintile 1 (poorest) (reference) Sud-Kivu 4.33 economic status Quintile 2 1.06
Quintile 3 1.13 Bas-Congo 7.45 Quintile 4 1.73
Quintile 5 (richest) 2.89
Place of Rural (reference) Nord-Kivu 9.15 residence Urban 1.25
1 2 4 1 2 4 8 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
20 5. Democratic Republic of the Congo
7. State of inequality: childhood immunization. Geneva: for children in the richest quintile than for children World Health Organization; 2016. in the poorest quintile. Mother’s ethnicity was a 8. Zivich PN, Kiketa L, Kawende B, Lapika B, Yotebieng relevant factor associated with DTP3 immunization M. Vaccination coverage and timelines among children coverage. Compared to children in the Uele 0–6 months in Kinshasa, the Democratic Republic of the Congo: a prospective cohort study. Matern Child Lac Albert subgroup, the following ethnicity Health J. 2017 May;21(5):1055–64. subgroups were three or more times as likely to be covered: Bas-Kasaï et Kwilu-Kwango; Basele-K, Man. and Kivu; and Kasai, Katanga, Tanganyika. After controlling for other factors, children in the secondary education or higher subgroup were 1.4 times more likely to be covered than those in the no education subgroup (the association was near significant).
The other factors demonstrated non-significant associations with DTP3 immunization coverage. After controlling for other characteristics, child’s sex, birth order, mother’s age at birth, sex of the household head and place of residence were not significantly associated with coverage.
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). 2. The Constitution of the Democratic Republic of the Congo, 2005. Kinshasa: Democratic Republic of the Congo; 2005. 3. Levels and trends in child mortality report 2017. Estimates developed by the UN Inter-agency Group for Child Mortality Estimation. United Nations Children’s Fund, World Health Organization, World Bank and United Nations [Internet]. New York: United Nations Children’s Fund; 2017 (http://www.childmortality.org/ Interactive visuals and tables files_v21/download/IGME%20report%202017%20 child%20mortality%20final.pdf, accessed 21 February Interactive visuals and tables allow for further exploration 2018). of inequality in childhood immunization in the Democratic 4. Global Vaccine Action Plan: priority country reports on progress towards GVAP-RVAP goals. Annex to the Republic of the Congo. GVAP Secretariat Annual Report 2016. Geneva; World To access interactive visuals: Health Organization, 2016. 5. Gerard SP, Kyrousis E, Zachariah R. Measles in the SCAN HERE: or VISIT: Democratic Republic of the Congo: an urgent wake-up call to adapt vaccination implementation strategies. http://apps.who.int/gho/ Public Health Action. 2014 March 21;4(1):6–8. data/view.wrapper.HE- VIZ21?lang=en&menu=hide 6. Ashbaugh HR, Hoff NA, Doshi RH, Alfonso VH, Gadoth A, Mukadi P et al. Predictors of measles vaccination coverage among children 6–59 months of age in the Democratic Republic of the Congo. Vaccine. 2018 To access interactive tables, see Appendix 1. January;36(4):587–93.
21 6. Ethiopia
Country context and communities, those with higher levels of parental education, and those residing in certain geographic Ethiopia has a population of about 100 million, areas (7–9). Sustained and accelerated improvements with 14.9% under 5 years of age (1). The country in coverage are impeded by: difficulties reaching is comprised of 11 subnational regions, including nomadic and remote communities; vaccine supply nine regional states and two chartered cities. chain and cold-chain issues; irregular or infrequent Between 1990 and 2016, the rate of under-5 availability of immunization services at health mortality in Ethiopia reduced by more than two facilities; and low monitoring and supervision thirds, from 203.2 deaths per 1000 live births in capacity (10). Ethiopia is prone to natural and 1990 to 58.4 deaths per 1000 live births in 2016 human-induced disasters (including droughts, (2). Ethiopia surpassed the child health Millennium floods, earthquakes, etc.) which may interrupt health Development Goal 4, largely due to the rapid services including immunization. development of community-level primary health care and the scale-up of high-impact interventions Descriptive overview across health, nutrition and other development sectors (3). Disaggregation by background characteristics The Expanded Programme on Immunization was In 2016, DTP3 immunization coverage among one- first launched in Ethiopia in 1980 with six antigens, year-olds in Ethiopia was 53%. Figure 6.1 presents followed by the progressive introduction of new and coverage of DTP3 immunization in Ethiopia underutilized vaccines; currently, over 10 antigens disaggregated by socioeconomic, demographic are included in routine immunization programme. and geographic characteristics. Ethiopia’s national immunization programme Comprehensive Multi-Year Plan 2016–2020 aims There was no difference in coverage between to achieve coverage levels of at least 90% nationally females and males (53%), or between 1st and and 80% in every district by 2020 (4). The regional 2nd–3rd born children (59%). Children of higher governments are the primary implementers of birth orders, however, demonstrated lower levels of immunization programmes, with financial, technical coverage, with just 43% coverage among 6th born and material support from international and local or higher children. partners, and operational support from health officers (nurses), and community-based health Considering mothers’ characteristics, coverage extension workers. The Routine Immunization was lowest in the 15–19 years subgroup (44%) Improvement Plan, developed following a large and was highest among the 20–34 years subgroup national survey in 2012, has been successful in (55%); for mothers aged 35–49 years, coverage increasing immunization coverage in the 51 poorest- was 51%. There was a 17 percentage point increase performing zones (5). in DTP3 immunization coverage moving from the no education (45%) to the primary school (62%) Immunization coverage among one-year-olds in subgroups, and an 18 percentage point increase Ethiopia varies across population subgroups (6). moving from the primary to the secondary school Immunization coverage was found to be higher or more (80%) subgroups. The ethnicity-related among those children from wealthier households inequality demonstrated a substantial difference
22 6. Ethiopia
FIGURE 6.1. DTP3 immunization coverage among one-year-olds in Ethiopia, disaggregated by background characteristics (DHS 2016)
Characteristic Subgroup Characteristic Subgroup 53.2 Child's sex Female 53.3 Subnational Affar 20.1 Male 52.9 region Birth order 1st born 59.4
2nd-3rd born 58.7 Somali 36.3
4th-5th born 53.4
6th born or higher 43.0 Oromiya 39.9 Mother's age 15-19 years 43.6 at birth 20-34 years 55.3
35-49 years 50.7 Gambela 54.8
Mother's No education 45.3 education Primary school 62.3 Harari 58.7 Secondary school or more 79.8
Mother's Affar 21.7 ethnicity Amhara 66.4 SNNPR 59.0
Oromo 40.6
Somalie 37.1 Amhara 63.8 Tigray 81.9
Other 57.4
Sex of Female 45.3 Benishangul-Gumuz 76.2 household head Male 54.5
Household Quintile 1 (poorest) 36.4 economic Tigray 81.4 status Quintile 2 50.4
Quintile 3 51.4
Quintile 4 63.2 Dire Dawa 84.9
Quintile 5 (richest) 76.3
Place of Rural 49.7 95.7 residence Addis Ababa Urban 79.5
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average. of 60 percentage points, from 22% in the Affar Immunization coverage was 30 percentage points population to 82% in the Tigray population. higher among children in urban areas (80%) than children in rural areas (50%). Across regions, In Ethiopia, inequality according to household coverage varied markedly: while less than half of economic status presented a gradient pattern, with one-year-olds received three doses of DTP in Affar higher levels of coverage in wealthier households. (20%), Somali (36%) and Oromiya (40%), the Coverage in the richest quintile (76%) was more DTP immunization coverage was more than 80% in than double the level of coverage in the poorest Tigray (81%), Dire Dawa (85%) and Addis Ababa quintile (36%). (96%).
23 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Wealth- and education-related inequalities for DTP3 immunization coverage in Ethiopia is The ECI for Ethiopia showed a positive value available in Chapter 13. of 0.296, indicating that coverage was largely concentrated among wealthier households. The ECI Multiple regression analysis for education-related inequality in Ethiopia showed a positive value of 0.220, indicating that DTP3 Figure 6.2 shows the adjusted associations between immunization coverage was higher among children DTP3 immunization coverage among one-year-olds with more-highly educated mothers. Both wealth- and socioeconomic, demographic and geographic related and education-related ECIs were statistically factors. The association between subnational significant. The wealth-related concentration curve regions and DTP3 immunization coverage was
FIGURE 6.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Ethiopia, calculated as odds ratio (DHS 2016)
Characteristic Subgroup Characteristic Subgroup
Child's sex Male (reference) Subnational Affar (reference) region Female 0.93
Birth order 6th born or higher (reference) Oromiya 1.99 1st born 1.55
2nd-3rd born 1.52 Somali 2.74
4th-5th born 1.45
Mother's age 15-19 years (reference) Gambela 2.82 at birth 20-34 years 2.08
35-49 years 2.09 Harari 2.89
Mother's No education (reference) education
Primary school 1.82 SNNPR 4.07
Secondary school or more 1.66
Sex of Male (reference) Amhara 5.12 household head Female 0.57
Benishangul-Gumuz 11.83 Household Quintile 1 (poorest) (reference) economic status Quintile 2 1.57
Dire Dawa 13.31 Quintile 3 1.68
Quintile 4 2.45
Tigray 13.47 Quintile 5 (richest) 2.74
Place of Rural (reference) residence Addis Ababa 19.95 Urban 1.50
1 2 1 2 4 8 16 32 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
24 6. Ethiopia
strong and statistically significant. Compared to 4. Ethiopia National Expanded Programme on Affar, the odds of coverage were more than 10 times Immunization: Comprehensive Multi-Year Plan 2016– 2020. Addis Ababa: Ethiopia Federal Ministry of Health; higher in four regions (Addis Ababa, Tigray, Dire 2015. Dawa and Benishangul-Gumuz). 5. Thematic briefing note: immunization [Internet]. UNICEF Ethiopia Country Office; 2016 (https://www. Immunization coverage was associated with unicef.org/ethiopia/ECO_Thematic_Briefing_Note_ Immunization.pdf, accessed 5 April 2018). economic status and mother’s education, after 6. Global Health Observatory. Health Equity Monitor controlling for other characteristics. Compared [Internet]. Geneva: World Health Organization; 2018 to children in the poorest quintile, the odds of (http://www.who.int/gho/health_equity/en/, accessed receiving the DTP3 vaccine were 2.5 times higher 4 April 2018). in quintile 4, and 2.7 times higher in quintile 5. 7. Abadura SA, Lerebo WT, Kulkarni U, Mekonnen ZA. Individual and community level determinants of Children in the primary education subgroup were childhood full immunization in Ethiopia: a multilevel 1.8 times more likely to be covered than children in analysis. BMC Public Health [Internet]. 2015 December the no education subgroup. (http://bmcpublichealth.biomedcentral.com/ articles/10.1186/s12889-015-2315-z, accessed 5 April 2018). There was a positive relationship between 8. Lakew Y, Bekele A, Biadgilign S. Factors influencing immunization coverage and mother’s age at birth: full immunization coverage among 12–23 months of compared to the 15–19 years subgroup, the odds of age children in Ethiopia: evidence from the national coverage were twice as high among both the 20–34 demographic and health survey in 2011. BMC Public Health [Internet]. 2015 December (http:// years and the 35–49 years subgroups (though only bmcpublichealth.biomedcentral.com/articles/10.1186/ nearly significant in the 35–49 years subgroup). s12889-015-2078-6, accessed 5 April 2018). One-year-olds belonging to households headed 9. Skaftun EK, Ali M, Norheim OF; Thorne C, editor. by a female had a significantly lower chance (0.6 Understanding inequalities in child health in Ethiopia: health achievements are improving in the period 2000– times) of being covered. 2011. PLoS ONE. 2014 August 28;9(8):e106460. 10. Global Vaccine Action Plan: priority country reports Once background characteristics were accounted on progress towards GVAP-RVAP goals. Annex to the for, child’s sex, birth order and place of residence GVAP Secretariat Annual Report 2016. Geneva: World showed no significant association with DTP3 Health Organization; 2016. immunization coverage in Ethiopia.
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). Interactive visuals and tables 2. Levels and trends in child mortality report 2017. Interactive visuals and tables allow for further exploration Estimates developed by the UN Inter-agency Group for of inequality in childhood immunization in Ethiopia. Child Mortality Estimation. United Nations Children’s Fund, World Health Organization, World Bank and To access interactive visuals: United Nations [Internet]. New York: United Nations Children’s Fund; 2017 (http://www.childmortality.org/ SCAN HERE: or VISIT: files_v21/download/IGME%20report%202017%20 child%20mortality%20final.pdf, accessed 21 February http://apps.who.int/gho/ 2018). data/view.wrapper.HE- VIZ21?lang=en&menu=hide 3. Ruducha J, Mann C, Singh NS, Gemebo TD, Tessema NS, Baschieri A et al. How Ethiopia achieved Millennium Development Goal 4 through multisectoral interventions: a Countdown to 2015 case study. Lancet To access interactive tables, see Appendix 1. Glob Health. 2017;5(11):e1142–e1151.
25 7. India
Country context lack of evidence for planning and research activities); and human resource shortages in management, India has a population of 1.309 billion, with 9.3% research and operations at all levels (4). under 5 years of age (1). The country consists of 29 states and 7 union territories, including the National In India, the National Family Health Survey (NFHS) Capital Territory of Delhi. The under-5 mortality is equivalent to the DHS conducted in other rate in India has fallen from 125.9 deaths per 1000 countries. live births in 1990 to 43.0 deaths per 1000 live births in 2016 (2). The high burden of childhood Descriptive overview mortality and morbidity in India reflects the poor quality of public health care in India, with uneven Disaggregation by background progress between subnational regions, and on the characteristics basis of socioeconomic status and sex (3). According to the 2015 NFHS, the DTP3 immunization coverage among one-year-olds in India was 79%. In 1978, India launched its Expanded Programme Figure 7.1 presents DTP3 immunization coverage in on Immunization, which was renamed as the India by population subgroups. Universal Immunization Program in 1985. The Universal Immunization Program is embedded Sex-related inequality was non-existent, as male as part of India’s National Health Policy, and the and female children presented the same level of National Vaccine Policy of 2011 provides broad coverage (79%). There was a gradient in coverage policy guidelines and frameworks to guide decision- according to birth order, with higher order births making in areas such as research and development, showing lower coverage. vaccine quality assessment, institutional processes, vaccine introduction issues and regulatory issues Looking at mother’s characteristics, the coverage (4). The Government of India launched Mission of DTP3 immunization was the same for the 15–19 Indradhanush in 2014 as a campaign to accelerate years and 20–34 years subgroups (79%), while progress towards achieving 90% full immunization coverage was lower in the 35–49 years subgroup coverage by 2020 (5). Initially targeting 201 (70%). A gradient in immunization coverage was low-performing districts, the campaign has observed across mother’s education level, with subsequently been expanded to over 350 districts, increasing coverage in more-educated subgroups. and has been successful in improving coverage in The gap between the no education subgroup and the underserved areas (6). subgroup with more than secondary education was 18 percentage points. There were small differentials Immunization coverage in India demonstrates in DTP3 coverage by mother’s caste/tribe: coverage multidimensional types of inequality, with variation was higher among those in the scheduled caste, across subgroups defined by social class, parental other backward class or other subgroups (coverage education, religion, wealth status, place of residence, around 80%), whereas coverage was lower in the and other individual and family characteristics (7–9). scheduled tribe subgroup (74%). Key challenges related to immunization activities include: weak health information systems and low There was no difference in coverage on the basis capacity for monitoring and evaluation (resulting in a of the sex of the household head. A gradient was
26 7. India
FIGURE 7.1. DTP3 immunization coverage among one-year-olds in India, disaggregated by background characteristics (NFHS 2015)
Characteristic Subgroup Characteristic Subgroup 78.8 Child's sex Female 78.8 Subnational Nagaland 52.5 region Arunachal Pradesh 52.7 Male 78.7 Mizoram 62.0 Birth order 1st born 83.0 Assam 66.7 Uttar Pradesh 67.0 2nd-3rd born 78.7 Tripura 71.1 4th-5th born 69.2 Rajasthan 71.9
6th born or higher 59.3 Gujarat 72.8 Madhya Pradesh 73.7 Mother's age 15-19 years 79.3 at birth Daman and Diu 74.0 20-34 years 79.0 Meghalaya 74.2 75.0 35-49 years 70.1 Dadra and Nagar Haveli Maharashtra 75.3 Mother's No education 68.6 76.9 education Haryana Primary school 77.9 Manipur 78.0 Karnataka 78.6 Secondary school 82.9 Uttarakhand 80.2 More than secondary school 87.0 Bihar 80.6
Mother's Scheduled caste 79.6 Jharkhand 82.7 caste/tribe Delhi 85.2 Scheduled tribe 73.8 Andaman and Nicobar Islands 85.2 Other backward class 78.6 Tamil Nadu 85.2 Himachal Pradesh 85.3 Other 80.4 Telangana 87.9 Sex of Female 78.6 Jammu and Kashmir 88.8 household head Male 78.8 Andhra Pradesh 89.0 Odisha 89.4 Household Quintile 1 (poorest) 70.4 economic Kerala 90.4 status Quintile 2 77.1 Chhattisgarh 91.7 93.0 Quintile 3 80.9 Sikkim West Bengal 93.0 Quintile 4 83.8 Goa 94.2 Quintile 5 (richest) 85.7 Punjab 94.5 Lakshadweep 95.1 Place of Rural 78.0 residence Chandigarh* 95.9 Urban 80.6 Puducherry 96.0
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average. *Estimate was based on 25–49 cases. observed across wealth quintiles, with the gap in (78%). Coverage across subnational regions varied coverage between the richest and poorest quintiles markedly. Nagaland and Arunachal Pradesh had the amounting to 16 percentage points (86% and 70%, lowest coverage at 53%, whereas 9 out of the 36 respectively). regions reported coverage of 90% or higher.
Inequality by place of residence in India was minimal, with a 3 percentage point difference between coverage in urban (81%) and rural areas
27 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Wealth- and education-related inequalities curve for DTP3 immunization coverage in India is The wealth-related inequality in DTP3 immunization available in Chapter 13. coverage in India amounted to a positive ECI value of 0.130. While small, this value indicates that Multiple regression analysis coverage is more concentrated among wealthier households. The education-related ECI in India Figure 7.2 shows odds ratios for the adjusted had a value of 0.150, suggesting that coverage was associations between DTP3 immunization coverage more concentrated in children with more-educated and selected socioeconomic, demographic and mothers. Both ECIs were largely statistically geographic factors. There was large variation in significant. The wealth-related concentration the odds of childhood immunization coverage
FIGURE 7.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in India, calculated as odds ratio (NFHS 2015)
Characteristic Subgroup Characteristic Subgroup
Child's sex Male (reference) Subnational Nagaland (reference) region Arunachal Pradesh 1.02 Female 1.01 Mizoram 1.20 Birth order 6th born or higher (reference) Tripura 1.67 Uttar Pradesh 1.83 1st born 1.81 Assam 1.88 1.51 2nd-3rd born Gujarat 1.93
4th-5th born 1.28 Daman and Diu 1.96 Haryana 2.03 Mother's age 15-19 years (reference) at birth Maharashtra 2.15 20-34 years 1.24 Rajasthan 2.16 Karnataka 2.43 35-49 years 1.27 Madhya Pradesh 2.52 Mother's No education (reference) Manipur 2.68 education Primary school 1.40 Uttarakhand 2.73 Dadra and Nagar Haveli 2.77 Secondary school 1.61 Meghalaya 2.79 More than secondary school 1.89 Himachal Pradesh 2.86 3.29 Mother's Scheduled tribe (reference) Andaman and Nicobar Islands caste/tribe Tamil Nadu 3.35 Scheduled caste 1.24 Delhi 4.09 Other backward class 1.12 Kerala 4.74 Bihar 4.85 Other 1.04 Telangana 4.90 Sex of Male (reference) Jharkhand 5.08 household head Female 0.97 Jammu and Kashmir 5.24 Andhra Pradesh 5.41 Household Quintile 1 (poorest) (reference) economic Odisha 7.94 status Quintile 2 1.37 Goa 8.26 Sikkim 8.46 Quintile 3 1.64 Punjab 9.39 Quintile 4 2.03 Chhattisgarh 10.04 Quintile 5 (richest) 2.26 Lakshadweep 11.65 West Bengal 12.54 Place of Rural (reference) residence Puducherry 13.01 Urban 0.75 Chandigarh 14.56
1 2 1 2 4 8 16 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
28 7. India
across regions. Compared to the reference region of References Nagaland, which had the lowest DTP3 immunization 1. United Nations Population Division. World population coverage in the country, several states/union prospects: 2017 revision [Internet]. New York: United territories demonstrated more than 10 times Nations; 2017 (https://esa.un.org/unpd/wpp, accessed higher odds of coverage, including Chandigarh, 21 February 2017). 2. Levels and trends in child mortality report 2017. Chhattisgarh, Lakshadweep, Puducherry and West Estimates developed by the UN Inter-agency Group for Bengal. Child Mortality Estimation. United Nations Children’s Fund, World Health Organization, World Bank and Immunization coverage was positively associated United Nations [Internet]. New York: United Nations Children’s Fund; 2017 (http://www.childmortality.org/ with economic status and mother’s education, files_v21/download/IGME%20report%202017%20 adjusting for other factors. A one-year-old child in child%20mortality%20final.pdf, accessed 21 February the richest quintile had a 2.3 times higher chance 2018). of being covered than a child in the poorest quintile. 3. Patel V, Parikh R, Nandraj S, Balasubramaniam P, Narayan K, Paul VK et al. Assuring health coverage for Compared to the no education subgroup, the odds all in India. Lancet. 2015;386(10011):2422–35. of DTP3 immunization coverage were 1.4 times 4. Ministry of Health & Family Welfare. National Vaccine higher in the primary school subgroup, 1.6 times Policy 2011. New Delhi: Government of India; 2011. higher in the secondary school subgroup, and 5. Mission Indradhanush [Internet]. (http://www. almost double in the more than secondary school missionindradhanush.in/index.html, accessed 13 April 2018). subgroup. 6. Bashar MA. Success story of mission Indradhanush: a road to achieve universal immunization. Natl J There was a weak, although statistically significant, Community Med. 2016;7:455–6. association between mother’s age at birth and 7. Shrivastwa N, Gillespie BW, Kolenic GE, Lepkowski JM, childhood immunization: the 20–34 years subgroup Boulton ML. Predictors of vaccination in India for children and the 35–49 years subgroup were 1.2 and 1.3 aged 12–36 months. Vaccine. 2015 Nov;33:D99–105. 8. Debnath A, Bhattacharjee N. Wealth-based inequality in times more likely, respectively, of being covered child immunization in India: a decomposition approach. than the 15–19 years subgroup. Birth order had J Biosoc Sci. 2017 August 14;1–14. a negative association with DTP3 immunization 9. Francis MR, Nohynek H, Larson H, Balraj V, Mohan coverage: relative to children 6th born or higher, VR, Kang G et al. Factors associated with routine the odds of coverage were significantly higher in all childhood vaccine uptake and reasons for non- vaccination in India: 1998–2008. Vaccine [Internet]. lower order birth subgroups. The chance of being 2017 August (http://linkinghub.elsevier.com/retrieve/ covered by DTP3 immunization among 1st born pii/S0264410X17310939, accessed 13 April 2018). children was nearly twice as large. Compared to those of mothers in the scheduled tribe subgroup, those in the scheduled caste subgroup had 1.2 times Interactive visuals and tables higher chance of being covered. Place of residence also demonstrated a significant association, as Interactive visuals and tables allow for further exploration children living in urban areas had a lower chance of inequality in childhood immunization in India. (0.75) of being covered than those in rural areas. To access interactive visuals:
SCAN HERE: or VISIT: The sex of the child and the sex of the household head showed non-significant associations with http://apps.who.int/gho/ data/view.wrapper.HE- DTP3 immunization coverage after adjusting for VIZ21?lang=en&menu=hide other factors.
To access interactive tables, see Appendix 1.
29 8. Indonesia
Country context posing challenges in reaching children in migrant and urban poor families with limited access to health Indonesia has a population of 258 million, with care. Other challenges in delivering immunizations 9.5% under 5 years of age (1). The under-5 mortality stem from supply-side issues such as geographic rate in Indonesia has decreased from 84.3 deaths disparity, topography, cold-chain maintenance and per 1000 live births in 1990, to 26.4 deaths per reluctance to implement multi-dose vial policy; 1000 live births in 2016 (2), with inequalities that demand-side issues include an increasing trend of favour population subgroups that are richer, more vaccine hesitancy, such as negative perceptions of educated or urban (3). The country is currently immunization side-effects and suspicion of haram comprised of 34 provinces, which span five main ingredients (5). islands and four archipelagos. (Note that the data in this report are from the 2012 DHS, conducted Descriptive overview before the creation of the province of North Kalimantan, and thus covering 33 provinces.) Disaggregation by background characteristics The immunization programme in Indonesia is guided DTP3 immunization coverage among one-year-olds by a comprehensive multi-year plan for immunization in Indonesia was 72%. Figure 8.1 presents DTP3 (2015–2019), and the country has a fully functional immunization coverage in Indonesia by population Independent National Technical Advisory Group subgroups. on immunization. Vaccines are primarily delivered by government health centres at the subdistrict Male and female children had similar levels of level and private providers. The Indonesian Ministry coverage. While 1st born and 2nd–3rd born children of Health has launched a number of national reported higher levels of coverage (76% and 74%, activities to promote immunization, including: respectively), immunization coverage was lower Backlog Fighting; National Immunization Week; among children born 4th–5th (61%) and much catch-up campaigns; Sustained Outreach Strategy lower among those born 6th or higher (37%). DTP3 for drop-out and follow-up in targeted districts; immunization coverage showed a reverse U-shape Strengthening Technical Assistance for Routine pattern by mother’s age at birth: coverage was Immunization (START) projects; and outbreak higher in the 20–34 years subgroup (74%) than the response (4). Social assistance programmes such as 15–19 years subgroup (66%) and the 35–49 years Programme Keluarga Harapan encourage the use subgroup (69%). of immunization services through conditional cash transfers to poor families (5). A gradient in immunization coverage was observed across mother’s education level and economic Immunization coverage in Indonesia varies by status, with increasing coverage in more-educated household economic status, mother’s education and richer subgroups. Both mother’s education and level, place of residence and province (4); other economic status demonstrated a marginal exclusion factors correlated with immunization status include pattern, whereby coverage was considerably lower birth order, family size, parental occupation, health in the most disadvantaged subgroup. There was insurance and distance to a health facility (6,7). no difference in coverage based on the sex of the The country faces a rapidly urbanizing population, household head.
30 8. Indonesia
FIGURE 8.1. DTP3 immunization coverage among one-year-olds in Indonesia, disaggregated by background characteristics (DHS 2012)
Characteristic Subgroup Characteristic Subgroup 35.3 72.2 Child's sex Female 71.2 Subnational Papua region Maluku 47.9 73.2 Male Banten 49.1
West Sulawesi 49.8 Birth order 1st born 76.4 Central Kalimatan 52.5 73.5 2nd-3rd born Aceh 57.6
West Papua 58.1 4th-5th born 60.5 South Sulawesi 60.3 36.7 6th born or higher North Sumatera 61.7
South Kalimatan 62.1 Mother's age 15-19 years 65.7 at birth North Maluku 62.2 73.6 20-34 years West Kalimatan 62.8
West Sumatera 62.9 35-49 years 69.1 Riau 67.9 28.1 Mother's No education Jambi 69.3 education South Sumatera 70.1 Primary school 62.0 West Nusa Tenggara 70.7 75.8 Secondary school Central Sulawesi 71.5
Bengkulu 71.9 More than secondary school 85.9 Gorontalo 72.7 72.9 Sex of Female Bangka Belitung 72.8 household head West Java 73.8 Male 72.1 Lampung 74.1 52.5 Household Quintile 1 (poorest) Riau Islands 74.2 economic status East Nusa Tenggara 76.4 Quintile 2 69.1 Southeast Sulawesi 76.4 75.0 Quintile 3 DKI Jakarta 77.5
East Kalimatan 80.4 Quintile 4 81.0 Central Java 82.7 85.1 Quintile 5 (richest) North Sulawesi 84.2
East Java 84.3 Place of Rural 67.1 residence Bali 89.2 Urban 77.6 DI Yogyakarta 96.4
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average.
Place of residence inequality in childhood Wealth- and education-related inequalities immunization showed a gap of 11 percentage points The wealth-related ECI for DTP3 immunization between urban and rural areas (78% in urban coverage showed a positive value of 0.261, indicating areas versus 67% in rural areas). Coverage across that coverage was more concentrated among richer provinces varied significantly in Indonesia, with households. The magnitude of inequality across some provinces such as DI Yogyakarta reaching different levels of education showed a positive almost universal coverage (96%), while others, ECI value of 0.226, indicating that DTP3 coverage such as Banten, Maluku, Papua and West Sulawesi, was higher among more-educated groups. Both had coverage levels below 50%. ECIs were significantly significant. The wealth-
31 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
related concentration curve for DTP3 immunization immunization coverage across subnational regions, coverage in Indonesia is available in Chapter 13. demonstrating a statistically significant association. For example, those living in the provinces of DI Multiple regression analysis Yogyakarta, Bali, East Nusa Tenggara and North Sulawesi were more than 8 times as likely to be Figure 8.2 shows associations between covered as those living in Banten. DTP3 immunization coverage and selected socioeconomic, demographic and geographic Immunization coverage was strongly associated factors, controlling for other factors. There with birth order, mother’s education, household was a large variation in the odds of childhood economic status and subnational region. The
FIGURE 8.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Indonesia, calculated as odds ratio (DHS 2012)
Characteristic Subgroup Characteristic Subgroup Subnational Banten (reference) Child's sex Male (reference) region Maluku 1.48 0.85 Female Papua 1.69 Aceh 1.94 Birth order 6th born or higher (reference) Central Kalimatan 2.04 3.94 1st born West Sumatera 2.14 South Sulawesi 2.45 2nd-3rd born 3.29 North Sumatera 2.55 2.17 4th-5th born South Kalimatan 2.58 DKI Jakarta 2.59 Mother's age 15-19 years (reference) at birth Riau 2.63 1.48 20-34 years West Papua 2.72 West Sulawesi 2.78 35-49 years 1.90 West Java 3.06 Mother's No education (reference) Riau Islands 3.12 education South Sumatera 3.18 Primary school 2.23 Bangka Belitung 3.38 3.27 Secondary school West Kalimatan 3.39 Jambi 3.53 More than secondary school 4.37 North Maluku 3.56 Sex of Male (reference) Bengkulu 3.63 household head West Nusa Tenggara 3.85 Female 0.89 Lampung 4.10 Household Quintile 1 (poorest) (reference) Gorontalo 4.69 economic status Central Sulawesi 5.04 Quintile 2 1.60 East Java 5.72 1.96 Quintile 3 Central Java 5.74 East Kalimatan 6.26 Quintile 4 2.83 Southeast Sulawesi 6.64 Quintile 5 (richest) 3.36 North Sulawesi 8.56 East Nusa Tenggara 8.67 Place of Rural (reference) residence Bali 9.44 Urban 1.08 DI Yogyakarta 24.37
1 2 4 1 2 4 8 16 32 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
32 8. Indonesia
6. Fernandez RC, Awofeso N, Rammohan A. Determinants association between birth order and childhood of apparent rural-urban differentials in measles immunization was negative: compared to children vaccination uptake in Indonesia. Rural Remote Health. born 6th or higher, children with lower birth order 2011;11(3):1702. were more likely to be covered. For instance, 1st 7. Herliana P, Douiri A. Determinants of immunisation born and 2nd–3rd born children were more than 3 coverage of children aged 12–59 months in Indonesia: a cross-sectional study. BMJ Open. 2017 December times as likely to be covered. By mother’s education 22;7(12):e015790. level, compared to children in the no education subgroup, the odds of DTP3 immunization coverage were more than double in the primary education subgroup, more than triple in the secondary education subgroup, and more than quadruple in the subgroup with more than secondary education. A one-year-old child in the richest quintile had a 3.4 times higher chance of being covered than a child in the poorest quintile.
Mother’s age at birth demonstrated an association with childhood immunization whereby children of mothers aged 35–49 years had a probability of being covered that was 1.9 times higher than children of mothers aged 15–19 years.
Child’s sex, sex of household head and place of residence did not show associations with immunization coverage after adjusting for other factors.
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). 2. Levels and trends in child mortality report 2017. Estimates developed by the UN Inter-agency Group for Child Mortality Estimation. United Nations Children’s Interactive visuals and tables Fund, World Health Organization, World Bank and United Nations [Internet]. New York: United Nations Interactive visuals and tables allow for further exploration Children’s Fund; 2017 (http://www.childmortality.org/ of inequality in childhood immunization in Indonesia. files_v21/download/IGME%20report%202017%20 child%20mortality%20final.pdf, accessed 21 February To access interactive visuals: 2018). 3. Indonesia Ministry of National Development Planning, SCAN HERE: or VISIT: UNICEF. SDG baseline report on children in Indonesia. http://apps.who.int/gho/ Jakarta: BAPPENAS and UNICEF; 2017. data/view.wrapper.HE- 4. State of health inequality: Indonesia. Geneva: World VIZ21?lang=en&menu=hide Health Organization; 2017. 5. Asia Pacific Observatory on Health Systems and Policies. The Republic of Indonesia health system review. New To access interactive tables, see Appendix 1. Delhi: World Health Organization, Regional Office for South-East Asia; 2017.
33 9. Kenya
Country context order (7,8). Childhood immunization activities are bolstered by high community-level awareness and Kenya has a population of 47 million, with 14.8% acceptance of immunization, and the existence of under 5 years of age (1). The under-5 mortality Expanded Programme on Immunization focal points rate was 98.1 deaths per 1000 live births in 1990, in every county. The current Comprehensive Multi- falling to 49.2 deaths per 1000 live births in 2016 Year Plan acknowledges concerns about vaccine (2). Key child health policies and strategies, such quality, as well as shortages and lack of training in as the removal of user fees for maternal and the health workforce, and low government funding child health care, the Kenya Essential Package for immunization activities (6). for Health, and the Community Health Strategy, have contributed to gains in child survival, though Descriptive overview geographic and socioeconomic inequalities in child health intervention coverage persist (3,4). Disaggregation by background Historically, Kenya consisted of eight provinces, characteristics which were restructured as 47 counties in the Overall, DTP3 immunization coverage among one- 2010 Constitution. (Note: the 2014 DHS retained year-olds in Kenya was 90%. Figure 9.1 presents the former provincial designations in its sampling disaggregated data for DTP3 immunization design, and thus in this report, the counties are coverage among one-year-olds. grouped according to these eight regions (5).) Male and female children had about the same level The Kenya Expanded Programme on Immunization of DTP3 immunization coverage. Inequality by was established in 1980 with the goal of achieving birth order was observed only for those born 6th universal coverage of childhood immunizations or higher: while coverage among 1st, 2nd–3rd and (6). The Expanded Programme on Immunization is 4th–5th born children ranged between 91% and managed by the Unit of Vaccine and Immunization 93%, it was 79% among child born 6th or higher. within the Ministry of Health, and guided by successive 5-year Comprehensive Multi-Year DTP3 immunization coverage was lower among Plans for immunization (6). National and county children of mothers aged 35–49 years (83%) than level governments have distinct roles and the younger subgroups, where coverage was 93% responsibilities in delivering immunization services: and 91% for mothers aged 15–19 years and 20–34 national governments establish standards and years, respectively. The no education subgroup guidelines, and provide technical assistance, while reported coverage of 78%, which increased sharply county governments oversee the planning and in the primary education subgroup (91%); there management of immunization service delivery at were almost no additional gains in coverage among the facility level, and are responsible for ensuring the higher education subgroup. DTP3 coverage adherence to national guidelines (6). varied with mother´s ethnic group: while the Maasai (77%) and the Somali (80%) demonstrated Inequalities in immunization coverage in Kenya have markedly lower immunization coverage than the been reported between provinces, and on the basis national average, the Kisii (96%) were closest to of maternal education, household wealth and birth universal coverage.
34 9. Kenya
FIGURE 9.1. DTP3 immunization coverage among one-year-olds in Kenya, disaggregated by background characteristics (DHS 2014)
Characteristic Subgroup Characteristic Subgroup Child's sex Female 90.6 90.1
Male 89.7 Subnational North Eastern 77.8 Birth order 1st born 93.5 region 2nd-3rd born 91.5
4th-5th born 91.0 6th born or higher 78.8 Nairobi 88.0 Mother's age 15-19 years 92.9 at birth 20-34 years 90.5 35-49 years 83.3
Mother's No education 77.7 education Primary school 90.7 Rift Valley 88.4 Secondary school or more 93.4
Mother's Kalenjin 91.2 ethnicity Kamba 91.9 Kikuya 92.9 Nyanza 89.8
Kisii 95.8
Luhya 88.2 Luo 87.6
Maasai 77.4 Western 90.3 Meru 94.2 Mijikenda / Swahili 91.3
Somali 80.3 87.9 Turkana Coast 92.2 Other 91.0
Sex of Female 89.6 household head Male 90.3 Household Quintile 1 (poorest) 83.7 93.7 economic Eastern status Quintile 2 90.4 Quintile 3 92.0 Quintile 4 93.9
Quintile 5 (richest) 92.8 Central 95.5 Place of Rural 89.4 residence Urban 91.4
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average.
There was no difference in coverage based on Immunization coverage was nearly the same the sex of the household head. Across wealth among children in urban areas (91%) and children quintiles, the maximum gap in DTP3 immunization in rural areas (89%). Across the eight subnational coverage remained less than 10 percentage points. regions, the lowest coverage was reported by the The poorest quintile demonstrated distinctly lower North Eastern region (78%), which was at least 10 coverage (84%) than the other four quintiles (90% percentage points lower than any other region. The to 94%). Central region presented the highest immunization coverage at 96%.
35 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Wealth- and education-related inequalities immunization coverage in Kenya is available in The wealth-related ECI for Kenya showed a positive Chapter 13. value of 0.079, indicating minor inequality in DTP3 immunization coverage Education-related Multiple regression analysis inequality, as measured by the ECI, showed a positive value of 0.091. This suggests that DTP3 coverage Figure 9.2 shows the adjusted associations between was slightly more concentrated among children of DTP3 immunization coverage and socioeconomic, more-educated mothers. Although relatively small, demographic and geographic factors. After inequalities in Kenya were statistically significant. controlling for other factors, there was statistically The wealth-related concentration curve for DTP3 significant inequality in DTP3 immunization
FIGURE 9.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Kenya, calculated as odds ratio (DHS 2014)
Characteristic Subgroup Characteristic Subgroup Child's sex Male (reference) Female 1.11 Subnational Nairobi (reference) Birth order 6th born or higher (reference) region 1st born 2.30 2nd-3rd born 2.11 4th-5th born 2.35 North Eastern 1.31 Mother's age 15-19 years (reference) at birth 20-34 years 0.84 35-49 years 0.70 Mother's No education (reference) education Primary school 2.43 Rift Valley 1.79 Secondary school or more 2.54 Mother's Luhya (reference) ethnicity Kalenjin 2.42 Kamba 1.32 Nyanza 2.31 Kikuya 1.52 Kisii 3.18 Luo 1.06
Maasai 1.49 Central 3.12 Meru 1.27 Mijikenda / Swahili 2.15 Somali 3.21 4.45 Turkana Western 3.18 Other 2.32 Sex of Male (reference) household head Female 0.91 Household Quintile 1 (poorest) (reference) Coast 3.18 economic status Quintile 2 1.40 Quintile 3 1.79 Quintile 4 2.62 Quintile 5 (richest) 2.61 Eastern 3.80 Place of Rural (reference) residence Urban 0.90
1 2 4 1 2 4 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
36 9. Kenya
coverage across regions in Kenya. For instance, 4. Brault MA, Ngure K, Haley CA, Kabaka S, Sergon K, the Eastern region had 3.8 times higher odds of Desta T et al.; Bhutta ZA, editor. The introduction of new policies and strategies to reduce inequities and improve immunization coverage than the reference region child health in Kenya: a country case study on progress of Nairobi. According to mother’s ethnicity, the in child survival, 2000–2013. PLoS ONE. 2017 August Turkana, Kisii and Kalenjin reported significantly 1;12(8):e0181777. higher odds of coverage than the reference ethnic 5. Kenya National Bureau of Statistics, Ministry of Health Kenya, National AIDS Control Council Kenya, subgroup of Luhya by 4.5, 3.2 and 2.4 times, Kenya Medical Research Institute, National Council respectively. for Population and Development Kenya. Kenya Demographic and Health Survey 2014 [Internet]. Rockville MD: USAID; 2015 (https://dhsprogram.com/ After controlling for background characteristics, publications/publication-fr308-dhs-final-reports.cfm, immunization coverage was positively associated accessed 22 May 2018). with household economic status and mother’s 6. Unit of Vaccines and Immunization Services. education. Relative to children in the poorest Comprehensive Multi-Year Plan for Immunization: July quintile, the odds of being covered by DTP3 was 2.6 2015–June 2019. Nairobi: Ministry of Health for Kenya; 2015. times higher for children in each of the richest two 7. Van Malderen C, Ogali I, Khasakhala A, Muchiri SN, quintiles (quintiles 4 and 5). Comparably, having Sparks C, Van Oyen H et al. Decomposing Kenyan socio- at least secondary education increased the odds of economic inequalities in skilled birth attendance and coverage by 2.5 times, compared to no education. measles immunization. Int J Equity Health. 2013;12(1):3. 8. Onsomu EO, Abuya BA, Okech IN, Moore D, Collins- McNeil J. Maternal education and immunization status Compared to children born 6th or higher, the among children in Kenya. Matern Child Health J. 2015 odds of reporting DTP3 coverage were more than August;19(8):1724–33. double for all other lower birth order subgroups, and associations were significant.
The remaining factors – child’s sex, mother’s age at birth, sex of the household head, and place of residence – did not demonstrate a significant association with DTP3 immunization.
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). 2. Levels and trends in child mortality report 2017. Interactive visuals and tables Estimates developed by the UN Inter-agency Group for Child Mortality Estimation. United Nations Children’s Interactive visuals and tables allow for further exploration Fund, World Health Organization, World Bank and of inequality in childhood immunization in Kenya. United Nations [Internet]. New York: United Nations Children’s Fund; 2017 (http://www.childmortality.org/ To access interactive visuals: files_v21/download/IGME%20report%202017%20 SCAN HERE: or VISIT: child%20mortality%20final.pdf, accessed 21 February 2018). http://apps.who.int/gho/ 3. Keats EC, Ngugi A, Macharia W, Akseer N, Khaemba EN, data/view.wrapper.HE- Bhatti Z et al. Progress and priorities for reproductive, VIZ21?lang=en&menu=hide maternal, newborn, and child health in Kenya: a Countdown to 2015 country case study. Lancet Glob Health. 2017;5(8):e782–e795. To access interactive tables, see Appendix 1.
37 10. Nigeria
Country context communication and promotion for immunization activities; lack of security in some parts of the Nigeria has a population of 181 million, with 17.2% country; and lack of high-quality data for monitoring under 5 years of age (1). The country is divided and evaluation (5). into 36 states and one Federal Capital Territory, which are grouped into six geopolitical zones. As Descriptive overview of 2016, the under-5 mortality rate in Nigeria was estimated at 104.3 deaths per 1000 live births, Disaggregation by background representing a reduction since 1990, when the characteristics country reported 212.9 under-5 deaths per 1000 The national coverage of DTP3 immunization live births (2). Improvements in child mortality have coverage among one-year-olds in Nigeria was 39%. occurred alongside a proliferation of child health Figure 10.1 presents coverage of DTP3 immunization policies and strategies, especially between 2005 disaggregated by background characteristics. and 2010 (3). Coverage differed by just 3 percentage points Nigeria introduced the Expanded Programme between males (40%) and females (37%). on Immunization in 1979, which was successful There was a negative relationship between DTP3 in increasing immunization coverage over immunization coverage and birth order: with the subsequent decade (4). Following declining increasing birth order, the immunization coverage immunization coverage in the early 1990s, the decreased. The level of coverage of one-year-olds National Programme on Immunization was that were 1st born children (50%) was about twice established in 1996, which merged with the as high as the level of coverage among children National Primary Health Care Development Agency born 6th or higher (26%). Considering DTP3 in 2007 (4). The National Primary Healthcare immunization coverage by mother’s age at birth, Development Agency is responsible for providing coverage was highest among the 20–34 years free vaccines and coordinating immunization subgroup (42%), and lowest in the 15–19 years vaccine procurement initiatives. In recent years, subgroup (22%). Nigeria has made some progress in increasing immunization coverage due in part to demand- Results demonstrated large education-related side factors such as increased awareness and and wealth-related inequalities in childhood community mobilization activities (5). immunization. The pattern of inequality was similar across education and wealth groups, showing step- National immunization coverage in Nigeria wise increases in coverage, moving from the most remains low, and varies across geographic disadvantaged to the most advantaged. Across the areas, and according to socioeconomic and four education subgroups, immunization coverage demographic factors such as ethnicity, parental increased by about 30 percentage points between occupation, wealth status and religious affiliation each subgroup, moving from the no education (6–9). At a population level, childhood immunization subgroup to the secondary education subgroup, programmes are adversely affected by: health and by about 20 percentage points between personnel shortages; insufficient funds and secondary education and more than secondary resources for immunization activities; inadequate education subgroups. Across wealth quintiles,
38 10. Nigeria
FIGURE 10.1. DTP3 immunization coverage among one-year-olds in Nigeria, disaggregated by background characteristics (DHS 2013)
Characteristic Subgroup Characteristic Subgroup 38.5 Child's sex Female 37.2
Male 39.7 Subnational North West 14.0 region Birth order 1st born 49.6
2nd-3rd born 42.2
4th-5th born 36.8
6th born or higher 26.3 North East 21.4
Mother's age 15-19 years 22.3 at birth 20-34 years 42.0
35-49 years 38.9
North Central 44.1 Mother's No education 12.2 education Primary school 40.4
Secondary school 70.3
More than secondary school 87.4 South West 65.8 Sex of Female 60.5 household head Male 36.0
Household Quintile 1 (poorest) 7.4 economic status Quintile 2 18.9 South South 70.0
Quintile 3 40.1
Quintile 4 60.1
Quintile 5 (richest) 79.6
South East 80.9 Place of Rural 25.2 residence Urban 62.4
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average.
coverage increased by about 20 percentage points regions. For example, while the best-performing between each quintile, from quintiles 2 to 5. South East region reported a coverage of 81%, the North West region showed coverage of only 14% Immunization coverage was 2.5 times higher – a gap of 67 percentage points. among children in urban areas (62%) than children in rural areas (25%). Across regions, there was Wealth- and education-related inequalities significant variation in coverage, with northern The wealth-related ECI for Nigeria showed a regions performing markedly worse than southern positive value of 0.612, indicating that coverage was
39 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
highly concentrated among the richest population. Multiple regression analysis Education-related inequality was also very large in Nigeria, as demonstrated by a positive ECI value Figure 10.2 shows the adjusted associations of 0.613. Similar to wealth-related inequality, between DTP3 immunization coverage among this result shows that DTP3 coverage was highly one-year-olds and socioeconomic, demographic concentrated among children with more-educated and geographic factors. Immunization coverage mothers. Both ECIs were statistically significant. was strongly associated with economic status and The wealth-related concentration curve for DTP3 mother’s education, even after controlling for other immunization coverage in Nigeria is available in characteristics. The odds of receiving the third dose Chapter 13.
FIGURE 10.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Nigeria, calculated as odds ratio (DHS 2013)
Characteristic Subgroup Characteristic Subgroup
Child's sex Male (reference)
Female 0.94 Subnational North West (reference) region Birth order 6th born or higher (reference)
1st born 1.38
2nd-3rd born 1.02
4th-5th born 0.92 North East 1.45
Mother's age 15-19 years (reference) at birth 20-34 years 1.89
35-49 years 2.38 North Central 2.49 Mother's No education (reference) education Primary school 1.84
Secondary school 3.73
More than secondary school 6.64 South West 2.13
Sex of Male (reference) household head Female 1.43
Household Quintile 1 (poorest) (reference) economic status Quintile 2 1.93 South South 3.28
Quintile 3 3.03
Quintile 4 4.82
Quintile 5 (richest) 7.27 South East 6.71 Place of Rural (reference) residence Urban 1.26
1 2 4 8 1 2 4 8 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
40 10. Nigeria
of the DTP vaccine among children in the richest 4. National Primary Health Care Development Agency. quintile were more than 7 times higher than for Comprehensive EPI Multi-Year Plan 2011–2015. Abuja: Federal Ministry of Health Nigeria; 2011. children in the poorest quintile. The subgroup with 5. Global Vaccine Action Plan: priority country reports the most education was more than 6 times more on progress towards GVAP-RVAP goals. Annex to the likely to be covered than the subgroup with the GVAP Secretariat Annual Report 2016. Geneva: World least education. Health Organization; 2016. 6. Antai D. Inequitable childhood immunization uptake After controlling for other factors, large and in Nigeria: a multilevel analysis of individual and contextual determinants. BMC Infect Dis [Internet]. statistically significant inequalities in DTP3 2009 December (http://bmcinfectdis.biomedcentral. immunization coverage across regions were com/articles/10.1186/1471-2334-9-181, accessed 6 evident. Compared to the North West, all other April 2018). regions reported odds of coverage that were 7. Ataguba JE, Ojo KO, Ichoku HE. Explaining socio-economic inequalities in immunization between 1.5 and 6.7 times larger. coverage in Nigeria. Health Policy Plan. 2016 November;31(9):1212–24. There was a significant positive relationship 8. Oleribe O, Kumar V, Awosika-Olumo A, Taylor SD. between immunization coverage and mother’s age Individual and socioeconomic factors associated with childhood immunization coverage in Nigeria. Pan Afr at birth: compared to the 15–19 years subgroup, Med J [Internet]. 2017 (http://www.panafrican-med- the odds of coverage were 2.4 times higher in the journal.com/content/article/26/220/full/, accessed 6 35–49 years subgroup and 1.9 times higher among April 2018). the 20–34 years subgroup. Children in female- 9. State of inequality: childhood immunization. Geneva: World Health Organization; 2016. headed households had 1.4 times higher probability of immunization than children in male-headed households.
The sex of the child, birth order and place of residence showed no significant associations with DTP3 immunization coverage.
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). 2. Levels and trends in child mortality report 2017. Estimates developed by the UN Inter-agency Group for Interactive visuals and tables Child Mortality Estimation. United Nations Children’s Fund, World Health Organization, World Bank and Interactive visuals and tables allow for further exploration United Nations [Internet]. New York: United Nations of inequality in childhood immunization in Nigeria. Children’s Fund; 2017 (http://www.childmortality.org/ files_v21/download/IGME%20report%202017%20 To access interactive visuals: child%20mortality%20final.pdf, accessed 21 February 2018). SCAN HERE: or VISIT: 3. Kana MA, Doctor HV, Peleteiro B, Lunet N, Barros H. http://apps.who.int/gho/ Maternal and child health interventions in Nigeria: a data/view.wrapper.HE- systematic review of published studies from 1990 to VIZ21?lang=en&menu=hide 2014. BMC Public Health [Internet]. 2015 December (http://bmcpublichealth.biomedcentral.com/ articles/10.1186/s12889-015-1688-3, accessed 6 April 2018). To access interactive tables, see Appendix 1.
41 11. Pakistan
Country context gaps have persisted or worsened in recent years (11). Barriers to immunization are related to disease Pakistan has a population of 189 million, with 13.0% awareness and risk perception, knowledge and under 5 years of age (1). The country is divided perceptions about vaccines, transportation and into eight subnational regions. Six of these regions wait time issues, vaccine and vaccinator availability, were included in the report: four provinces; ICT and immunization recordkeeping (6). Strategies Islamabad; and the administrative territory of Gilgit to increase childhood immunization coverage Baltistan. The 2012 DHS excluded two regions – encompass efforts to build health worker capacity, Azad Jammu and Kashmir (Pakistan administered improve data management and health information Kashmir) and Federally Administered Tribal Areas systems, create demand for routine immunizations, (2) – which were thus not included in this report. and upgrade cold-chain infrastructure (6). Over the past decades, the under-5 mortality rate in Pakistan has decreased from 138.8 deaths per 1000 Descriptive overview live births in 1990 to 78.8 deaths per 1000 live births in 2016 (3). Despite progress, certain aspects Disaggregation by background of child health such as nutrition have worsened, characteristics and the country faces security, social, political, Among one-year-olds in Pakistan, national economic and environmental challenges that have DTP3 immunization coverage was 65%. Figure implications for health services (4). 11.1 presents coverage of DTP3 immunization by socioeconomic, demographic and geographic Pakistan initiated its Expanded Programme on characteristics. Immunization in 1978, which after funding disruptions in the early 1990s, was revived in 1996–1997 (5). The There was a 4 percentage point difference in decentralization of the Pakistani health system in coverage between males (67%) and females 2011 shifted certain roles and responsibilities for (63%), which was not statistically significant. DTP3 childhood immunization services from the federal to coverage decreased with increasing birth order: 1st the provincial level: while federal authorities remain born children reported coverage of 73%, whereas responsible for providing technical coordination only half of children from the highest birth order and support as well as delivering on national subgroup (6th or higher) were covered (50%). commitments at the global and regional levels, The level of coverage across mother’s age at birth provincial governments oversee the implementation showed a reverse U-shape pattern, where coverage and operation of immunization services (6,7). peaked in the 20–34 years subgroup (68%) and National-level strategic planning for immunization was a little more than 10 percentage points lower is detailed in the 2014–2018 Comprehensive Multi- in both the younger and older subgroups (55% and Year Plan, and subnational Comprehensive Multi- 57%, respectively). Year Plans guide planning, decision-making and budget allocations at local levels (7). The education-related and wealth-related gradient in DTP3 coverage presented patterns of marginal Immunization coverage has been shown to vary by exclusion, characterized by the most disadvantaged household economic status, maternal education subgroup (i.e. the least-educated or poorest) level and place of residence (8–10); such coverage presenting markedly lower coverage than the other more advantaged subgroups.
42 11. Pakistan
FIGURE 11.1. DTP3 immunization coverage among one-year-olds in Pakistan, disaggregated by background characteristics (DHS 2012)
Characteristic Subgroup Characteristic Subgroup 65.3 Child's sex Female 63.2
Male 67.4 Subnational Balochistan 27.9 region Birth order 1st born 73.0
2nd-3rd born 66.3
4th-5th born 66.1
6th born or higher 50.3 Sindh 38.6
Mother's age 15-19 years 55.3 at birth 20-34 years 67.6
35-49 years 57.1
Gilgit Baltistan 55.3 Mother's No education 51.1 education Primary school 74.4
Secondary school 85.6
More than secondary school 88.1 Khyber Pakhtunkhwa 70.2 Sex of Female 71.0 household head Male 64.8
Household Quintile 1 (poorest) 29.9 economic status Quintile 2 67.5 Punjab 76.3
Quintile 3 69.2
Quintile 4 78.9
Quintile 5 (richest) 88.0
ICT Islamabad 91.9 Place of Rural 59.1 residence Urban 79.1
0 50 100 0 50 100 Coverage (%) Coverage (%)
Notes: The red vertical line shows the national average. Data were not available for two subnational regions: Azad Jammu and Kashmir (Pakistan administered Kashmir) and Federally Administered Tribal Areas.
Children in urban areas had 20 percentage point Wealth- and education-related inequalities higher coverage than children in rural areas (79% The wealth-related inequality of DTP3 coverage versus 59%). Regions also showed substantial in Pakistan, measured by the ECI, had a positive variation in coverage. While coverage was only value of 0.435, indicating that coverage was 28% in Balochistan and 39% in Sindh, it reached more concentrated among wealthier households. 92% in the ICT Islamabad region. The ECI applied to summarize the magnitude of inequality across mother’s education revealed a
43 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
positive value of 0.334. This suggests a higher level Multiple regression analysis of coverage among children with more-educated mothers. Both wealth- and education-related Figure 11.2 shows the adjusted associations ECIs were statistically significant. The wealth- between DTP3 immunization coverage among related concentration curve for DTP3 immunization one-year-olds and selected socioeconomic, coverage in Pakistan is available in Chapter 13. demographic and geographic characteristics. The odds of coverage differed greatly across regions,
FIGURE 11.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Pakistan, calculated as odds ratio (DHS 2012)
Characteristic Subgroup Characteristic Subgroup
Child's sex Male (reference)
Female 0.88 Subnational Sindh (reference) region Birth order 6th born or higher (reference)
1st born 1.46
2nd-3rd born 1.05
4th-5th born 1.32 Balochistan 1.08
Mother's age 15-19 years (reference) at birth 20-34 years 1.65
35-49 years 1.84 Gilgit Baltistan 3.43 Mother's No education (reference) education Primary school 1.71
Secondary school 2.70
More than secondary school 2.38 Khyber Pakhtunkhwa 4.55
Sex of Male (reference) household head Female 1.07
Household Quintile 1 (poorest) (reference) economic status Quintile 2 3.18 Punjab 5.04
Quintile 3 2.67
Quintile 4 3.39
Quintile 5 (richest) 7.05 ICT Islamabad 7.77 Place of Rural (reference) residence Urban 1.31
1 2 4 8 1 2 4 8 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
Note: Data were not available for two subnational regions: Azad Jammu and Kashmir (Pakistan administered Kashmir) and Federally Administered Tribal Areas.
44 11. Pakistan
Health: progress and the way forward. Paediatr Int Child with all regions except Balochistan showing higher Health. 2015 October 2;35(4):287–97. odds of DTP3 immunization than the reference of 5. Owais A, Khowaja AR, Ali SA, Zaidi AKM. Pakistan’s Sindh. Notably, ICT Islamabad had 7.8 times higher expanded programme on immunization: an overview odds of coverage than Sindh. in the context of polio eradication and strategies for improving coverage. Vaccine. 2013 July;31(33):3313–19. 6. Global Vaccine Action Plan: priority country reports After controlling for other characteristics, there on progress towards GVAP-RVAP goals. Annex to the was a strong association between immunization GVAP Secretariat Annual Report 2016. Geneva: World coverage and household economic status and Health Organization; 2016. mother’s education. Relative to children in the 7. Ministry of National Health Services, Regulation and poorest quintile, the odds of being covered were Coordination. Expanded Program on Immunization: Comprehensive Multi-Year Plan 2015–2018. Islamabad: around 3 times higher for children in quintiles 2 Government of Pakistan; 2014. to 4; this ratio increased to 7 times for children in 8. Khan MT, Zaheer S, Shafique K. Maternal education, the richest quintile. The subgroup with mother’s empowerment, economic status and child polio education level of secondary education was vaccination uptake in Pakistan: a population based cross sectional study. BMJ Open. 2017;7(3):e013853. 2.7 times more likely to be covered than the no 9. Zaidi SMA, Khowaja S, Kumar Dharma V, Khan AJ, education subgroup, and the subgroup with more Chandir S. Coverage, timeliness, and determinants of than secondary education was 2.4 times more likely immunization completion in Pakistan: evidence from the to be covered. Demographic and Health Survey (2006–2007). Hum Vaccines Immunother. 2014 June 4;10(6):1712–20. 10. Murtaza F, Mustafa T, Awan R. Determinants of Compared to mothers 15–19 years of age, the nonimmunization of children under 5 years of age in odds of reporting coverage were 1.7 to 1.8 times Pakistan. J Fam Community Med. 2016;23(1):32. higher in the older age subgroups, although only 11. State of inequality: childhood immunization. Geneva: significant for the 20–34 years subgroup. Other World Health Organization; 2016. factors including child’s sex, birth order and sex of household head demonstrated non-significant associations with DTP3 immunization coverage, after controlling for background characteristics.
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). 2. Pakistan Demographic and Health Survey 2012–2013 [Internet]. Islamabad: National Institute of Population Interactive visuals and tables Studies Pakistan and ICF International; 2013 (http:// dhsprogram.com/pubs/pdf/FR290/FR290.pdf, Interactive visuals and tables allow for further exploration accessed 22 May 2018). of inequality in childhood immunization in Pakistan. 3. Levels and trends in child mortality report 2017. Estimates developed by the UN Inter-agency Group for To access interactive visuals: Child Mortality Estimation. United Nations Children’s SCAN HERE: or VISIT: Fund, World Health Organization, World Bank and United Nations [Internet]. New York: United Nations http://apps.who.int/gho/ Children’s Fund; 2017 (http://www.childmortality.org/ data/view.wrapper.HE- files_v21/download/IGME%20report%202017%20 VIZ21?lang=en&menu=hide child%20mortality%20final.pdf, accessed 21 February 2018). 4. Rizvi A, Bhatti Z, Das JK, Bhutta ZA. Pakistan and the To access interactive tables, see Appendix 1. Millennium Development Goals for Maternal and Child
45 12. Uganda
Country context stockouts at local levels, and inadequate monitoring and supervision (5,6). Uganda has a population of 40 million, with 18.7% under 5 years of age (1). The country saw a decrease Descriptive overview in the under-5 mortality rate from 175.0 to 53.0 deaths per 1000 live births between 1990 and 2016 Disaggregation by background (2). Child health outcomes are unevenly distributed characteristics across the country, as health service availability, As of 2016, the overall coverage of DTP3 quality and uptake are variable (3). As of 2016, immunization among one-year-olds was 79%. the country was divided into 112 administrative Figure 12.1 presents DTP3 immunization coverage districts; for the purpose of the 2016 Uganda DHS in Uganda disaggregated by background and the analyses in this report, these districts were characteristics. grouped into 15 subnational regions (4). Overall, coverage across all subgroups remained Immunization activities in Uganda are supported within about 10 percentage points of the national by a functional National Immunization Technical average of 79%. Male and female children had Advisory Group (5). The Expanded Programme on about the same level of coverage. Little variation in Immunization has had a history of mixed success, coverage was reported across birth order subgroups with progressive improvements realized between ranging from 1st to 5th born children (around 80%), 2000 and 2006, followed by a period of declining however, coverage was lower in higher order births performance (6). The Expanded Programme on (75%). Immunization Revitalization Plan, spanning 2012 to 2014, brought funding and attention to DTP3 immunization coverage by mother’s age at immunization service delivery in poor-performing birth was similar across all three subgroups, with districts through: strengthening outreach and slightly lower coverage in the 15–19 years subgroup promotion activities; increasing supervisory visits; (76%) than the 20–34 years and 35–49 years and implementing Reach Every District and Reach subgroups (80% in both). For education subgroups, Every Child strategies (5). In 2016, the Government DTP3 coverage was similar in the no education of Uganda enacted the Immunization Act, declaring (77%), primary education (79%) and secondary the compulsory immunization of children and other education (80%) subgroups, however, increased groups, and establishing an Immunization Fund to by a larger margin for mothers with more than support activities (7). secondary education (88%).
Immunization services in Uganda are primarily Male- and female-headed households had about provided by nurses, midwives and nursing the same level of coverage. The wealth-related assistants, and volunteer village health teams assist gradient showed very little inequality, with a with social mobilization activities (5). Staff shortages, maximum difference of just 4 percentage points especially in hard-to-reach and underserved areas, between quintile 4 (82%) and quintile 5 (78%). is one critical barrier to universal immunization coverage; other barriers include irregular and Rural and urban areas reported immunization insufficient outreach activities, vaccine shortages or coverage of 80% and 78%, respectively. Coverage
46 12. Uganda
FIGURE 12.1. DTP3 immunization coverage among one-year-olds in Uganda, disaggregated by background characteristics (DHS 2016)
Characteristic Subgroup Characteristic Subgroup 79.3 Child's sex Female 79.0 Subnational Busoga 71.2 region Male 79.6
Bugishu 73.1 Birth order 1st born 81.2
2nd-3rd born 80.4 Tooro 75.2
4th-5th born 79.5
North Central 75.5 6th born or higher 74.9
Mother's age 15-19 years 76.2 South Central 75.5 at birth 20-34 years 80.0 Bukedi 76.9 35-49 years 79.7
Mother's No education 76.5 Bunyoro 80.8 education Primary school 78.6 Kampala 80.9 Secondary school 79.5
More than secondary school 87.6 Lango 82.5
Sex of Female 77.7 household West Nile 83.1 head Male 79.8
Household Quintile 1 (poorest) 78.6 Ankole 83.4 economic status Quintile 2 79.2 Acholi 86.0 Quintile 3 79.0
86.8 Quintile 4 81.6 Karamoja
Quintile 5 (richest) 78.4 Kigezi 88.1
Place of Rural 79.8 residence Teso 90.7 Urban 77.6
0 50 100 0 50 100 Coverage (%) Coverage (%)
Note: The red vertical line shows the national average.
across regions, however, varied more markedly, was no inequality in coverage among children with with Busoga reporting the lowest coverage (71%) different levels of household wealth. By contrast, and Teso reporting the highest coverage (91%). the education-related ECI in Uganda had a positive value of 0.055. This value indicates that coverage Wealth- and education-related inequalities was more concentrated among children with more- The wealth-related ECI for DTP3 immunization educated mothers. While small, the education- coverage in Uganda had a value of 0.010, which was related ECI in Uganda was statistically significant. not statistically significant. This suggests that there The wealth-related concentration curve for DTP3
47 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
immunization coverage in Uganda is available in Teso were more than 4 times as likely to be covered Chapter 13. as those living in Busoga, the region with the lowest DTP3 immunization coverage. Multiple regression analysis There was a significant positive relationship Figure 12.2 shows that, controlling for other factors, between immunization coverage and mother’s age there was a large and statistically significant at birth: compared to the 15–19 years subgroup, the variation in the odds of childhood immunization odds of coverage were 2.4 times higher in the 35– across subnational regions in Uganda. For example, 49 years subgroup and 1.7 times higher among the one-year-olds living in the regions of Karamoja and 20–34 years subgroup. Birth order demonstrated a
FIGURE 12.2. Adjusted associations between DTP3 immunization coverage among one-year-olds and background characteristics in Uganda, calculated as odds ratio (DHS 2016))
Characteristic Subgroup Characteristic Subgroup
Child's sex Male (reference) Subnational Busoga (reference) region Female 0.98
Bugishu 1.15 Birth order 6th born or higher (reference)
2.25 1st born North Central 1.16
2nd-3rd born 1.71 South Central 1.18 4th-5th born 1.50
Mother's age 15-19 years (reference) Tooro 1.21 at birth 20-34 years 1.69 Bukedi 1.40 35-49 years 2.43
Mother's No education (reference) Kampala 1.74 education Primary school 1.23 Bunyoro 1.80 Secondary school 1.34
More than secondary school 2.17 Ankole 1.94
Sex of Male (reference) household Lango 2.29 head Female 0.82
2.38 Household Quintile 1 (poorest) (reference) West Nile economic status Quintile 2 1.33 Kigezi 2.70 Quintile 3 1.40
Acholi 3.16 Quintile 4 1.56
Quintile 5 (richest) 1.29 Karamoja 4.44
Place of Rural (reference) residence Teso 4.51 Urban 0.76
1 2 1 2 4 Odds ratio Odds ratio
Significant (p-value<0.05) Not significant
48 12. Uganda
significant association with immunization coverage. Most notably, 1st born children were 2.3 times more likely to report DTP3 immunization coverage than children born 6th or higher. Mother’s education was also associated with DTP3 immunization when comparing the most- and least-educated, with the most-educated subgroup having a 2.2 times higher chance of immunization than the least-educated subgroup.
Overall, immunization was not significantly associated with household economic status. Child sex, sex of household head and place of residence demonstrated no association with childhood immunization, when accounting for other background characteristics
References 1. United Nations Population Division. World population prospects: 2017 revision [Internet]. New York: United Nations; 2017 (https://esa.un.org/unpd/wpp, accessed 21 February 2017). 2. Levels and trends in child mortality report 2017. Estimates developed by the UN Inter-agency Group for Child Mortality Estimation. United Nations Children’s Fund, World Health Organization, World Bank and United Nations [Internet]. New York: United Nations Children’s Fund; 2017 (http://www.childmortality.org/ files_v21/download/IGME%20report%202017%20 child%20mortality%20final.pdf, accessed 21 February 2018). 3. Republic of Uganda, UNICEF. Country Programme Action Plan 2016–2020. Kampala: UNICEF Uganda Country Office; 2015. 4. Uganda Bureau of Statistics, ICF International. Uganda Demographic and Health Survey 2011. Kampala: ICF International Inc.; 2012. Interactive visuals and tables 5. Global Vaccine Action Plan: priority country reports on progress towards GVAP-RVAP goals. Annex to the Interactive visuals and tables allow for further exploration GVAP Secretariat Annual Report 2016. Geneva: World of inequality in childhood immunization in Uganda. Health Organization; 2016. 6. Republic of Uganda. Uganda National Expanded To access interactive visuals: Programme on Immunization Multi Year Plan 2012– 2016: Updated cMPY August 2013. Geneva: World SCAN HERE: or VISIT: Health Organization and UNICEF; 2013. http://apps.who.int/gho/ 7. Immunisation Act, 2016 [Internet]. Kampala: Republic of data/view.wrapper.HE- Uganda; 2016 (https://www.sabin.org/sites/sabin.org/ VIZ21?lang=en&menu=hide files/uganda_immunization_act_2015.pdf, accessed 27 May 2018). To access interactive tables, see Appendix 1.
49 13. Multicountry assessment
Integrating the country-specific findings presented chapter begins by comparing DTP3 immunization in Chapters 3 to 12, this chapter provides a coverage nationally and by within-country multicountry overview of DTP3 immunization subgroups. Then, the chapter assesses differences coverage and inequality across the Gavi, the in within-country inequality using concentration Vaccine Alliance 10 tier-1 countries. A multicountry curves and inequality summary measures (wealth- perspective permits benchmarking – the comparison and education-related concentration indices). of data from similar countries to get a sense of how Finally, the chapter documents differences in the any one country performs in relation to others. This adjusted associations between DTP3 coverage and selected socioeconomic, demographic and geographic factors. The interactive visuals and Interactive visuals and tables tables that accompany the report provide additional Interactive visuals and tables facilitate further exploration opportunities for multicountry data exploration. of inequality in childhood immunization across the 10 priority countries. Descriptive overview To access interactive visuals: National childhood immunization coverage SCAN HERE: or VISIT: Countries reported variable levels of national DTP3 http://apps.who.int/gho/ immunization coverage (Figure 13.1). The national data/view.wrapper.HE- immunization coverage was markedly different VIZ22?lang=en&menu=hide across countries: the countries with the lowest coverage were Chad (34%) and Nigeria (39%); Kenya reported the highest national coverage To access interactive tables, see Appendix 1. at 90%.
FIGURE 13.1. National DTP3 immunization coverage among one-year-olds in 10 priority countries (DHS 2012–2016 and NFHS 2015)
100 90.1
78.8 79.3 80 72.2 65.3 ) 60.5 % ( 60 58.3 53.2 a g e o v e r C 40 38.5 33.8
20
0 Afghanistan Chad Democratic Ethiopia India Indonesia Kenya Nigeria Pakistan Uganda Republic of the Congo
50 13. Multicountry assessment
Disaggregation by background characteristics
Child’s sex The 10 priority countries demonstrated almost no points or less; Pakistan reported inequality of 4 variation in DTP3 immunization coverage according percentage points, favouring boys. In all cases, to the sex of the child (Figure 13.2). In nine of the the differences between boys and girls were not 10 countries, the gap was around 2 percentage statistically significant.
FIGURE 13.2. DTP3 immunization coverage among one-year-olds by child’s sex in 10 priority countries (DHS 2012–2016 and NFHS 2015)
Afghanistan
Chad
Democratic Republic of the Congo
Ethiopia
India
Indonesia
Kenya
Nigeria
Pakistan
Uganda
0 10 20 30 40 50 60 70 80 90 100 Coverage (%)
Female Male
Birth order Overall, DTP3 immunization coverage tended to be birth order, with gaps between the lowest and highest among 1st born children and lower among highest birth order group spanning more than higher birth order children (Figure 13.3). Across 20 percentage points; two countries (Chad and countries, coverage was markedly lower among Democratic Republic of the Congo), in contrast, 6th born children (or higher) than across other showed little variation in immunization coverage subgroups, with some exceptions. India, Indonesia, according to birth order (less than 5 percentage Nigeria and Pakistan presented particularly points difference between the best- and worst- large differentials in immunization coverage by performing subgroups).
51 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
FIGURE 13.3. DTP3 immunization coverage among one-year-olds by birth order in 10 priority countries (DHS 2012–2016 and NFHS 2015)
Afghanistan
Chad
Democratic Republic of the Congo
Ethiopia
India
Indonesia
Kenya
Nigeria
Pakistan
Uganda
0 10 20 30 40 50 60 70 80 90 100 Coverage (%)
1st born 2nd-3rd born 4th-5th born 6th born or higher
Mother’s age at birth DTP3 immunization coverage was highest among There was more across country variation in DTP3 children whose mothers were 20–34 years in coverage among the subgroup of mothers aged 15– seven of the study countries (with the exception 19 years than in either of the other two subgroups. of Afghanistan, India and Kenya) (Figure 13.4). In While the difference between the country with the the Democratic Republic of the Congo, however, highest and lowest coverage for the 15–19 years the differences in mean coverage between age subgroup was 71 percentage points, this difference subgroups were not statistically significant. was 55 percentage points for the 20–34 years subgroup and 51 percentage points for the 35–49 years subgroup.
FIGURE 13.4. DTP3 immunization coverage among one-year-olds by mother’s age at birth in 10 priority countries (DHS 2012–2016 and NFHS 2015)
Afghanistan
Chad
Democratic Republic of the Congo
Ethiopia
India
Indonesia
Kenya
Nigeria
Pakistan
Uganda
0 10 20 30 40 50 60 70 80 90 100 Coverage (%)
15-19 years 20-34 years 35-49 years 52 13. Multicountry assessment
Mother’s education Across all 10 priority countries, DTP3 immunization coverage were larger between the no education coverage was consistently higher among children and primary education subgroups than between the of mothers with more education (Figure 13.5). primary education and higher education subgroups. For example, median coverage for mothers with The Democratic Republic of the Congo and Uganda, primary school was 14 percentage points higher by contrast, showed notably larger increases in than for mothers with no education. coverage in higher education subgroups relative to lower education subgroups. The pattern of inequality across mother’s education subgroups differed across countries, noting that Within each education subgroup, DTP3 coverage some countries had four education subgroups varied greatly across countries. The range in (India, Indonesia, Nigeria, Pakistan and Uganda) coverage across countries in the subgroup of and other countries had three (Afghanistan, Chad, mothers with no education was 66 percentage Democratic Republic of the Congo, Ethiopia and points, spanning from 12% in Nigeria to 78% in Kenya). For instance, Chad and Ethiopia and to some Kenya. In the subgroup with higher education, extent, Nigeria, demonstrated a linear gradient the gap between the best- and worst-performing with approximately similar increases in coverage country was 39 percentage points (from 54% in across education subgroups. In Afghanistan, India, Chad to 93% in Kenya), indicating a lower but still Indonesia, Kenya and Pakistan, differences in relevant variation in coverage.
FIGURE 13.5. DTP3 immunization coverage among one-year-olds by mother’s education in 10 priority countries (DHS 2012–2016 and NFHS 2015)
Afghanistan
Chad
Democratic Republic of the Congo
Ethiopia
Kenya
0 10 20 30 40 50 60 70 80 90 100 Coverage (%)
No education Primary school Secondary school or more
India
Indonesia
Nigeria
Pakistan
Uganda
0 10 20 30 40 50 60 70 80 90 100 Coverage (%)
No education Primary school Secondary school More than secondary school
53 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Household economic status Figure 13.6 shows DTP3 immunization coverage a linear relationship with approximately equal by wealth quintile for each of the 10 priority gains in coverage across wealth quintiles. countries. In the majority of countries (7 out of Indonesia and Pakistan demonstrated patterns of 10), immunization coverage increased in each marginal exclusion where the poorest quintile had progressively wealthier quintile; in 9 out of 10 substantially lower coverage than the other four countries, DTP3 immunization coverage was quintiles. significantly higher in the richest quintile compared to the poorest. Uganda was the only country Looking at cross-country performance by quintiles that reported no significant difference between revealed greatest variation in the poorest quintile, the richest and poorest quintiles. The sharpest and least variation among the richest quintile. gradients were evident in Nigeria, Pakistan and The range between the quintile 1 subgroup with Ethiopia where, relative to the poorest quintile, the highest coverage (84% in Kenya) and the coverage in the richest quintile was about 10, 3 quintile 1 subgroup with the lowest coverage (7% and 2 times higher, respectively. Nigeria presented in Nigeria) was 76 percentage points. Among the largest absolute inequality, with a rich–poor performance in quintile 5, this range was much difference spanning over 70 percentage points. narrower. With the exception of the outlier of Chad, which showed a notably low level of coverage even Different patterns of inequality were detected among the richest, the subgroup performance across countries. For example, Nigeria showed ranged from 70% in Afghanistan to 93% in Kenya: a queuing pattern where coverage increased in 23 percentage points.
FIGURE 13.6. DTP3 immunization coverage among one-year-olds, by household economic status in 10 priority countries (DHS 2012–2016 and NFHS 2015)
Afghanistan
Chad
Democratic Republic of the Congo
Ethiopia
India
Indonesia
Kenya
Nigeria
Pakistan
Uganda
0 10 20 30 40 50 60 70 80 90 100 Coverage (%)
Quintile 1 (poorest) Quintile 2 Quintile 3 Quintile 4 Quintile 5 (richest)
54 13. Multicountry assessment
Place of residence Urban areas tended to report higher levels of Across countries, the variation in the level of childhood immunization coverage than rural areas; coverage in rural areas was larger than in urban the magnitude of the urban–rural gap, however, areas: in the rural subgroup, coverage ranged from varied greatly across countries (Figure 13.7). While 25% in Nigeria to 89% in Kenya (64 percentage Nigeria and Ethiopia had an absolute difference points), whereas in the urban subgroup, coverage between urban and rural areas of 37 and 30 ranged from 40% in Chad to 91% in Kenya (51 percentage points, respectively, this difference was percentage points). minimal in India, Kenya and Uganda, amounting to 2–3 percentage points.
FIGURE 13.7. DTP3 immunization coverage among one-year-olds by place of residence in 10 priority countries (DHS 2012–2016 and NFHS 2015)
Afghanistan
Chad
Democratic Republic of the Congo
Ethiopia
India
Indonesia
Kenya
Nigeria
Pakistan
Uganda
0 10 20 30 40 50 60 70 80 90 100 Coverage (%)
Rural Urban
Wealth- and education-related inequalities
Concentration indices Erreygers concentration indices (ECI) depicting above, the largest inequality was observed in the magnitude of wealth- and education-related Nigeria where the wealth-related ECI was 0.612, inequality in each of the 10 priority countries are followed by Pakistan with an ECI value of 0.435. shown in Figures 13.8 and 13.9. The wealth-related The Democratic Republic of the Congo, Ethiopia ECI for DTP3 immunization coverage was positive and Indonesia had similar ECI values, falling just and statistically significant for all countries, except short of 0.3, confirming that DTP3 coverage in Uganda, indicating that immunization coverage these countries was concentrated among the richer. among one-year-olds tended to be concentrated Chad, India and Kenya had lower levels of wealth- among wealthier households (Figure 13.8). related inequality, and Uganda demonstrated no Consistent with the concentration curves shown inequality.
55 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
FIGURE 13.8. Wealth-related Erreygers concentration index for DTP3 immunization coverage among one-year-olds in 10 priority countries (DHS 2012–2016 and NFHS 2015)
0.612 0.6
0.5 0.435 n d e x i 0.4 o n i a t r 0.296 0.3 0.287 0.261 o n c e t C 0.2 0.191
0.130 0.100 0.1 0.079
0.010 0.0 Afghanistan Chad Democratic Ethiopia India Indonesia Kenya Nigeria Pakistan Uganda Republic of the Congo
The ECI values for education-related inequality inequality (ECI=0.334). Uganda reported a were all positive, indicating higher coverage small but statistically significant positive ECI for among children of mothers with more education education-related inequality (0.055), suggesting (Figure 13.9). Nigeria was the country with the that, although DTP3 coverage did not differ by largest education-related inequality in DTP3 household economics status, it did vary by mother’s immunization coverage (ECI=0.613), and Pakistan level of education. reported the second largest education-related
FIGURE 13.9. Education-related Erreygers concentration index for DTP3 immunization coverage among one-year-olds in 10 priority countries (DHS 2012–2016 and NFHS 2015)
0.613 0.6
0.5 n d e x i 0.4 o n i 0.334 a t r 0.3
o n c e t 0.220 0.226 C 0.200 0.2 0.184 0.150 0.135
0.1 0.091 0.055
0.0 Afghanistan Chad Democratic Ethiopia India Indonesia Kenya Nigeria Pakistan Uganda Republic of the Congo
56 13. Multicountry assessment
Concentration curves Concentration curves illustrating wealth-related households. Countries demonstrated variation in inequality in DTP3 immunization coverage how far the concentration curve fell from the line across the 10 priority countries are displayed of equality, indicating differing levels of wealth- in Figure 13.10. The concentration curves show related inequality. Nigeria had the most pronounced the cumulative share of DTP3 coverage plotted inequality, depicted by the concentration curve that against the cumulative share of children ranked falls below other curves at all points, whereas the from poorest to richest household. concentration curve for Kenya fell close to the line of equality. The concentration curve for Uganda is For 9 out of 10 countries (excepting Uganda), almost on the line of equality, indicating that DTP3 the concentration curve was below the line of immunization coverage was equally distributed equality, demonstrating that DTP3 coverage was across children of different household economic more concentrated among children from richer status.
FIGURE 13.10. Concentration curves for DTP3 immunization coverage by household economic status in 10 priority countries (DHS 2012–2016 and NFHS 2015) 1.0 0.8 0.6 0.4 0.2 Cumulative share of DTP3 immunization coverage 0.0 0.0 0.2 0.4 0.6 0.8 1.0
Cumulative share of children ranked by household economic status
Afghanistan Indonesia
Chad Kenya
Democratic Republic of the Congo Nigeria
Ethiopia Pakistan
India Uganda
Note: The red diagonal line represents the line of equality.
57 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Characteristics associated with in Nigeria and Pakistan, this chance was 7 times DTP3 immunization coverage higher (Figure 13.11). across countries The odds of DTP3 immunization were higher Figures 13.11 and 13.12 present estimated odds among the higher education subgroups relative to ratios for the adjusted association between DTP3 the subgroup with no education (Figure 13.12). The immunization coverage and select socioeconomic, adjusted association between mother’s education demographic and geographic factors across priority and coverage, however, varied largely across countries. countries. The association was not statistically significant in the Democratic Republic of the Congo In all countries except Uganda, children in the and Ethiopia. In Nigeria and Indonesia, however, richest quintile had a significantly higher chance the odds of being vaccinated were 6.6 and 4.4 of being vaccinated compared with those from the times higher, respectively, among the most-educated poorest quintile, after controlling for other factors; subgroup compared to the least-educated subgroup.
FIGURE 13.11. Adjusted associations between household economic status and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for the richest quintile compared with the poorest quintile (DHS 2012–2016 and NFHS 2015)
Afghanistan 2.68
Chad 2.48
Democratic Republic of the Congo 2.89
Ethiopia 2.74
India 2.26
Indonesia 3.36
Kenya 2.61
Nigeria 7.27
Pakistan 7.05
Uganda 1.29
1 2 4 8 Odds ratio
Significant (p-value<0.05) Not significant
Note: The adjusted odds ratios compare quintile 5 against the reference subgroup of quintile 1, adjusting for other characteristics: child’s sex; birth order; mother’s age at birth; mother’s education; mother’s ethnicity or caste/tribe (where applicable); sex of household head (except Afghanistan); place of residence; and subnational region.
58 13. Multicountry assessment
FIGURE 13.12. Adjusted associations between mother’s education and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for the most educated compared with the least educated (DHS 2012–2016 and NFHS 2015)
Afghanistan 1.87
Chad 1.95
Democratic Republic of the Congo 1.44
Ethiopia 1.66
India 1.89
Indonesia 4.37
Kenya 2.54
Nigeria 6.64
Pakistan 2.38
Uganda 2.17
1 2 4 Odds ratio
Significant (p-value<0.05) Not significant
Notes: In Afghanistan, Chad, the Democratic Republic of the Congo, Ethiopia and Kenya, the most-educated subgroup was the combined category of “secondary school or more”, applied due to small sample sizes of women with higher levels of education; in all other countries, the most-educated subgroup was “more than secondary school”. The adjusted odds ratios compare the most-educated subgroup against the reference subgroup of no education, adjusting for other characteristics: child’s sex; birth order; mother’s age at birth; mother’s ethnicity or caste/tribe (where applicable); sex of household head (except Afghanistan); household economic status; place of residence; and subnational region.
After controlling for other factors, there were mixed close to and not statistically different from the associations between DTP3 coverage and mother’s subgroup with mothers aged 15–19 years; however, age at birth (Figure 13.13). In the Democratic Republic in Ethiopia and Nigeria, the odds of coverage in the of the Congo and Kenya, the odds of coverage among 20–34 years subgroup were around twice as high, the subgroup with mothers aged 20–34 years were compared to the 15–19 years subgroup.
FIGURE 13.13. Adjusted associations between mother’s age at birth and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for the 20–34 years subgroup compared with the 15–19 years subgroup (DHS 2012–2016 and NFHS 2015)
Afghanistan 1.72
Chad 1.48
Democratic Republic of the Congo 0.99
Ethiopia 2.08
India 1.24
Indonesia 1.48
Kenya 0.84
Nigeria 1.89
Pakistan 1.65
Uganda 1.69
1 2 Odds ratio
Significant (p-value<0.05) Not significant
Note: The adjusted odds ratios compare the 20–34 years subgroup against the reference subgroup of 15–19 years, adjusting for other characteristics: child’s sex; birth order; mother’s education; mother’s ethnicity or caste/tribe (where applicable); sex of household head (except Afghanistan); household economic status; place of residence; and subnational region. 59 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Six countries out of 10 did not report statistically born children compared to 6th born or higher; in significant associations between DTP3 immunization Kenya and Uganda the odds were 2.3 times as high coverage and birth order (Figure 13.14). In Indonesia, and in India, 1.8 times as high. however, the odds were 4.0 times higher among 1st
FIGURE 13.14. Adjusted associations between birth order and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for the 1st born compared with the 6th born or higher (DHS 2012–2016 and NFHS 2015)
Afghanistan 1.21
Chad 0.91
Democratic Republic of the Congo 0.88
Ethiopia 1.55
India 1.81
Indonesia 3.94
Kenya 2.30
Nigeria 1.38
Pakistan 1.46
Uganda 2.25
1 2 4 Odds ratio
Significant (p-value<0.05) Not significant
Note: The adjusted odds ratios compare the 1st born subgroup against the reference 6th born or higher subgroup, adjusting for other characteristics: child’s sex; mother’s age at birth; mother’s education; mother’s ethnicity or caste/tribe (where applicable); sex of household head (except Afghanistan); household economic status; place of residence; and subnational region.
Most countries (7 out of 9) did not report a households had a lower chance of coverage (0.6 significant association between DTP3 coverage and times) than children in male-headed households; the sex of household head (Figure 13.15). Ethiopia and in Nigeria, children in female-headed households Nigeria demonstrated significant, though opposite had 1.4 times higher chance of immunization than associations: in Ethiopia, children in female-headed children in male-headed households.
60
Notes: The adjusted odds ratios compare females against the reference subgroup males, adjusting for other characteristics: child’s sex; birth order; mother’s age at birth; mother’s education; mother’s ethnicity or caste/tribe (where applicable); household economic status; place of residence; and subnational region. Afghanistan was not included in this analysis because nearly all households (99%) were headed by males. 13. Multicountry assessment
FIGURE 13.15. Adjusted associations between sex of household head and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for females compared with males (DHS 2012–2016 and NFHS 2015)
Chad 1.00
Democratic Republic of the Congo 1.21
Ethiopia 0.57
India 0.97
Indonesia 0.89
Kenya 0.91
Nigeria 1.43
Pakistan 1.07
Uganda 0.82
0.5 1 Odds ratio Significant (p-value<0.05) Not significant
Notes: The adjusted odds ratios compare females against the reference subgroup males, adjusting for other characteristics: child’s sex; birth order; mother’s age at birth; mother’s education; mother’s ethnicity or caste/tribe (where applicable); household economic status; place of residence; and subnational region. Afghanistan was not included in this analysis because nearly all households (99%) were headed by males.
In 9 out of the 10 priority countries, there was no Children living in urban areas of India, however, statistically significant association between DTP3 had a significantly lower chance (0.75 times) of immunization coverage and place of residence, reporting high coverage comparing to those living after controlling for other factors (Figure 13.16). in rural areas.
FIGURE 13.16. Adjusted associations between place of residence and DTP3 immunization coverage among one-year-olds in 10 priority countries, calculated as odds ratio for urban compared with rural (DHS 2012–2016 and NFHS 2015)
Afghanistan 1.21
Chad 0.84
Democratic Republic of the Congo 1.25
Ethiopia 1.50
India 0.75
Indonesia 1.08
Kenya 0.90
Nigeria 1.26
Pakistan 1.31
Uganda 0.76
1 Odds ratio
Significant (p-value<0.05) Not significant
Note: The adjusted odds ratios compare the urban subgroup against the reference rural subgroup, adjusting for other characteristics: child’s sex; birth order; mother’s age at birth; mother’s education; mother’s ethnicity or caste/tribe (where applicable); sex of household head (except Afghanistan); household economic status; and subnational region.
61 14. Conclusion
Taking stock • Most countries (7 out of 10) reported inequality by birth order with lowest immunization coverage For many countries, universal childhood immuniza- among children born 6th or higher. tion has not yet been achieved. Taking a closer look at the Gavi, the Vaccine Alliance 10 tier-1 countries The ECI, a summary measure of absolute – Afghanistan, Chad, Democratic Republic of the inequality, drew on data from all subgroups to Congo, Ethiopia, India, Indonesia, Kenya, Nigeria, express within-country inequality by wealth and Pakistan and Uganda – this report described within- education. Nine out of 10 countries (all but Uganda) country inequalities in childhood immunization showed wealth-related inequality that favoured and probed further into factors associated with richer households, and all 10 countries showed DTP3 coverage. Using comparable DHS and NFHS education-related inequality that favoured children data and analysis methods, multicountry analyses with more-educated mothers. For both wealth were conducted to assess performance across the and education, Nigeria had the most pronounced 10 countries. inequality, followed by Pakistan. Kenya and Uganda were the countries with the lowest levels of wealth- The report conveys several important findings. and education-related inequality. Although the 10 priority countries were all home to large numbers of unvaccinated or under-vaccinated Multiple regression analyses permitted a closer children, the magnitude and patterns of inequality look at the associations between each of the varied markedly. The descriptive analyses of selected factors and childhood immunization disaggregated data across different dimensions coverage, adjusting for other characteristics. In of inequality revealed certain similarities and each of the 10 countries, the odds of childhood differences across the 10 countries. immunization coverage varied significantly across subnational regions. All countries demonstrated • None of the countries showed statistically positive associations between mother’s education significant inequality in childhood immunization level and childhood immunization coverage – in based on child sex. all countries but Uganda, this association was • Inequalities by mother’s education, household statistically significant (and in Uganda the economic status and subnational region were association was nearly significant). Most countries, prevalent in nearly all countries, with variable except for Uganda, reported positive associations patterns. between household economic status and childhood • In all five countries where disaggregated data by immunization coverage. mother’s ethnicity or caste/tribe were presented, inequalities were evident. In several cases, the patterns of childhood • All countries but one (Democratic Republic of immunization coverage suggested by the descriptive the Congo) reported inequality by mother’s age analyses were not significant after adjusting for at birth. other factors in the multiple regression analyses. For • The majority of countries (8 out of 10) reported example, after adjusting for other factors, just one statistically significant higher coverage in country (India) retained a significant association urban than rural areas; in six of these countries, between coverage and place of residence. differences were larger than 10 percentage points.
62 14. Conclusion
Compounded vulnerability and advantage
Children who experience multiple forms of disadvantage are less likely to be vaccinated than children who experience a single type of disadvantage. For example, a child who lives in an underserved region may be at an increased risk of remaining unvaccinated if the child also belongs to a poor household. Conversely, children who belong to multiple subgroups that hold advantages are more likely to be covered. In many countries, children who are 1st born and whose mothers are highly educated have compounded advantage. Certain living conditions and characteristics compound to exacerbate vulnerability or advantage. For instance: In Afghanistan in 2015, a child of a teenaged mother with no education had one third the chance of being vaccinated as a child of a mother 20–49 years of age with secondary education or higher; if the child of the uneducated, teenaged mother belonged to the poorest 20%, this chance dropped to one ninth (compared to a child of a highly educated mother aged 20–49 years in the richest 20%). In Chad in 2014–2015, a child of a mother aged 20–34 years with secondary education or higher and belonging to the richest 20% had up to 7.2 times higher chance of receiving DTP3 immunization compared with a child of a teenaged mother with no education, from the poorest 20% household. In Ethiopia in 2016, the chance of receiving the third dose of DTP vaccine was 6.7 times higher for a child whose mother was 20–49 years of age and primary school educated, and who lived in a male-headed household, compared with a child of a teenaged mother with no education in a female-headed household. In India in 2015–2016, children with highly educated mothers aged 20–49 years who belonged to the richest 20% of the population had a 5.3 times higher chance of being vaccinated, compared with children born to teenaged mothers with no education, in the poorest 20% of the population. In Indonesia in 2012, a child who was part of a household in the richest 20%, and whose mother was aged 35–49 years, had a 6.4 times greater chance of being vaccinated compared to a child living in a household of the poorest 20%, and whose mother was a teenager. In Kenya in 2014, children had a higher chance of being vaccinated if they belong to the richest 40% of households and their mother had at least primary school education: compared to those in the poorest 20% and whose mother had no education, their chances were 6.3 times higher. In Nigeria in 2013, children of mothers aged 20–34 years who were highly educated, living in a rich household in the South South region were among the most advantaged in terms of childhood immunization: their chance of being vaccinated was 300 times higher than children with teenaged mothers with no education, living in poor households in the North West region. In Pakistan in 2012–2013, a child of a mother aged 20–34 years with higher than secondary education and from the richest 20% of the population had a 28 times higher chance of being vaccinated, compared with a child of a teenaged mother with no education and from the poorest 20% of the population.
Note: the above examples do not necessarily represent the most advantaged or disadvantaged subgroups; there may be other combinations of compounded vulnerability and advantage.
63 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
Understanding the analysis The report covered only one indicator of childhood immunization (DTP3 immunization coverage), and Evaluating performance based on national the findings may not be directly applicable to other averages alone masks the situation in population childhood immunization indicators. Limitations subgroups. The analyses of this report drew from inherent to the underlying dataset were evident data disaggregated by a range of socioeconomic, in this report (for example, the omission of demographic and geographic factors; these data certain subnational regions in Pakistan). For some lend insight into the gaps in childhood immunization countries, low sample sizes for certain ethnic and coverage within countries. The findings help education subgroups necessitated the combination to reveal missed opportunities for childhood of multiple subgroups into one, changing the nature immunization by identifying poor-performing of the resulting analysis. subgroups and suggesting factors associated with lower coverage. A more nuanced and detailed understanding of the results and their underlying data can be derived from This report is novel because it combined several the interactive visuals and tables that accompany the types of analyses to yield a comprehensive report. For all disaggregated estimates, the interactive multicountry overview of inequalities in childhood visuals include information about population share – immunization and the factors that underlie the percentage of the population that is represented immunization coverage. Descriptive analyses of by a given population subgroup. Population share disaggregated data set the stage of how coverage offers important insight into interpreting findings that varied between subgroups. Expanding on this, arise from disaggregated data and associations. For concentration index, a summary measure of example, across countries, the percentage of children inequality, and concentration curves, the graphical of mothers with no education varied: in Afghanistan, demonstration, generated a concise representation 81% belonged to the no education subgroup, of the situation across economic status and whereas in Indonesia, less than 2% belonged to the education subgroups, which could be compared no education subgroup. The interactive visuals and across multiple countries (benchmarking). Then, tables indicate where estimates were based on low multiple regression analyses provided a closer sample sizes and include 95% confidence intervals examination of factors that contributed to coverage. for all estimates. Taken together, these analyses provided more insight into explanations of national coverage and Moving forward equity in immunization coverage and can be used to generate hypotheses for further investigation. The 10 countries featured in this report each face a distinct situation with regard to childhood All analyses have limitations. An awareness of immunization coverage. When paired with in-depth the relevant limitations of the underlying data knowledge of the country context nationally and and methods helps to avoid misinterpretation locally – including the political situation, existing of the findings. In this report, the analyses drew policies, living conditions, culture, etc. – the results from cross-sectional data, and thus claims about of this report can be used to inform equity-oriented causality should be avoided. While the data used in policies, programmes and practices to promote this report were all collected as part of the DHS or universal childhood immunization coverage. NFHS, the year of the most recent available survey varied for the 10 countries, ranging from 2012 to Characteristic patterns of inequality can be linked 2016. National coverage of childhood immunization to broad policy recommendations (1). Generally, can fluctuate over time, and the findings should low national levels of coverage are best addressed be understood as a reflection of the survey year. by universally oriented policy approaches that
64 14. Conclusion
Innov8 approach for reviewing national health programmes to leave no one behind The Innov8 approach for reviewing national health programmes to leave no one behind is a WHO resource that facilitates the application of national health inequality monitoring to national health programming. As part of the 8-step approach, Innov8 guides multidisciplinary review teams through identifying populations left behind, assessing barriers and facilitating factors, identifying mechanisms that lead to health inequalities, and strengthening monitoring and evaluation. For more information about Innov8 approach for reviewing national health programmes to leave no one behind, see: http://www.who. int/life-course/publications/innov8-technical-handbook/en/. also promote accelerated improvements in towards subgroups with lower coverage, such as disadvantaged subgroups. For instance, Chad, with larger families, poorer household, less-educated low national coverage of 34%, would benefit from mothers, and those living in poor-performing regions strengthening efforts aimed at overall improvement, (Banten, Maluku, Papua and West Kalimantan), with a focus on the five subnational regions that would be positioned to reduce inequalities and report coverage of 10% or less. Nigeria, a country further improve the national coverage. with similarly low national coverage (39%), demonstrated elevated within-country inequality Of the 10 priority countries featured in this according to many dimensions. In addition to a report, Uganda reported the lowest levels of focus on overall improvement, policy approaches inequality. Here, sustained efforts, supported by in Nigeria require special consideration of: northern ongoing monitoring, are warranted to ensure that regions (especially the North West and the North inequalities in childhood immunization coverage are East); poor households; families of less-educated minimized, and that all children have the chance to women (nearly half of mothers have no education); benefit from immunization. rural residents; and male-headed households. In Afghanistan and Ethiopia, DTP3 coverage was fairly This report drew from the latest publicly low overall, with pronounced inequality by certain available DHS and NFHS data to explore national characteristics, which indicate potential entry points performance and within-country inequalities. The for targeted policies and programmes. For instance, findings of this report serve as a basis to identify in Afghanistan, immunization coverage was areas for additional inquiry. The questions that arise almost zero among children of some ethnic groups from these findings are vast and may be pursued (especially the Nuristanis), and regions (including through diverse types of research. For example, Nooristan and Urozgan). Likewise, Ethiopia reported expanded quantitative analyses can be performed elevated inequality by mother’s ethnicity, with Affar, to delve into questions related to the inequalities Oromo and Somallie subgroups falling behind. described in the report, such as how the national performance and within-country inequalities have In contrast, countries with high inequality changed over time and which factors are associated alongside higher national levels of coverage require with these changes. In countries where the DHS or a targeted policy approach that prioritizes the other national surveys have a very large sample size most vulnerable. For instance, Kenya, with 90% and/or are representative at the state or province childhood immunization coverage, displayed level (including India, Indonesia and Pakistan), marginal exclusion patterns of inequality across further studies can be done to interrogate factors certain dimensions, indicating that renewed efforts associated with coverage at subnational levels. are warranted to reach the poorest, least-educated Investigations of a qualitative nature may explore and largest families. Relative to the other priority the root causes and drivers of inequality. Qualitative countries, Indonesia also had high national childhood studies are important in their own right, and to immunization coverage at 72%. Policies targeted add context to quantitative findings, for example,
65 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
exploring how and why inequalities arise. WHO society; progression towards universal health is now refining approaches (including qualitative coverage; and adoption of a comprehensive primary methods) to assess demand- and supply-side health care approach (7). This report demonstrates barriers to effective coverage with health services (2). that childhood immunization coverage is linked to factors beyond the health system. Coordinated One emerging area for further exploration is efforts across health and other sectors should aim childhood immunization among the urban poor. to reduce barriers to childhood immunization and With rising rates of rural-to-urban migration in reinforce strategies that work well. many countries, concerns about restricted access to health services, including childhood immunization, Detailed explorations of health inequalities, among those in informal settlements are mounting encompassing multiple health topics and the (3–5). At present, however, there is a paucity of factors that affect performance, contribute to wider data available about these subpopulations, and ambitions to pursue sustainable development. the extent of inequality and its root causes are not Pledging that no one will be left behind, the United fully understood. For instance, the DHS, a major Nations 2030 Agenda for Sustainable Development source of household survey data in many low- and requires a comprehensive understanding of where middle-income countries, does not systematically missed opportunities lie, and the barriers that capture this population through its sampling design. impede progress. Expanded and targeted data collection efforts, for both quantitative and qualitative research, are References required to better understand this issue. 1. Handbook on health inequality monitoring: with a special focus on low- and middle-income countries. Ongoing monitoring and evaluation of childhood Geneva: World Health Organization; 2013. immunization, including inequalities, drivers and 2. Towards universal coverage for preventive chemotherapy for Neglected Tropical Diseases: guidance for assessing determinants, are important to systematically track “who is being left behind and why”. Geneva: World the impact of policies and programmes, and to Health Organization; 2017. identify bottlenecks in programme delivery and 3. Awoh AB, Plugge E. Immunisation coverage in rural– implementation. To this end, health information urban migrant children in low- and middle-income countries (LMICs): a systematic review and meta- systems should be strengthened to adequately analysis. J Epidemiol Community Health. 2016 collect, manage and analyse data, and report on March;70(3):305–11. information about health and health determinants. In 4. Fotso J-C, Ezeh AC, Madise NJ, Ciera J. Progress many low- and middle-income countries, additional towards the child mortality millennium development investments are needed to establish components goal in urban sub-Saharan Africa: the dynamics of population growth, immunization, and access to of health information systems. This includes: data clean water. BMC Public Health [Internet]. 2007 collection through population health surveys, December (http://bmcpublichealth.biomedcentral. facility-based administrative data, as well as civil com/articles/10.1186/1471-2458-7-218, accessed 23 May 2018). registration and vital statistics and censuses; and 5. Soura AB, Mberu B, Elungata P, Lankoande B, capacity-building for data analysis and reporting. Millogo R, Beguy D et al. Understanding inequities in child vaccination rates among the urban poor: Overall, sustainable and equity-oriented childhood evidence from Nairobi and Ouagadougou Health and Demographic Surveillance Systems. J Urban Health. immunization programmes rely on strong health 2015 February;92(1):39–54. systems that are attentive and responsive to 6. Brearley L, Eggers R, Steinglass R, Vandelaer J. Applying the realities of the populations that they serve an equity lens in the Decade of Vaccines. Vaccine. 2013 (6). Equity-oriented health systems share certain April;31:B103–7. characteristics: intersectoral action to promote 7. Gilson L, Doherty J, Lowenson R, Francis V. Challenging health; inclusion of population groups and civil inequity through health systems. Final report. Knowledge Network on Health Systems. WHO Commission on the Social Determinants of Health. Geneva: World Health 66 Organization; 2007. Appendix 1
Interactive tables contain the complete set of data from the 10 priority countries, for further reference. Interactive tables The information in these tables corresponds Interactive tables include information about disaggregated with the descriptive overview and multivariate data and multiple regression models. regression analysis components of the report. To access the interactive tables: The descriptive overview tables include national average, disaggregated data estimates, 95% SCAN HERE: or VISIT: confidence intervals and population share; cases of http://apps.who.int/gho/ limited data availability are noted. The multivariate data/view.wrapper.HE- regression analysis tables include odds ratios and VIZ23?lang=en&menu=hide 95% confidence intervals, as well as information about the characteristics of the model.
67 Index
A indicator (DTP3 immunization coverage) xii, 4–5, 10, 14, 18, Afghanistan xiii, 10–13 22, 26, 30, 34, 38, 42, 46, 50 birth order 10, 11, 12, 13, 52, 60 monitoring 66 child’s sex 10, 11, 12, 13, 51 promoting 2 compounded vulnerability and advantage 63 Child’s sex 6, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, descriptive analysis 10–12 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, education-related inequality 12, 56 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 51, 62 Erreygers normalized concentration index 12, 55, 56 Comprehensive Multi-Year Plan 18, 22, 30, 34, 42 household economic status 11, 12, 13, 54, 57, 58 Concentration curve 6, 7, 57 interactive visuals 13 Concentration index see Erreygers normalized concentration mother’s age at birth 10, 11, 12, 13, 52, 59 index mother’s education 10, 11, 12, 13, 53, 59 mother’s ethnicity 10, 11 D multiple regression analysis 12–13 Data analysis xii, 5–8, 64 national DTP3 immunization coverage 10, 50 Data sources xii, 4 place of residence 11, 12, 13, 55, 61 Democratic Republic of the Congo xiii, 18–21 policy recommendations 65 birth order 18, 19, 20, 21, 51, 52, 60 sex of household head 10, 11 child’s sex 18, 19, 20, 21, 51 subnational region 11, 12, 13 descriptive analysis 18–20 wealth-related inequality 12, 56, 57 education-related inequality 19–20, 56 Erreygers normalized concentration index 19–20, 55, 56 B household economic status 18, 19, 20–21, 54, 57, 58 Benchmarking 50 interactive visuals 21 Birth order 6, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, mother’s age at birth 18, 19, 20, 21, 52, 59 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, mother’s education 18, 19, 20, 21, 53, 58, 59 40, 41, 42, 43, 44, 45, 46, 47, 48–49, 51–52, 60, 62 mother’s ethnicity 18, 19, 20, 21 multiple regression analysis 20–21 C national DTP3 immunization coverage 18, 50 Chad xiii, 14–17 place of residence 18, 19, 20, 21, 55, 61 birth order 14, 15, 16, 17, 51, 52, 60 sex of household head 18, 19, 20, 21, 61 child’s sex 14, 15, 16, 17, 51 subnational region 18–19, 20 compounded vulnerability and advantage 63 wealth-related inequality 19–20, 55, 56, 57 descriptive analysis 14–16 Demographic and Health Surveys (DHS) 4 education-related inequality 15–16, 56 Descriptive analysis xii, 5–7, 10–12, 14–16, 18–20, 22–24, Erreygers normalized concentration index 15–16, 55, 56 26–28, 30–32, 34–36, 38–40, 42–44, 46–48, 50–57, 62 household economic status 14, 15, 16, 17, 54, 57, 58 DTP3 immunization coverage xii, 4–5, 10, 14, 18, 22, 26, 30, interactive visuals 17 34, 38, 42, 46, 50 mother’s age at birth 14, 15, 16, 17, 52, 59 mother’s education 14, 15, 16, 17, 53, 59 E multiple regression analysis 16–17 Education-related inequality 6–7, 12, 15–16, 19–20, 24, 28, national DTP3 immunization coverage 14, 50 31–32, 36, 39–40, 43–44, 47–48, 55–57, 62 place of residence 14, 15, 16, 17, 55, 61 Erreygers normalized concentration index (ECI) 6, 12, 15–16, policy recommendations 65 19–20, 24, 28, 31–32, 36, 39–40, 43–44, 47–48, 55–56, 62 sex of household head 14, 15, 16, 17, 61 Ethiopia xiii–xiv, 22–25 subnational region 14–15, 16 birth order 22, 23, 24, 25, 52, 60 wealth-related inequality 15–16, 55, 56, 57 child’s sex 22, 23, 24, 25, 51 Childhood immunization compounded vulnerability and advantage 63 child mortality reduction 1 descriptive analysis 22–24 challenges to 1, 14, 18, 22, 26, 30, 38, 42, 46 education-related inequality 24, 56
68 Index
Erreygers normalized concentration index 24, 55, 56 Erreygers normalized concentration index 31–32, 55, 56 household economic status 23, 24, 25, 54, 57, 58 household economic status 30, 31, 32, 33, 54, 57, 58 interactive visuals 25 interactive visuals 33 mother’s age at birth 22, 23, 24, 25, 52, 59 mother’s age at birth 30, 31, 32, 33, 52, 59 mother’s education 22, 23, 24, 25, 53, 58, 59 mother’s education 30, 31, 32, 33, 53, 58, 59 mother’s ethnicity 22–23 multiple regression analysis 32–33 multiple regression analysis 24–25 national DTP3 immunization coverage 30, 50 national DTP3 immunization coverage 22, 50 place of residence 31, 32, 33, 55, 61 place of residence 23, 24, 25, 55, 61 policy recommendations 65 policy recommendations 65 sex of household head 30, 31, 32, 33, 61 sex of household head 23, 24, 25, 60, 61 subnational region 31, 32 subnational region 23, 24–25 wealth-related inequality 31–32, 55, 56, 57 wealth-related inequality 24, 55, 56, 57 Innov8 approach for reviewing national health programmes to Expanded Programme on Immunization 1, 10, 18, 22, 26, 34, leave no one behind 65 38, 42, 46 Interaction effects 8 Revitalization Plan 46 Interactive visual and tables 8, 13, 17, 21, 25, 29, 33, 37, 41, 45, 49, 50, 64, 67 G Gavi, the Vaccine Alliance xii, 1 J Global Vaccine Action Plan 1 Joint hypothesis testing 8 H K Health information systems 66 Kenya xiv–xv, 34–37 Health systems 1, 66 birth order 34, 35, 36, 37, 52, 60 Household economic status 5, 6, 7, 11, 12, 13, 14, 15, 16, 17, child’s sex 34, 35, 36, 37, 51 18, 19, 20–21, 23, 24, 25, 26–27, 28, 29, 30, 31, 32, 33, 35, 36, compounded vulnerability and advantage 63 37, 38–39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 54, 57, 58, 62 descriptive analysis 34–36 education-related inequality 36, 56 I Erreygers normalized concentration index 36, 55, 56 India xiv, 26–29 household economic status 35, 36, 37, 54, 57, 58 birth order 26, 27, 28, 29, 51, 52, 60 interactive visuals 37 child’s sex 26, 27, 28, 29, 51 mother’s age at birth 34, 35, 36, 37, 52, 59 compounded vulnerability and advantage 63 mother’s education 34, 35, 36, 37, 53, 59 descriptive analysis 26–28 mother’s ethnicity 34, 35, 36, 37 education-related inequality 28, 56 multiple regression analysis 36–37 Erreygers normalized concentration index 28, 55, 56 national DTP3 immunization coverage 34, 50 household economic status 26–27, 28, 29, 54, 57, 58 place of residence 35, 36, 37, 55, 61 interactive visuals 29 policy recommendations 65 mother’s age at birth 26, 27, 28, 29, 52, 59 sex of household head 35, 36, 37, 61 mother’s education 26, 27, 28, 29, 53, 59 subnational region 35, 36, 37 mother’s caste/tribe 26, 27, 28, 29 wealth-related inequality 36, 55, 56, 57 multiple regression analysis 28–29 national DTP3 immunization coverage 26, 50 L National Family Health Survey (NFHS) 4, 26 Logistic regression model xii, 7 place of residence 27, 28, 29, 55, 61 sex of household head 26, 27, 28, 29, 61 M subnational region 27, 28, 29 Monitoring childhood immunization 66 wealth-related inequality 28, 55, 56, 57 Mother’s age at birth 5, 6, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, Indonesia xiv, 30–33 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, birth order 30, 31, 32, 33, 51, 52, 60 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 52, 59, 62 child’s sex 30, 31, 32, 33, 51 Mother’s education 5, 6, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, compounded vulnerability and advantage 63 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, descriptive analysis 30–32 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 53, 58, 59, 62 education-related inequality 31–32, 56 Mother’s ethnicity or caste/tribe 5, 6, 10, 11, 18, 19, 20, 21,
69 EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION
22–23, 26, 27, 28, 29, 34, 35, 36, 37, 62 23, 24, 25, 27, 28, 29, 31, 32, 33, 35, 36, 37, 39, 40, 41, 43, 44, Multicollinearity 8 46, 47, 48, 49, 55, 61, 62 Multiple regression analysis xii, 5, 7–8, 12–13, 16–17, 20–21, Policy recommendations 64–65 24–25, 28–29, 32–33, 36–37, 40–41, 44–45, 48–49, 62 Population share 8, 64 N R National Family Health Survey (NFHS) 4, 26 Rural-to-urban migration 66 National immunization programmes 1, 10, 18, 22, 26, 30, 34, 38, 42, 46 S Nigeria xv, 38–41 Sex differences see Child’s sex birth order 38, 39, 40, 41, 51, 52, 60 Sex of household head 6, 10, 11, 14, 15, 16, 17, 18, 19, 20, 21, child’s sex 38, 39, 40, 41, 51 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 35, 36, 37, 39, 40, 41, compounded vulnerability and advantage 63 43, 44, 45, 46, 47, 48, 49, 60–61 descriptive analysis 38–40 Subnational region 5, 6, 11, 12, 13, 14–15, 16, 18–19, 20, 23, education-related inequality 39–40, 56 24–25, 27, 28, 29, 31, 32, 35, 36, 37, 39, 40, 41, 43, 44–45, Erreygers normalized concentration index 39–40, 55, 56 46–47, 48, 62 household economic status 38–39, 40, 41, 54, 57, 58 interactive visuals 41 U mother’s age at birth 38, 39, 40, 41, 52, 59 Uganda xv–xvi, 46–49 mother’s education 38, 39, 40, 41, 53, 58, 59 birth order 46, 47, 48–49, 52, 60 multiple regression analysis 40–41 child’s sex 46, 47, 48, 49, 51 national DTP3 immunization coverage 38, 50 descriptive analysis 46–48 place of residence 39, 40, 41, 55, 61 education-related inequality 47–48, 56 policy recommendations 65 Erreygers normalized concentration index 47–48, 55, 56 sex of household head 39, 40, 41, 60, 61 household economic status 46, 47, 48, 49, 54, 57, 58 subnational region 39, 40, 41 interactive visuals 49 wealth-related inequality 39–40, 55, 56, 57 mother’s age at birth 46, 47, 48, 52, 59 mother’s education 46, 47, 48, 49, 53, 59 O multiple regression analysis 48–49 Odds ratio 7–8 national DTP3 immunization coverage 46, 50 place of residence 46, 47, 48, 49, 55, 61 P policy recommendations 65 P-value 8 sex of household head 46, 47, 48, 49, 61 Pakistan xv, 42–45 subnational region 46–47, 48 birth order 42, 43, 44, 45, 51, 52, 60 wealth-related inequality 47–48, 55, 56, 57 child’s sex 42, 43, 44, 45, 51 United Nations 2030 Agenda for Sustainable Development 66 compounded vulnerability and advantage 63 Urban poor 66 Comprehensive Multi-Year Plan 42 descriptive analysis 42–44 W education-related inequality 43–44, 56 Wealth-related inequality 6–7, 12, 15–16, 19–20, 24, 28, Erreygers normalized concentration index 43–44, 55, 56 31–32, 36, 39–40, 43–44, 47–48, 55–57, 62 household economic status 42, 43, 44, 45, 54, 57, 58 World Health Organization (WHO) interactive visuals 45 Expanded Programme on Immunization 1, 10, 18, 22, 34, mother’s age at birth 42, 43, 44, 45, 52, 59 38, 42, 46 mother’s education 42, 43, 44, 45, 53, 59 Global Vaccine Action Plan 1 multiple regression analysis 44–45 Innov8 approach 65 national DTP3 immunization coverage 42, 50 place of residence 43, 44, 55, 61 sex of household head 43, 44, 45, 61 subnational region 43, 44–45 wealth-related inequality 43–44, 55, 56, 57 Place of residence 6, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21,
70
http://www.who.int/gho/health_equity/report_2018_immunization/en/ Childhood immunization Childhood INTERACTIVE VISUALIZATION OF HEALTH DATA HEALTH OF VISUALIZATION INTERACTIVE EXPLORATIONS OF INEQUALITY EXPLORATIONS
EXPLORATIONS OF INEQUALITY: CHILDHOOD IMMUNIZATION WHO SWITZERLAND 20, AVENUE APPIA 20, AVENUE CH-1211 GENEVA 27 CH-1211 GENEVA WORLD HEALTH ORGANIZATION HEALTH WORLD HEALTH METRICS AND MEASUREMENT CLUSTER METRICS AND MEASUREMENT HEALTH DEPARTMENT OF INFORMATION, EVIDENCE AND RESEARCH OF INFORMATION, DEPARTMENT ISBN 978 92 4 156561 5 http://www.who.int/gho/health_equity/report_2018_immunization/en/