Increased immunization coverage addresses the equity gap in Nepal

Abstract Objective To compare immunization coverage and equity distribution of coverage between 2001 and 2014 in Nepal. Methods We used data from the Demographic and Health Surveys carried out in 2001, 2006 and 2011 together with data from the 2014 Multiple Indicator Cluster Survey. We calculated the proportion, in mean percentage, of children who had received bacille Calmette–Guérin (BCG) vaccine, three doses of polio vaccine, three doses of diphtheria–pertussis–tetanus (DPT) vaccine and measles vaccine. To measure inequities between wealth quintiles, we calculated the slope index of inequality (SII) and relative index of inequality (RII) for all surveys. Findings From 2001 to 2014, the proportion of children who received all vaccines at the age of 12 months increased from 68.8% (95% confidence interval, CI: 67.5–70.1) to 82.4% (95% CI: 80.7–84.0). While coverage of BCG, DPT and measles immunization statistically increased during the study period, the proportion of children who received the third dose of polio vaccine decreased from 93.3% (95% CI: 92.7–93.9) to 88.1% (95% CI: 86.8–89.3). The poorest wealth quintile showed the greatest improvement in immunization coverage, from 58% to 77.9%, while the wealthiest quintile only improved from 84.8% to 86.0%. The SII for children who received all vaccines improved from 0.070 (95% CI: 0.061–0.078) to 0.026 (95% CI: 0.013–0.039) and RII improved from 1.13 to 1.03. Conclusion The improvement in immunization coverage between 2001 and 2014 in Nepal can mainly be attributed to the interventions targeting the disadvantaged populations.


Introduction
Immunization is a proven, cost-effective intervention to reduce morbidity and mortality from vaccine-preventable diseases. 1 Each year immunization averts 2.5 million deaths in children younger than 5 years. 2 Globally in 2011, 103 million (83% of total) children received all three doses of diphtheria-pertussistetanus (DPT3) vaccine, but an estimated 22 million children did not complete such vaccination. 1 Gaps in immunization coverage exist between and within countries, and in some places, the gap is increasing. For example, the average DPT3 coverage in low-income countries was 15 percentage points lower than that of high-income countries in 2011. 1,3,4 To reach universal immunization coverage and to increase equity, countries need to focus on targeted interventions that reach the most disadvantaged populations, rather than only focusing on increasing coverage at the national level. 1,5 For example, countries with the most rapid increase in overall coverage of measles vaccination also show the greatest improvement in coverage of the population in the poorest wealth quintile. 6 One intervention to reach disadvantaged populations is by targeted communication, which aims to create a demand for immunization in a population. Communication through community mobilization, where influential local leaders support vaccination, could have a positive influence on the community's trust and willingness to vaccinate their children. 7 Which communication intervention to use in hard-to-reach populations to improve immunization coverage has to be based on which vaccination barriers these populations are facing.
To increase immunization coverage in Nepal, the government has invested in efforts and resources to improve the service delivery system of the national immunization programme. 8 During the last two decades, national health policies and a health sector strategy have been implemented to improve coverage and the health and survival of children. [9][10][11] One of the main strategic pillars of the National communication strategy on maternal, newborn, child health 2011-2016, is community mobilization of unvaccinated and under-vaccinated communities to increase coverage. 12 Furthermore, to increase access to vaccination in geographic areas with low coverage, the government has identified bottlenecks in supply of services and has mobilized local resources to intensify outreach clinics and mobile clinics. [13][14][15] The Nepalese National Immunization Programme, created in 1979, aims to vaccinate all children in the country. The programme is guided by a comprehensive multi-year plan for 2017-2022. 16 According to this plan, the public sector is the primary provider of immunization services, although the private sector is increasingly providing such services. The government provides all vaccine included in the programme freeof-charge. In 1989, the programme provided immunization services for six vaccine-preventable diseases, which were oral polio vaccine, DPT vaccine, tetanus toxoid, bacille Calmette-Guérin (BCG) for tuberculosis and measles vaccine. Between 2000 and 2015, the programme introduced six new vaccines: hepatitis B vaccine, Haemophilus influenza B vaccine, rubella vaccine, Japanese encephalitis vaccine, injectable polio vaccine and pneumococcal vaccine. In 2015, the programme provided immunization for 11 vaccine-preventable diseases (Table 1).
Health facilities and outreach clinics are the main providers of immunization services in Nepal. Each district has about 3-5 outreach clinics per village. In urban areas of the district, government health facilities, municipal clinics, private hospitals and private clinics deliver immunization services, in coordination and collaboration with the district health offices.
The objective of this study was to investigate changes in immunization coverage and equity gaps between 2001 and 2014, by using data from nationwide surveys.
Objective To compare immunization coverage and equity distribution of coverage between 2001 and 2014 in Nepal. Methods We used data from the Demographic and Health Surveys carried out in 2001, 2006 and 2011 together with data from the 2014 Multiple Indicator Cluster Survey. We calculated the proportion, in mean percentage, of children who had received bacille Calmette-Guérin (BCG) vaccine, three doses of polio vaccine, three doses of diphtheria-pertussis-tetanus (DPT) vaccine and measles vaccine. To measure inequities between wealth quintiles, we calculated the slope index of inequality (SII) and relative index of inequality (RII) for all surveys. Findings From 2001 to 2014, the proportion of children who received all vaccines at the age of 12 months increased from 68.8% (95% confidence interval, CI: 67.5-70.1) to 82.4% (95% CI: 80.7-84.0). While coverage of BCG, DPT and measles immunization statistically increased during the study period, the proportion of children who received the third dose of polio vaccine decreased from 93.3% (95% CI: 92.7-93.9) to 88.1% (95% CI: 86.8-89.3). The poorest wealth quintile showed the greatest improvement in immunization coverage, from 58% to 77.9%, while the wealthiest quintile only improved from 84.8% to 86.0%. The SII for children who received all vaccines improved from 0.070 (95% CI: 0.061-0.078) to 0.026 (95% CI: 0.013-0.039) and RII improved from 1.13 to 1.03. Conclusion The improvement in immunization coverage between 2001 and 2014 in Nepal can mainly be attributed to the interventions targeting the disadvantaged populations.
Equity distribution of immunization coverage in Nepal Ashish KC et al.

Survey data
We

Variables
For comparison reasons, we chose immunizations that were included in all four surveys. The variables comprised of: BCG vaccination at birth; three doses of polio vaccine at 1, 2 and 3 months of age; three doses of DPT vaccine at 1, 2 and 3 months of age; and measles vaccine at 9 months of age. To calculate fully immunized children, we created a composite variable for children who had received all of these vaccines at the age of 12 months.
To measure inequity, driven by socioeconomic determinants, 21 we used the theoretical framework of the Commission on Social Determinants of Health. This framework emphasizes people's social position based on structural determinants as the main driver of inequity. 21 To evaluate levels of inequity in immunization coverage, we stratified the immunization coverage data by socioeconomic determinants, including maternal education, area of residence, sex of child and family wealth. We did not include intermediary determinants, such as migration and maternal employment status. We dichotomized the determinants to detect differences between the most advantaged and the most disadvantaged populations. Maternal education was dichotomized as women with no education and women with a primary education or higher. Family areas of residence were divided between mountain areas and hill or terrain regions; the most disadvantaged families were those living in mountain areas. The families' wealth index was constructed through principal component analysis of scores based on families' ownership of durable assets, housing characteristics and access to services. From these indexes, we created wealth quintiles. These quintiles were then ranked from bottom to top, as poorest (first), poorer (second), middle (third), richer (fourth) and richest (fifth). To detect inequity between wealth quintiles, dichotomization was done as follows: the poor belonging to the poorest quintile, and the non-poor belonging to the other four quintiles. Even though the framework of the Commission on Social Determinants of Health considered ethnicity and caste as structural determinants, we were not able to include these variables in the analyses, since they were not captured in MICS.

Data analysis
For each vaccine we calculated the proportion of children who had received the recommended immunization for their age. For the composite variable, we calculated the proportion of children who had received all vaccination at the age of 12 months.
We calculated the slope index of inequality (SII) and the relative index of inequality (RII), which are two regression-based measures, for immunization coverage based on wealth quintiles. Both indexes take into account the whole socioeconomic distribution and remove variability in the size of socioeconomic groups as a source of variation in the magnitude of inequalities in health. SII is the absolute difference of immunization coverage between the richest and the poorest wealth quintile. 22 RII is the ratio of the immunization coverage estimate between wealth quintiles, calculated on the basis of the systematic association between coverage estimate and the socioeconomic status for all wealth quintiles. 22 An RII value of 1 indicates equity, while any value larger than 1 indicates inequity.
To analyse if there was an association between any of the socioeconomic determinants and immunization coverage, we used logistic regression models. We used univariate logistic regression to detect associations between coverage and socioeconomic variables, such as wealth quintile, geographic location, sex of child and maternal education. For those variables where we detected an association in the univariate logistic regression model, we did a multivariate logistic regression to adjust the effect of each socioeconomic variable on the immunization coverage. Results are presented as adjusted odds ratios (aORs). In this analysis, the sex of the child is an effect modifier of immunization coverage. Therefore, we checked whether sex of the child was associated with immunization coverage and found that no multicollinearity was detected during the initial analyses testing for effect modification.
We used SPSS version 20.0 (IBM, Chicago, United States of America) for the statistical analyses and a P-value of 0.05 was considered significant.

Results
After deleting survey responses for children with no information on immunization, we included 6931 (  Equity distribution of immunization coverage in Nepal Ashish KC et al. Overall, the surveys showed improvements in immunization coverage between 2001 and 2014 (Table 2) The equity in immunization coverage between wealth quintiles also improved over the study period (Table 3). For children fully immunized at the age of 12 months, the absolute difference in coverage between the wealthiest quintile and the poorest quintile significantly decreased between 2001 and 2014, from an SII of 0.070 (95% CI: 0.061-0.078) to an SII of 0.026 (95% CI: 0.013-0.039). Improvements in equity were also seen for coverage in BCG vaccination, from a SII of 0.054 (95% CI: 0.047-0.061) to anSII of 0.012 (95% CI: 0.005-0.019); in DPT vaccination from 0.068 (95% CI: 0.060-0.076) to 0.016 (95% CI: 0.005-0.026); and in measles vaccination from 0.054 (95% CI: 0.046-0.063) to 0.019 (95% CI: 0.009-0.029). Likewise, the RII improved for these vaccines ( Table 3). The equity for polio vaccination coverage did not improve over the study period. However, the SII and RII were already low in 2001: 0.017 (95% CI: 0.012-0.022) and 1.02, respectively, and were similar in 2014 (SII: 0.017; 95% CI: 0.008-0.027 and RII: 1.02). The decrease in polio vaccination coverage during the study period did not seem to affect equity, since the decrease was similar in all wealth quintiles.
For all wealth quintiles, the proportion of children who had received all doses of immunization at the age of 12 months increased during the study period. However, the improvement in coverage was much greater in the poorest quintiles (Fig. 1). For the poorest quintile, coverages of BCG, DPT and measles immunizations improved over time (  Table 5). The analysis also revealed a similar trend for mother's education level. Children of non-educated mothers were four times more likely not to be fully immunized in 2001 (aOR: 4.13; 95% CI: 3.52-4.85), which was reduced to 1.5-fold in 2014 (aOR: 1.53; 95% CI: 1.20-1.96; Table 5). BCG: bacille Calmette-Guérin; CI: confidence interval; DPT; diphtheria-pertussis-tetanus; RII: relative index of inequality; SII: slope index of inequality. a Children who had received BCG vaccine, three doses of diphtheria-tetanus-pertussis vaccine, three doses of polio vaccine and measles vaccines were categorized as fully immunized. Notes: SII is the absolute difference of immunization coverage between the richest and the poorest wealth quintile. 22 RII is the ratio of the immunization coverage estimate between wealth quintiles, calculated on the basis of the systematic association between coverage estimate and the socioeconomic status for all wealth quintiles.

Research
Equity distribution of immunization coverage in Nepal Ashish KC et al.

Discussion
This study shows that immunization coverage of Nepalese children aged 12 months or younger has improved significantly between 2001 and 2014; an increase that has been accompanied by improved equity. The improved coverage can be attributed to the concentrated efforts of the Nepalese government in collaboration with nongovernmental organizations, which have focused on hard-to-reach and disadvantaged populations with low immunization coverage. Improvements have been the greatest in the poorer and less educated populations, resulting in a decreased equity gap. This achievement is important since the National Immunization Programme strives to achieve herd immunity within a population. 16 To accomplish such immunity, wide population coverage is needed and failure to reach certain populations might leave them susceptible to vaccine-preventable disease outbreaks.
The results show a setback in polio vaccination coverage. The coverage of the first dose of oral polio vaccine decreased over the study period, from a high level in 2001, to significantly lower between 2001 and 2014. This decrease might be because people may not perceive polio as such a great threat anymore and therefore think immunization is no longer necessary. In 1996, the government initiated a polio eradication programme, which conducted national immunization days targeting all children younger than 5 years with two doses of oral polio vaccine. In subsequent years, Nepal implemented all four strategies required for polio eradication: strengthening routine immunization, supplementary immunization activities, surveillance and outbreak response. The last person infected with indigenous wild polio virus was detected in 1999 and Nepal was declared to have eradicated polio in 2000. 15 Between 2000 and 2010, 30 individuals in the country have been detected with imported wild polio virus and in the past 5 years, Nepal has been free of polio.
Previous studies conducted in the WHO South-East Asian Region, have examined the relationship between various equity parameters and immunization coverage. In line with our results, a study from Cambodia showed that socioeconomic factors were most strongly associated with decreased access to immuniza- Equity distribution of immunization coverage in Nepal Ashish KC et al.
tion services, compared to other factors such as gender or rural/urban location. 23 Furthermore, social distance, that is the distance a person is from health care in terms of acceptability and affordability, was discussed as the main determinant of immunization access as compared to geographical distance. 23 Our study confirms that geographical location, or the urban/rural divide, does not contribute to differences in immunization coverage. Another study from Pakistan found similar inequities when examining the uptake of measles vaccination. 24 Children who lived further than 5 km away from vaccination facilities, had poor roofing, parents with a poorly paid job or mothers who were uneducated, were less likely to be vaccinated in both urban and rural areas. 24 Finally, a systematic review of studies done in India found an association between immunization coverage and the sex of the child, birth order, residence, household wealth, parental education and access to healthcare services. 25 In our study, the sex of the child did not significantly influence coverage. Girls were only slightly less likely to be fully immunized at the age of 12 months compared to boys, a difference that did not show significance at any time point after adjustment for other socioeconomic determinants. This study has several limitations. First, we used data from two different surveys with different survey methods, which could limit the comparison between time periods. For example, DHS and MICS define and calculate wealth index differently. Second, the absence of ethnicity data in MICS meant that we could not study ethnicity as a social determinant of inequity for immunization coverage. Third, the MICS sample size was smaller than in three previous DHS. Fourth, to calculate SII and RII, the methods require a somewhat even distribution of inequity in the groups compared, otherwise differences in the poorest or richest quintiles compared to the overall population can be hidden and the results misleading. 22 For our data, there was a large inequity between poorest and richest quintiles, which masked the magnitude of those inequities when calculating the SII. However, using a logistic regression model, we were able to detect these in-  Notes: Fully immunized children is a composite variable for children who had received all the scheduled vaccines (bacille Calmette-Guérin, three doses of polio at 1, 2 and 3 months of age, three doses of pentavalent at 1, 2 and 3 months of age and measles at 9 months). The first wealth quintile represents the poorest and the fifth quintile the wealthiest.

Research
Equity distribution of immunization coverage in Nepal Ashish KC et al.

Résumé
Une meilleure couverture vaccinale pour combler le fossé en matière d'équité au Népal This study from Nepal demonstrates how directed interventions for the most disadvantaged people can be an important strategy to reach universal immunization coverage and also to help reach Target 3.8 of the sustainable development goals. 26 Part of this target states that all people should have the right to vaccination, which is also one of the aims of the Human Right to Health Movement. 26,27 Even though the increase in overall immunization coverage addressed the equity gap, inequities still exist in Nepal. However, the trend in the past 15 years is promising and indicates the possibility that Nepal will be able to achieve universal immunization coverage. ■

Funding:
The United Nations Children's Fund, Nepal, funded the study.

Equity distribution of immunization coverage in Nepal
Ashish KC et al.