**4. Discussion**

Our mapping confirmed that most countries have already been proposing pro-equity interventions using Gavi HSS funds for several years and is consistent with the results from UNICEF's mapping of JA reports [15]. It is the first time we are aware that such an analysis of pro-equity interventions in HSS proposals from all Gavi-supported countries was conducted. Importantly, we found that outreach and tailoring location of service delivery was very commonly presented in those proposals. This aligns with the observation that one of the main bottlenecks to getting children immunised reported in the proposals was the long distances to health facilities. This bottleneck is also well documented in the literature as a significant barrier to access to immunisation services [22]. Countries mostly addressed this bottleneck by conducting outreach and mobile sessions in hard-to-reach areas or areas with no health facility nearby. However, sustainability concerns and costeffectiveness of conducting outreach and mobile sessions as compared to other long-term strategies were rarely discussed in the HSS proposals. One explanation for this might be the fact that outreach and mobile sessions have been used for a long time and are generally considered necessary to increase immunisation coverage in those contexts.

District microplans and RED strategies were also very common and were assembled into one category because they were frequently planned to be implemented concurrently and since microplanning is often the only component of RED that is implemented. The RED approach was developed by the World Health Organization (WHO), UNICEF and other partners in the Gavi Alliance to improve immunisation coverage and it includes five operational components aimed at improving vaccination coverage: re-establishment of regular outreach services, supportive supervision, on-site training, community links with service delivery, monitoring and use of data for action, and better planning and management of human and financial resources [20]. These components thus have a large span and since they were grouped together under "RED strategies" in the proposals, we could not have a clear vision on what was planned exactly in each country. It was interesting to see, however, how widespread this overall approach has become as a strategy to improve immunisation coverage and increase equity.

Interestingly, two categories were reported in the UNICEF pro-equity mapping of JA reports that were, however, not found in the HSS proposals. These were "peer support group for health providers" and "security to allow immunisation services to happen safely".

Additionally, the results showed that several proposals focused on reaching remoterural and hard-to-reach areas, but few prioritised the other ERG settings, namely urban poor and conflict-affected areas. To note that only proposals that explicitly stated they would prioritise those areas and developed key interventions to improve coverage there were included in the results. It is possible that other planned pro-equity interventions would address barriers in those areas, but were not clearly acknowledged as doing so. These results were not surprising, however, as we analysed HSS proposals submitted starting from 2014, while the ERG priority areas were only defined in 2018. Furthermore, recent evidence shows that unlike what we might have expected, less than 50% of ZD children live in ERG settings worldwide, suggesting that although they are key areas for prioritisation, it is unlikely that we will make considerable progress by solely targeting those settings [23]. Still, many proposals did mention having large pockets of unimmunised children in large urban areas and slums. It is estimated that 28% of un- or under-vaccinated children lived in urban and peri-urban areas and up to 15% lived in conflict-affected areas in 2020 [24]. It would thus be beneficial to pay special attention to those populations. In this sense, it is worth noting renewed efforts to reach conflict-affected populations with the launch of the Zero-Dose Immunisation Programme (ZIP) in June 2022. This initiative led by Gavi in partnership with the International Rescue Committee and World Vision aims

to identify and reach ZD children in the Horn of Africa and the Sahel regions, prioritising children living in conflict settings, mobile populations, and cross-border refugees [25].

Interestingly, the results of the correlation analysis suggest that "bundles" of interventions are commonly used at the country-level as part of the strategy to increase immunisation coverage in the priority areas. Indeed, we found that outreach sessions and tailoring location of immunisation services was often implemented along with developing district microplans/RED strategies, community-level education activities as well as communication strategies to generate demand. Conversely, the analysis revealed how often certain interventions were not bundled, potentially limiting their sustainability and effectiveness. This links to increasing evidence that there are no silver bullets but rather bundles, or packages, of evidence-based interventions tailored to local context that are needed to increase immunisation coverage [26,27]. Learning efforts exploring these bundles of interventions to better reach ZD children and how to use the interventions synergistically to build off one another would be worth exploring. Developing a theory of change, among other things, would be a useful exercise to justify the bundling of activities and validate their effectiveness through implementation research or other approaches [28].

The finding that the rationale, or theory, behind the selection of specific pro-equity interventions was seldom provided in the proposals does not suggest that there is no rationale, but only that it was not clearly formulated. This made it difficult to assess the relevance and intended effects of interventions in different contexts. The few instances when a rationale was presented were generally for interventions that were not commonly found in other countries' HSS proposals. One might reasonably assume that there was less established evidence supporting the implementation of these rarer interventions, thus the need to justify them in the proposals. Documenting the assumptions and reasoning for specific interventions, especially less common ones, would be beneficial to monitor and measure their effectiveness. For example, by developing a theory of change or a strong logical model based on evidence of good results from other similar programs.

Furthermore, in the case of bundles, theories of change would be helpful to articulate how different interventions are expected to work synergistically to produce change. It would also help with the monitoring and evaluation of programmes. Considering this, it would be highly useful to build an evidence base of interventions, namely through implementation research, that may be used in those bundles. This has been conducted, for example, in the field of family planning, where over sixty organisations have endorsed and participated in the development, dissemination and implementation of a repository of evidence-based interventions coined "High Impact Practices", or HIPs [29]. The group explores practices that have demonstrated impact and generate evidence around replicability, scalability, sustainability, and cost-effectiveness of the different interventions and disseminate information namely through evidence briefs. Building similar evidence on interventions aiming to reach ZD children would be extremely valuable and would help prioritisation and strategic planning for future investments. This work could also be used for advocacy, design and implementation of programs, development of policies and guidelines, and identify knowledge gaps for future research.

Furthermore, several of the interventions listed in the proposals were not considered pro-equity according to our definition (and were thus not included in the database nor analysis), but they could easily become so if they were targeted or tailored to specific populations. For example, social mobilisation activities to generate demand aimed at the entire population of a country through mass media could become pro-equity by adapting the messaging to specific target communities. Along the same lines, capacity building of health workers could become a pro-equity intervention if the health workers received adapted training on interpersonal relations with specific vulnerable groups such as refugees, for example, or if they served a low-performing area. In short, countries do not necessarily have to go back to the drawing board to design 'pro-equity' interventions but should build on existing interventions and tailor and target them to areas and/or populations with large numbers of ZD children. Accurately identifying who and where zero-dose children

is evidently a critical pre-requisite to be able to do this effectively. This is not to say that innovative interventions are not needed to reach ZD children, but both strategies can be used coincidently. It is also not to say that simply targeting and tailoring an intervention to a subgroup will necessarily be effective. Understanding the context, including the different vulnerabilities and barriers faced by a particular community, as well as building humancentred designs will be critical to appropriately reach the remaining unimmunised children.

A crucial point at the centre of the ongoing work around zero-dose children is the lack of an agreed upon definition of what constitutes a "pro-equity" approach. Vega & Irwin first highlighted in 2004 that pro-equity health policy should not only consider socioeconomic status, but all other social and systemic factors that influence health [30]. Wagner more specifically referred to pro-equity approaches as promoting equity for women and girls, special education needs and "marginalized" populations [31]. In the current article, we defined pro-equity interventions as "tailored or targeted approaches towards un- or under-immunised children and missed communities". Dadari and colleagues, for their part, defined them as "strategies designed to reach underserved children and populations" [15]. However, there is no formal definition and none of the current ones explicitly address the intersectional nature of inequities. Intersectionality, a concept first coined in African American feminist literature, describes the ways in which different inequalities are linked together and are mutually reinforcing in perpetuating discrimination and disadvantage [32]. Promoting health equity has been a priority for a long time now and much effort has gone towards it. However, the fact that inequities remain today may in part be explained by the fact that even though research shows the importance of many social determinants on health, we often take a siloed view on how to address them. Policy and action have mostly failed to recognise their intersectional nature and have instead focused extensively on addressing inequities related to socioeconomic status or on specific programs without addressing social determinants, which may generate short-term results but may not promote sustainability [33]. Stakeholders should thus reflect on what can be completed differently, such as building packages of interventions addressing different, overlapping vulnerabilities for example, that might help us bridge the gap to promote equity and reach ZD children and missed communities. Having a clear and common definition of what constitutes a pro-equity intervention would be important to avoid working in silos and to help test the effectiveness of pro-equity approaches in reaching zero-dose children.

The limitations of this study must be acknowledged. First, the HSS proposals provided an incomplete picture of pro-equity interventions being implemented at the country-level. They were limited in scope and reflected what countries planned to do with Gavi HSS funds, but Gavi is not the only source of funds for programmes. This might have led to a loss of perspective of other sectors in the findings and analysis. Secondly, the documents analysed did not report on implemented activities, but rather on plans susceptible to change in the countries' dynamic contexts, and thus did not necessarily reflect what truly happened in the field. Furthermore, considering the length of each proposal, a search by keyword was performed. Even though we conducted a manual screening of sections likely to contain relevant information, it is likely that some information was missed. Finally, a number of subjective assessments had to be made during data extraction to decide the category of each intervention. However, inter-rater reliability was not assessed since there was only one analyst and steps were taken to maintain objectivity and avoid bias via building on the existing codes developed by UNICEF for the JA mapping and reviewing the findings with peers.

#### **5. Conclusions**

The findings from this mapping provide a portfolio analysis of HSS pro-equity programming in all Gavi-supported countries and can inform discussions on what may or may not need to change to better reach ZD children and missed communities in the future. Further mapping should be conducted to provide a more complete picture of pro-equity strategies being implemented in those countries beyond interventions funded by Gavi

through HSS grants. The results can also help identify specific interventions that require further attention for further evidence synthesis, case studies and implementation research to learn more about their effectiveness, feasibility, acceptability, implementation cost and sustainability, among other factors. In addition to exploring new interventions, research should be conducted to investigate how to better design and implement commonly used interventions such as the ones identified in this mapping (e.g., outreach sessions, tailoring the location of service delivery, microplanning and community-level education activities) and adjust them to better reach the ZD children that are the key priority of Gavi 5.0 and IA2030.

**Author Contributions:** Conceptualization, J.D., G.C.C. and H.W.R.; data extraction, J.D.; methodology, J.D., G.C.C., H.W.R., A.B.S., V.A.F. and M.J.; analysis, J.D. and G.C.C.; writing—original draft preparation, J.D.; writing—review and editing, J.D., G.C.C., H.W.R., A.B.S., V.A.F. and M.J. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** All data supporting reported results in this article are from HSS proposals publicly available on Gavi, the Vaccine Alliance's website gavi.org (accessed on 20 December 2022).

**Acknowledgments:** We would like to acknowledge the contribution and support provided by colleagues at Gavi including Hope Johnson, Esther Saville, Dan Hogan, Binay Kumar, Maria Patyna, Hamidreza Setayesh, Nilgun Aydogan and Patience Musanhu. We would also like to extend our gratitude for the feedback provided by Robert Scherpbier, Niklas Danielsson and Asm Shahabuddin at UNICEF.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **Appendix A**

**Table A1.** List of all pro-equity intervention categories found in HSS proposals relating to reaching ZD/under-immunised children and corresponding definitions.



#### **Table A1.** *Cont.*


#### **Table A1.** *Cont.*

**Figure A1.** Complete correlation matrix (including all intervention categories).
