{"paper_id":"fac112e6-be2d-47bc-ad81-1ed190cbfdcc","body_text":"Understanding the political economy of \nreforming global health initiatives – \ninsights from global and country levels\nSophie Witter1*, Natasha Palmer1, Rosemary James2, Shehla Zaidi3, Severine Carillon4, Rene English5, \nGiulia Loffreda1, Emilie Venables2, Shifa Salman Habib3, Jeff Tan3, Fatouma Hane4, Maria Paola \nBertone1, Seyed-Moeen Hosseinalipour2, Valery Ridde4,6, Asad Shoaib3, Adama Faye4, Lilian Dudley5, \nKaren Daniels5, Karl Blanchet2\n1- Queen Margaret University, Edinburgh, Scotland\n2- Geneva Centre of Humanitarian Studies, Faculty of Medicine, University of Geneva, \nSwitzerland\n3-Aga Khan University International (in the UK) with Aga Khan University Pakistan\n4- Cheikh Anta Diop University, Dakar, Senegal\n5- Division of Health Systems and Public Health, Department of Global Health, Stellenbosch \nUniversity, South Africa\n6- Institut de Recherche pour le Développement, France\n* corresponding author. Email: switter@qmu.ac.uk\nShort title: Understanding the political economy of the global health initiatives\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \nNOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice.\n\n2\nAbstract\nIntroduction\nSince 2000, the number and role of global health initiatives has been growing, with these platforms \nplaying an increasingly important role in pooling and disbursing funds dedicated to specific global \nhealth priorities. While recognising their important contribution, there has also been a growth in \nconcerns about distortions and inefficiencies linked to the GHIs and attempts to improve their \nalignment with country health systems. There is a growing momentum to adjust GHIs to the current \nbroader range of global health threats, such as non-communicable diseases, humanitarian crises and \nclimate change. However, these reform attempts hit up against political economy realities of the \ncurrent structures.\nMethods\nIn this article, we draw on research conducted as part of the Future of Global Health Initiatives process. \nThe study adopted a cross-sectional, mixed-methods approach, drawing from a range of data \nsources and data collection methods, including a global and regional level analysis as well as three \nembedded country case studies in Pakistan, South Africa and Senegal. All data was collected from \nFebruary to July 2023. 271 documents were analysed in the course of the study, along with data \nfrom 335 key informants and meeting participants in 66 countries and across a range of \nconstituencies. For this paper, data were analysed using a political economy framework which \nfocused on actors, context (especially governance and financing) and framing.\nFindings\nIn relation to actors, the GHIs themselves have become increasingly complex (internally and in their \ninterrelations with other global health actors and one another). They have a large range of clients \n(including at national level and amongst multilateral agencies) which function as collaborators as well \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n3\nas competitors. Historically there have been few incentives within any of the actors to maximise \ncollaboration given the competitive funding landscape. Power to exert pressure for reforms sits \nultimately with bilateral and private funders, though single-issue northern NGOs are also cited as \nimportant influencers. Funders have not collaborated to enable reforms, despite concerns amongst a \nnumber of them, because of the helpful functional role of GHIs, which serves funder interests. Some \nkey global boards are reported to be engineered for stasis, and there are widespread concerns about \nlack of transparency and over-claiming (by some GHIs) of their results. Narratives about achievements \nand challenges are important to enable or block reforms and are vigorously contested, with \nstakeholders often selecting different outcomes to emphasise in justifying positions.\nConclusion\nGHIs have played an important role in the global health ecosystem but despite formal accountability \nstructures to include recipient governments, substantive accountability has been focused upwards to \nfunders, with risk management strategies which prioritise tracking resources more than improved \nnational health system performance. Achieving consensus on reforms will be challenging but funding \npressures and new threats are creating a sense of urgency, which may shift positions. Political \neconomy analysis can model and influence these debates.\nKeywords: Global health initiatives; political economy analysis; governance; financing; South Africa; \nPakistan; Senegal\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n4\nIntroduction\nThe global health system has undergone significant expansion over the past few decades, linked \nin part to efforts to reach the Millennium Development Goals (MDGs). This has included a \ncontinued increase in both the number and diversity of actors and the volume of funding. It is of \nnote that there has also been a marked increase in the distribution of development assistance for \nhealth (DAH) through Global Health Initiatives (GHIs), which are international partnerships that \naim to address specific goals in global health.  Many GHIs have been established since the early \n2000s, driven by the creation of the Global Fund to fight AIDS, Tuberculosis and Malaria (GFATM) \nand Gavi (the Vaccine Alliance), which accounted for 14% of DAH by 2019 (1). Four “mega-\ntrends” in DAH of proliferation, verticalization, circumvention of government systems, and \nfragmentation are identified, which go beyond but include the role of the GHIs. In relation to \nhealth financing, it is also important to note that DAH still forms a large part of the health budgets \nfor many low-income countries (LICs) in particular, and yet that the overall amount of financing \nfor health is not adequate to fund the achievement of the Sustainable Development Goals (SDGs). \n(2)\nEmerging challenges such as climate change, humanitarian crises, antimicrobial resistance, and a \nrise in non-communicable diseases over this timescale also suggest a need to find ways of \napproaching global health which are less vertically focussed on infectious diseases. Plateauing \nDAH and shrinking fiscal space post-COVID-19, a stormy geopolitical context, growing health \nneeds and costly health technologies are amongst the additional expected stressors. These factors \nargue for an urgent review to ensure that all global health resources are used as effectively as \npossible. The mismatch of DAH overall to global and country burden of disease suggests scope for \nimprovement.\nThe FGHI process was a time-bound multi-stakeholder exercise to explore how Global Health \nInitiatives (GHIs) contribute to progress towards Universal Health Coverage (UHC) and the \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n5\nbroader SDGs 2030 Agenda, and how this could be strengthened from the perspective of recipient \ncountries. The process, which ran 2022-23, aimed to make recommendations on how GHIs could \nbe more efficient, effective and equitable and to catalyse collective action to ensure that they were \nfit for purpose through 2030 and beyond. It led to the endorsement of the Lusaka Agenda in \nDecember 2023, which outlines five key shifts and a call to action for all GHI stakeholders to \nstrengthen the contribution of GHIs to achieving UHC (3).\nThis article draws from research commissioned as an input into that process (4). It was focused \non six GHIs, which differ in form and function: the GFATM, Gavi, the Global Financing Facility for \nWomen, Children, and Adolescents (GFF), Unitaid, the Foundation for Innovative New Diagnostics \n(FIND), and the Coalition for Epidemic Preparedness Innovations (CEPI) (Table 1), however in \nthis article we focus on the three main GHIs which account for the majority of funding invested in \nlow- and middle-income countries (GFATM, Gavi and the GFF). The study adopted a UHC lens and \nfocused on countries’ experiences with the GHIs as a group and the wider aid ecosystem. \nTable 1. The six Global Health Initiatives selected for the FGHI study\nGlobal health \ninitiative (GHI)\nMain objective Country-level function Approximate size\nCountry-level grants and technical assistance\nThe Global Fund \nto Fight AIDS, \nTuberculosis and \nMalaria (GFATM)\nEst. 2002. \nHeadquartered in \nGeneva\nTo attract leverage and \ninvest additional resources \nto end epidemics of HIV, TB, \nmalaria, reduce health \ninequities and support \nattainment of the SDGs\nGrants and technical \nassistance for disease \nprogrammes and health \nsystem strengthening \nrelating to these \nprogrammes.\n$5.2 billion per \nyear1. (67)\nCountry eligibility is \nbased on income \nclassification and \ndisease burden of \nHIV, TB, and/or \n1  By taking the last replenishment total and dividing by the three-year cycle; not a measure of actual \nexpenditure per year\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n6\nmalaria.\nGavi, the vaccine \nalliance\nEst. 2000. \nHeadquartered in \nGeneva.\nTo save lives and increase \npeople’s health by \nincreasing the equitable and \nsustainable use of vaccines\nGrants and technical \nassistance for vaccination \nprogrammes and health \nsystem strengthening \nrelating to these \nprogrammes.\nUS$21.3 billion in \ndonor contributions \nand pledges from \n2021-2025.(68)\nCountry eligibility \ndepends on Gross \nNational Income per \ncapita.\nModel based on leveraging concessional finance\nGlobal Financing \nFacility (GFF)\nEst. 2015. \nHeadquartered in \nWashington, D.C.\nTo end all preventable \nmaternal, child and \nadolescent deaths by 2030, \nthrough a health systems \nstrengthening approach\nGrants (as seed funding) \nand technical assistance \nrooted in a broad \ninvestment case, rooted \nthrough government \nsystems.\nAs of June 30, 2020, \nthe GFF Trust Fund \nhad US$602 million \nin grants under \nimplementation—\nlinked to US$4.7 \nbillion of World \nBank IDA/IBRD \nfinancing\nAims to mobilize \nmore than US$57 \nbillion from 2015 to \n2030 (69)\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n7\nResearch & development and market shaping\nUnitaid\nEst. 2006. \nHeadquartered in \nGeneva\n1-To accelerate the \nintroduction and adoption \nof key health products \n2-To create systemic \nconditions for sustainable, \nequitable access \n3-To foster inclusive and \ndemand-driven \npartnerships for innovation\nGlobal (late stage) \nresearch & development \n(R&D) and \nimplementation for new \ninnovations, including \ncreation of sustainable \nmarket conditions for \nequitable access.\nPortfolio budget of \nUS$164 million in \n2023.(70)\nRequested US$1.5 \nbillion for the 2023-\n2027 investment \ncase.(71)\nFIND, the global \nalliance for \ndiagnostics\nEst. 2003. \nHeadquartered in \nGeneva.\nTo drive equitable access to \nreliable diagnosis through \ncollective action\nGlobal R&D for new \ndiagnostics\nRequested US$100–\n120 million per year \nfor 2021-2023 (72)\nCoalition for \nEpidemic \nPreparedness \nInnovations \n(CEPI)\nEst. 2016. \nHeadquartered in \nLondon.\nTo accelerate the \ndevelopment of vaccines \nand other biologic \ncountermeasures against \nepidemic and pandemic \nthreats to be accessible to \nall\nGlobal R&D for new \nvaccines and other \nmeasures to prevent \nepidemics and pandemics\nApproximately $200 \nmillion per year. \nOverall target of \nfunds of USD \n$1billion. (73)\nThis article reports on the political economy underlying the current role of GHIs in the global \nhealth system and attempts to reform them.  While critiques of GHIs have been expressed and \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n8\npublished for decades (5–8) and incremental reforms undertaken within organisations, \nreforming fundamental aspects such as mandates, governance, transparency and priorities, and \nhow GHIs and other DAH actors cooperate with one another and engage with national health \nsystems, has been challenging.    \nMaterials and methods\nThe study adopted a cross-sectional, mixed-methods approach, drawing from a range of data \nsources and data collection methods, including a global and regional level analysis as well as three \nembedded country case studies in Pakistan, South Africa and Senegal.  Case study countries were \nselected based on offering a range of national government’s experiences with GHIs, having a \nvariety of GHIs’ investments and having in-country strong academic partners. \nData sources\nThe study was conducted between February and  July 2023 and drew on a number of data sources, \nwhich are detailed more fully in  (4) : 1) a rapid scoping review of available peer-reviewed and \ngrey literature (271 documents in total), 2) global and country level burden of disease and health \nfinancing data, 3) global-level key informants (KIs) interviews, 4) three in-depth country case \nstudies, 5) regional consultations with key stakeholders in all six World Health Organization \n(WHO) regions, 6) an online survey targeted to KIs who could not join the interviews or \nconsultations and Board members of the GHIs, and 7) consultative meetings, including one co-\nhosted by the Africa Centre for Disease Control and Prevention (CDC) in Addis Ababa in June 2023 \nto discuss preliminary findings. The study participants (total of 335) were based in 66 countries \n(Table 2). \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n9\nTable 2 Number and category of study participants by data source\nData Stream Number of \nparticipants\nCategory of participants\nGlobal-level interviews 76 GHI (n=18), Academic (n=11), Multilateral (n=16), Bilateral \ndonor (n=15), CSO (n=10), Private Sector (n=4), Foundation \n(n=2)\nCountry-level \ninterviews (Pakistan, \nSenegal, South Africa)\n63 Government (n=22), CSO (n=10), Academic (n=10), \nImplementation partner (n=4), Technical/Financial partner \n(n=6), National and provincial disease programme (n=4), \nTechnical Assistance provider (n=1), Multilateral (n=3), \nRegional organisation (n=2), Private Sector (n=1)\nRegional consultations \n(all six WHO regions)\n77 Multilateral (n=23), CSO (n=23), Implementing government \n(n=17), Academic (n=11), Implementation partner (n=3)\nProduct Development \nPartnership Coalition \nConsultation\n6 Product development partnership member (n=6)\nTargeted online survey 46 Academic (n=15), CSO (n=11), GHI (n=6), Implementing \ngovernment (n=4), Bilateral donor (n=4), Multilateral (n=4), \nFoundation (n=1), Other (n=2)\nHybrid Deliberative \nDiscussion co-hosted by \nAfrica CDC\n45\n(30 in-person, 15 \nonline)\nIn-person: Government (n=9), FGHI (n=4), CSO (n=4), \nMultilateral, (n=3), Regional organization (n=3), Africa CDC \n(n=3), Bilateral donor (n=2), Foundation (n=2)\n \nOnline: CSO (n=2), Product development partnership (n=1), \nGovernment (n=2), Foundation (n=5), Bilateral donor (n=2), \nIndependent global health consultant from the African \ncontinent (n=1), Multilateral (n=1), Academic (n=1)\nFGHI Steering Group \nConsultative Meeting\n22 Multilateral (n=2), Recipient government (n=3), CSO (n=2), \nBilateral donor (n=8), Foundation (n=5), FGHI (n=2)\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n10\nData Stream Number of \nparticipants\nCategory of participants\nTotal number of \nstudy participants*\n335 CSO (n=62, 19%)\nGovernment (n=57, 17%)\nMultilateral (n=52, 16%)\nAcademic (n=48, 14%)\nBilateral donor (n=31, 9%)\nGHI (n=24, 7%)\nFoundation (n=15, 4%)\nPDP (n=7, 2%)\nFGHI (n = 6, 2%)\nPrivate Sector (n=5, 1%)\nOther (n=29, 8%)\n*some participants may have been counted twice (e.g. if they participated in both an interview and a \nconsultation)\nStudy participants were purposely selected based on their level of experience working with GHIs \nand their membership of relevant constituencies (GHIs, academia, multilateral or bilateral \ndonors, civil society organizations (CSOs), private sector and philanthropic foundations). A first \nlist of informants was drafted by the FGHI Secretariat and then completed by the professional \nnetwork of the research consortium. During the course of the study, new KIs were recruited based \non suggestions from people interviewed (snowball technique).\nData analysis\nAll data sources were synthesised to inform this paper. The qualitative data were recorded, \ntranscribed, and coded inductively and deductively by a team of three researchers trained in \nqualitative research. The researcher consortium convened frequently to discuss the emerging \nfindings, and during analysis examined similarities and differences among GHIs and across \nparticipant categories. Political economy analysis (PEA)(9–12) was used throughout the study to \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n11\ninform the analysis and synthesis. Such an approach allowed the team to reflect on the dynamic \ninteraction between actors, their relative power and respective interests and incentives, and \nelements of the broader context, and how the outcome of the interaction affects the likelihood \nand content of future changes. In particular, the study focused on analysing actors, context and \nframing related to the GHIs and the wider global health ecosystem (Figure 1). \nFigure 1 Political economy framework guiding the qualitative analysis and synthesis\nSource: Adapted from (11)\n Actors: a detailed analysis of the stakeholders was carried out in each country and at global \nlevel. We identified as domains for the stakeholder analysis: (i) interest and position in \nrelation to changes whether the stakeholder supports, opposes or is neutral about changes to \nstatus quo on GHIs and its motivations for this; and (ii) power and influence (i.e., the potential \nability of the stakeholder to affect implementation of changes to status quo). The stakeholder \nanalysis was informed by guidance (13–15). \n Context: we collected and analysed information concerning the broader context in which the \nstakeholders operate and how it can constrain or support change, focusing on governance \nstructures and financial elements, which emerged from analysis as most relevant. \n Framing: building on recent literature (2) which acknowledges the critical influence of frames \nand framing in policy processes, we explored (but in less depth) the role and power of \nnarratives and discourses, and how they shape the debate around GHIs. \nEthical considerations\nThe study was approved by the ethics review boards of University of Geneva, Cheikh Anta Diop \nUniversity, Stellenbosch University, and Aga Khan University, Pakistan. Informed consent (oral \nand written, according to the circumstances) was obtained from the study participants to \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n12\nparticipate, and to record the qualitative data, which was pseudonymised to protect the study \nparticipants from being identified.\nStudy limitations\nThe study set out to capture the views of highly expert stakeholders with deep insights into the \nworkings of the GHIs, but also different perspectives on the topic, representing all the key parts \nof the global health system. It is important to note several limitations in this work, largely as a \nresult of a tight timeframe. The data we collected were qualitative and based on interviews, \nconsultations and a rapid non-systematic literature review. It is also important to highlight that \nthis is a contested area, and there were conflicting positions, which we reflect in this article.  \nThe country case studies were not meant to be a representative sample, but rather chosen due to \nstrong research partnerships within the country, as well as to compare a range of contexts in \nwhich the GHIs of focus are active. Findings of one country are not meant to be generalisable to \nother contexts, but to shed light on the dynamics that occur around GHIs and different \nexperiences of country stakeholders. \nResults\nActors\nThere has been a significant increase in the number and diversity of actors within the system \n(16). Whilst 30 years ago, it comprised primarily of bilateral and multilateral arrangements \nbetween nation-states, it is now a varied landscape, which also includes private firms, \nphilanthropies, non-governmental organisations (NGOs) and GHIs (17). The increase in DAH \ndisbursements from 1990-2015 was accompanied by a five-fold increase in the number of actors \ninvolved in global health, with a particularly rapid rate of growth in the number of CSOs between \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n13\n2005-2011 (17). In addition, there has been a marked increase in the distribution of DAH through \nGHIs, driven by the creation of the GFATM and Gavi (1). \nThere have also been changes to the GHI’s funding to partners: recent analysis suggested that \nGFATM’s share of disbursements to governmental organisations has been declining, from 80 \npercent in 2003 to 40 percent of all disbursements in 2021 (18). Many of the CSOs funded are \nfocussed in specific health areas: separate work has found that over one-third of CSO channels \nare only providing funds for the implementation of programmes in one health area e.g. HIV/AIDS, \nmalaria, child and maternal health or nutrition (17). \nOver recent decades, many GHIs have grown rapidly and become major players in the global \nhealth system. They are active at global, regional and country level. Some of the longest-standing \nGHIs such as GFATM and Gavi have evolved into large and complex organisations with the size of \ntheir secretariats reflecting this institutional growth.  They have inevitably developed their own \ninternal dynamics and priorities. GHIs now raise and channel 14% of DAH (1,19)  and have taken \non a growing range of roles, most recently including COVID-19 responses.   \nKey stakeholder groups involved in this ecosystem include:\n GHIs, which are instrumental in creating and responding to specific agendas by mobilising \nfunding and collective action.  Within the GHIs themselves, it is useful to distinguish several \npotential loci of power and influence. The Boards are the official mechanism of governance, \nbut other parts of the organisations such as the Secretariats or technical teams can also be \nimportant actors. In the case of the GFATM, for example, there are other bodies which act \nindependently, such as the Office of the Inspector General and the Technical Review Panel \nand Technical Evaluation Reference Group, which has since been replaced by the Independent \nEvaluation Panel (IEP) (20);   \n Recipients of GHI funding include health ministries (national or sub-national), United \nNations (UN) agencies, international and local NGOs, CSOs, private sector (e.g. consultancy, \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n14\ndigital start-ups, pharmaceutical), higher education institutions and research institutions. \nMany actors are keen to continue to receive funding from GHIs;   \n Donor agencies (bilateral, multilateral and private foundations), which constitute the main \nfunders of the GHIs;   \n Multilateral agencies (such as WHO, other United Nations (UN) agencies, World Bank) and \nregional development banks, which work in the same field as the GHIs, often have country \npresence, and can act as collaborators or competitors (or hosts, in the case of the World Bank \nfor the GFF).\n Political and interest groups, which exert pressure on donor governments and GHIs (lobby \nand campaigning groups, international NGOs, transnational corporations).\nHistorically there have been few incentives within any of the actors to maximise collaboration \ngiven the competitive funding landscape, but recently interactions between actors are becoming \nincreasingly intricate, with some GHIs as central players (16)  and growing inter-agency \npartnerships even between the GHIs. (21)\nThe types of power and influence wielded depends on the scope of the actor, which is summarised \nin Table 3 with reference to broad categories (acknowledging that there are nuances within each). \nMethods of wielding power are diverse, including funding power, influencing through formal \ngovernance structures like Boards, and normative power from organisations like WHO. The \nfunders of GHIs were identified as the most powerful actors in the global analysis; they are the \nonly actors that hold the ultimate sanction of withdrawing funding from the GHI ecosystem.  The \nBoards were identified as the principal mechanism through which they can wield that power, but \nit was observed that this was not always exercised successfully. Reasons for this include that \nbilateral donors have diverse focal areas and tend to function in accordance with their own \ninterests and values.  This means that donor coordination and alignment can be weak. They are \neach accountable for their tax-payer-funded investments, hence they seek reassurance on \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n15\nfiduciary risks, as well as measurable impact. This also makes them attentive to the views of \ninterest groups within their own countries.  In addition, DAH departments within high income \ncountry (HIC) governments are required to be accountable to the wider foreign and economic \npolicies and objectives of the country, and this creates additional layers of tensions and \ncompromises for a purely health agenda. Some bilateral donors favour disease-specific \ninvestments, while others are more system-oriented. However, they too benefit from the GHIs as \nan efficient (for them) vehicle for aid spending. Some academic and CSO KIs perceived bilateral \ndonors as prioritizing visible and rapid results to safeguard the health security of their own \ncitizens, such as addressing infectious diseases and preventing their cross-border spread. \nPhilanthropic foundations (which also fund GHIs) may have other interests, including using the \nGHIs as vehicles for projection of influence. \nTable 3 Summary of interest and influence of major stakeholder groups \nActors Interest and position Power and influence\nGHIs Interest in maintaining existence, which \nrequires demonstration of results and being \nadaptable, expanding mandate where new \nneeds are demonstrated. \nEach GHI has its own incentives, which in \nfunding GHIs are focused on fund flows and \naccountability for these.\nPower formally sits with Boards, \nmade up of diverse constituencies. \nHowever, not all constituencies are \nequally empowered or coordinated, \nleaving considerable influence in \nhands of senior leadership of GHIs. \nSix-monthly meetings of a few hours \ncannot provide sufficient oversight \nso other modes of control slip in.\nBilateral \nfunders\nGHIs provide a useful platform for joint action \nfor bilaterals, which are their major funders. \nEach bilateral has to reflect domestic priorities \nbut some (broadly, a European bloc, with \nConsiderable influence over GHIs as \nmajor funders (in proportion to \ncontributions, broadly), however \nthat influence is undermined by lack \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n16\nothers such as Japan and Canada) are more \ncommitted to integrated services and UHC, \nwith higher risk tolerance to achieve more \nsustainable results. Whilst others (such as the \nUS) are more committed to domestic political \npriorities, such as HIV, although this may be \nchanging.\nof coordination between them on \nreform agendas.\nMultilateral \norganisations\nMultilateral organisations play multiple roles \nin relation to GHIs, including:\n- technical partners (e.g. through Accelerators, \nproviding thematic coordination, and also \nthrough co-financing of programmes at \ncountry level, for example with the World \nBank)\n- rivals for bilateral and wider funding\n- providing technical guidance to GHIs (e.g. \nWHO disease programmes and health system \nteams)\n- grantees and implementing partners (e.g. \nUNDP)\n- suppliers (e.g. UNICEF as a major purchaser \nof vaccines for Gavi) \nConsequently, their interests are very mixed \nacross the different organisations, as well as \ninternally within each one\nInfluence at global level varies. A \nnumber, such as WHO, have \nnormative power which affects the \nGHIs. Others are important as \npartners and implementers at \ncountry level.  Some KI argued that \nthe weakness of WHO was one of the \nfactors in the large role of the GHIs. \nMany efforts have been made to \ncoordinate this group with the GHIs, \nhowever, their influence is not strong \nenough to override internal \nincentives of GHIs. \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n17\nPrivate \nfoundations\nPrivate foundations have contributed \nimportant sums to the GHIs, especially the \nGates Foundation, which has invested in Gavi \nand the Global Fund in particular and is \nsupportive of them, albeit sometimes as a \n‘critical friend’.\nThe Gates Foundation has significant \ninfluence through funding and board \nmembership on some of the GHIs, \nwhile also supporting coordination \nmechanisms, such as the \nAccelerators.\nRecipient \ngovernment \nagencies \n(national and \nsub-national)\nGovernment agencies have a broad interest in \nreceiving financial, material and technical \nsupport from GHIs but there are diverse \nconstituencies internally, with some \nstakeholders, such as disease programme \ndirectors and those represented on national \nGHI governance bodies, gaining resources and \nprivileges (such as attending international \nmeetings), while others with more integrated \nportfolios, such as planning, can find their jobs \nharder to do. \nRecipient governments exercise \npower through their presence on the \nGHI Boards, as well as in local \ndecision-making on grant \napplications etc. However, there was \nscepticism as to how formal board \nmembership translated into real \ndecision-making power by KIIs, due \nto informational barriers as well as \nthe frequency and structure of \nmeetings. Power in relation to grants \nwas limited by bureaucratic \nrequirements, though some \ncountries have shown agility in \nmaking these work better for them.\nPower dynamics on local governance \nbodies, such as the CCMs, will depend \non the balance of constituencies and \nindividuals (their interests, networks \nand capacities).\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n18\nNon-\ngovernmental \norganisations, \nconsultants \nand \nacademics \n(local and \ninternational)\nNGOs play diverse roles in relation to the GHIs, \nincluding as lobbyists, board members \n(representing civil society), sub-contracted \nconsultants, and implementing partners. These \ncreate different positions. \n- Some NGOs and consulting agencies, \nespecially the large HIV-focused ones, have \nbeen strongly supportive of organisations \nlike the Global Fund and resistant to \nreforms. Several universities in LMICs play \nthe role of service providers and are \npowerful advocates of GHI funds.\n- In the middle are some implementers and \nconsultants, which may have critiques but \nare not able to voice them easily, due to \ntheir financial dependence.\n- At the other end, are highly independent \nand hostile academics and CSOs which \nhave highlighted the many problems \ncreated by the current operating \nmodalities. \nThe major NGOs which can mobilise \npublic pressure on funder \ngovernments and/or are \nrepresented on governing boards are \nreported to have considerable \ninfluence over the GHIs. Others \n(implementing NGOs, consultants \nand academics) have less influence \non major issues, though they are \nengaged in technical consultations on \nmore detailed areas, such as when \norganisational strategies are being \nrevised.\nPrivate sector \nproviders and \nsuppliers\nThe private sector has varied interests as it \nplays multiple roles in relation to the GHIs, \nincluding as supplier of inputs, partners in \nproduct development etc. \nThe private sector is often \nrepresented on GHI boards but does \nnot feel very well engaged, according \nto our (limited number of) \ninterviews.\nSource: summarised by team based on analysis of KIIs\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n19\nWithin the GHIs, senior leadership was seen as highly influential, not least because of the \nchallenges noted for Boards (further discussed in the context section below). Technical power \nalso sits with the GHI Secretariats, and especially the country grant managers (more so than \ntechnical advisory staff), who are in charge of fund disbursement, which is a key performance \nmetric for GHIs, according to KIs. \n“It's the same program managers who developed the same applications or hired the same \nconsultants to write the same applications. There are three-year time horizons, it's short-term. \nShort-term money, short-term thinking and the grant managers…all of the incentives for the grant \nmanagers are to get the money out the door. That's honestly the main key performance indicator: \nGet the money out the door.” (Global KI)\nThe degree of financial dependency is a key variable in the position of national actors. In crisis-\naffected regions such as the Sahel, struggling with a reduction of domestic funding for health and \nthe withdrawal of the main technical and financial partners, dependence on GHIs has increased \nand their support is highlighted as critical.  (Southern and East Africa regional consultation KI).\nMany of the actor groups, as noted in Table 2, have mixed positions and incentives because of the \ndifferent roles they are playing and resources they may receive from the GHIs. The variation can \nbe between departments within organisations as much as between organisations. Their power or \ninfluence is also varied. At country level, local NGOs were not reported to be influential on GHIs \nin general. South Africa presents a contrasting picture in that the Treatment Action Campaign was \ninfluential in improving access to prevention and treatment options for HIV in particular. (22) \nGlobally, however the single interest lobby groups that campaign on certain health targets were \nviewed as highly influential in mobilising public opinion amongst voters and taxpayers. They can \neffectively bring pressure upon bilateral donors about how DAH budgets are allocated. This is \nreported by KI to be one reason why such a large proportion of the Global Fund’s budget (50%) \nis allocated to HIV.\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n20\n“The epidemiology suggests that there should be more money for TB than HIV, and there's no other \nmoney. It's not like there's another PEPFAR for TB.” (Global KI)\nThe GHIs, by holding a significant portion of global health resources, have had an impact on the \nrole of actors within some countries.  This is particularly true for NGOs and some UN agencies.  At \nthe country level, some UN agencies and large NGOs are reliant on GHIs for “soft-funding” to pay \nkey members of staff on their programmes. For instance, there has been a transformation of the \nUN from primarily a normative agency to a supplier and subcontractor, in many cases heavily \ndependent on GHI funding. The Pakistan case illustrates this phenomenon. Pakistan receives \nextensive funding for polio eradication and much of the effort is invested in eradication \ncampaigns. UN agencies manage the campaigns, deploying a large number of staff and consultants \nsupported by GHI project funding. However, government stakeholders are of the opinion that \ndirect delivery campaigns, even if bringing good results, limit the development of country \nownership and leadership (Pakistan KI). At the same time, some NGOs have also experienced a \nshift from advocating for health issues to assuming supply roles in response to the influence of \nGHIs. \nWHO was often described by KIs at country level as weaker in its partner coordination role than \ndesirable, absent from some of the roles perceived to be important parts of its function, and not \nmanaging to support UHC effectively. There are also potential conflicts of interests and \ninefficiencies as WHO applies to GHIs for funding from some country budgets, and also assumes \nthe role of a supplier of both technical assistance and services in the presence of a weak \ngovernment system. In all of these scenarios there is a risk that government systems are \neffectively bypassed and are not strengthened, with funding flows tilted more towards UN \nagencies and NGOs.  Another key actor in several countries is The World Bank, in some cases \nproviding its finance and convening power to bring bilateral funders and GHIs together for \ninvestment on specific country priorities. \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n21\nFinally, more peripheral actors include the academic community, which is minimally involved in \nthe implementation of GHI grants, though some evaluate their impact. They were amongst the \nmost critical, highlighting problems with the whole current model of external aid and conflicts of \ninterest in the aid landscape. This is also reflected in the literature which questions the role of \n“philanthrocapitalism”(23,24), use of for-profit consulting firms (25), and the pharmaceutical \nsector’s influence on GHIs.  \nA particular facet of the current complex global health funding environment around which there \nwas considerable tension is the use of short-term consultants, particularly at country level where \nthis is seen as boosting private interests and incomes over public service development (26) and \nagain bypassing the strengthening of national health systems. Domestically there can be a \nrevolving door of key, knowledgeable and highly skilled individuals between government, NGOs, \nGHIs and independent advisory work. They can also represent an unfortunate brain drain out of \ncentral government roles. \nIn addition, there can be a plethora of technical assistance both from the region and globally, often \nfunded by GHIs or other partners, sometimes with unclear terms of reference, possibly \noverlapping activities and not aligned to country needs. The interests of international consultants \nversus local ones also emerged as a tension in all three country case studies.\n“The Global Fund and other partners are helping Senegal to apply for grants and submit high-quality \napplications. Unicef, for example, recruits a consultant to support the country, notably at CCM level, \nas part of the elaboration of the GCS7. They have procedures, which require specific expertise, \nmaintain the consultancy market and do not necessarily encourage local capacity building” (Senegal \nKI).\nSome country KIs highlighted the way in which the complex systems operated by GHIs privilege \nexperts and the disempowering effects this has on government staff.\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n22\n“The experts are in charge and have taken total control of the organization. In some countries, 20 \nexperts come and write a concept note … No concept note is written without experts.” (SEARO KI). \nHealth staff are another constituency, which often benefits from GHI funds in the form of per \ndiems and salary supplements, which can however have very distorting effects on the health \nworkforce (27–31). In-country health staff who are highly trained and knowledgeable about GHIs \nare sometimes recruited by the GHIs and assume roles as experts responsible for monitoring \ngrant implementation, either in-country or at the GHI headquarters (Senegal KI). In South Africa, \nhealth staff are often recruited from the same geographical areas where GHIs support service \ndelivery, and are paid higher salaries than those working within the public sector, leading to \nweaknesses within the system (South African KI)\nPrivate sector KIs at global and country levels were willing to be more engaged with the GHIs but \ndid not feel very much so at present.\n“Engagement of private sector is important. All initial GHIs gave less importance to the private \nsector. The common notion was that private sector is not permanent and can go away. However, it \nis there to stay. Private sector and government sector are there to complement each other. \nStrengths of the private sector can better used to find an out of the box solution” (Pakistan KI) \nContext \nGovernance \nThe Boards of some of the GHIs were seen as innovative when first set up, with representatives \nfrom a range of constituencies, including implementing countries, donor countries, CSOs and the \nprivate sector.  The GFATM's Board has equal voting seats for donors and implementers, with 10 \nconstituencies respectively. Within the 10 voting implementer constituencies, seven are \nimplementer governments. Gavi also has representation from the vaccine industry and research \nand technical health institutes.  Instead of a traditional board, the GFF established an Investors \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n23\nGroup (32), which includes a range of actors, including UN agencies, recipient and donor \ngovernments, CSO, private sector, and youth representatives, and a Trust Fund Committee.\nWhile the Boards of the GHIs are designed to monitor and ensure performance, there were \nvarying perspectives on where the authority to challenge and rectify issues actually resided and \nhow it was effectively exercised. Despite being theoretically representative, several KIs indicated \nthat the Boards of some bigger GHIs have been structured in a way that fosters a balance of \nconstituencies, resulting in rather slow and inefficient decision-making.  Furthermore, KIs \nhighlighted that the boards of GHIs can be very large and unwieldy, and this can also make \nconsensus for change harder to reach. In addition, Boards can be at a disadvantage as Board \nmembers typically have short tenures, and this maintains an asymmetry in organisational \nknowledge and skills between the Boards and Secretariat, which has institutional memory. \nIn addition, KIs noted that there is a mismatch in the profiles of board members from the Global \nSouth and Global North, impacting their ability to effectively contribute and engage in decision-\nmaking processes.  There are two key elements to this that came up in our interviews. The first is \nthat the people sitting on Boards from the Global North are not of equivalent seniority to those \nrepresenting the Global South - the example of government ministers representing the South \nwhilst the North is represented by ‘bureaucrats’ from donor agencies was given.  Second, the \nnature of the interaction appears to be unequal, with several KIs stating that it was not possible \nto “speak out” in Board meetings. Concerns were raised regarding the effectiveness of Board \nprocesses in facilitating active and open debates, especially for country representatives. It was \nobserved that specific influential bilateral organisations, as well as certain large NGOs, hold more \npower than the recipient countries themselves.  At county level, NGOs represented on boards may \nsometimes represent their own interests, more than those of the recipient communities (South \nAfrican KI).\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n24\n“On paper [GHI Boards are] diverse but I don't think that the practical spaces that they provide \nactually allow people to speak in the way that they need to speak. It's all muted and it all becomes \npolitics and corridor speak. This is why I don’t go to [GHI] meetings anymore.” (Global KI)\nThese “corridors” are shared by GHIs and bi/multilaterals in Geneva and Washington DC, but \nnot with the Southern representatives, so it is more difficult for them to informally influence \ndecision making. In addition, the lines of accountability are reported to be skewed towards \nfunders, more than country health systems.\n‘The accountabilities are to the capital donors and to getting the money out of the door. And there's \nnot enough accountability to real results in country or to efficiency-oriented concerns.’ (Global KI) \nThe boards were also seen as not having the right technical expertise to address the challenges \nthat the GHIs and the global health system now need to face, in particular those of strengthening \nhealth systems and achieving UHC. \n“When you talk to [GFATM] about the importance of working with others to strengthen health \nsystems in a way that's not specific to HIV, you tend to get pretty blank looks… That's not what \nthey're there for… They're there to finish the job on HIV, and maybe TB and malaria.”  (Global KI) \nAnother aspect of unclear accountability at the global level was raised by some KIs in relation to \nthe lack of transparency of reporting by some GHIs on their activities and investments as well as \nindependent evaluations of their effect and cost efficiency. \nConsequently, this fragmented funding landscape leads to the proliferation of plans, funds, \nreporting mechanisms, and auditing processes. Such fragmentation not only contributes to \ninefficiency but also proves to be ineffective, overwhelming the capacity of the recipient country \nto effectively manage these resources.\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n25\n“You know there's multiple reporting channels, there's multiple. And so it's a complete nightmare \n(South African KI)\n“Gavi has its immunisation financing, technical support and then polio has its polio transition. And \nGFF has its UHC alignment. And we're just all pulling the same people to the same meetings. And \nthe organisations themselves aren't accountable for the fact we just distract and are selling our \nown products and justifying our own existence through these processes.” (Global KI)\nGovernance challenges were highlighted in the case studies - for example, in Senegal, where the \npresence of multiple governance structures across GHIs generates high transaction costs and \nrisks of uncoordinated initiatives for the government (120) (see also Boxes 1-3). Each GHI has its \nown operating methods, procedures, contracts and coordinating bodies. \nIn the case of the GFATM’s Country Coordinating Mechanism (CCM), some concerns regarding its \ncurrent make-up and operations were also raised, as it is typically representative of specific \ninterest groups who may also be funding recipients, aligned to the three diseases, while they may \nlack the technical expertise needed to develop strong health system strengthening (HSS) \nproposals. Other concerns relate to the possible blurring of roles and responsibilities, and \npotential conflicts of interest. For example, in South Africa, the Department of Health is both a \nmember of the CCM and a principal recipient. Furthermore, the South African National AIDS \nCouncil (SANAC) runs the CCM, which is positively viewed by some as indicating local leadership.  \nSANAC is however also a recipient of GFATM money and implements programmes within health \nfacilities. The Secretariat for SANAC is also the Secretariat of the GFATM. There is however strong \nCSO representation and SANAC is co-chaired by the country’s deputy President (33). \nNew institutional interests can also be set up as a result of siloed planning and funding:\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n26\n“The Global Fund model and the Gavi models are interesting. They say they are not going to \nestablish their own in-country presence, but what they've done is create their own in-country \ninstitutional monsters in some respects. We have the ministries of AIDS, right?” (Global KI)\nAt country level, accountability to GHIs (focused on managing financial risks) can take precedence \nover accountability to communities and national entities (for performance).\n\"Within the countries we lose a lot of efficiency because the country teams have to set up no objection \nprocedures, the fiduciary agencies have to validate the implementation, we lose efficiency. \nImplementers spend more time looking for ways to comply with FM [financial management] \ndirectives... regard is more focused on satisfying Geneva than communities” (SEARO KI)\nOther concerns included that reports are sent to ‘Geneva’ or to GHIs’ funders or stakeholders, but \nnot necessarily to the local policy-makers responsible for delivering health services (Addis \nconsultative meeting KI). Multiple KIs urged better country engagement and transparency \nregarding funding to enable collaborative action plans.\n“From a country perspective, I would give them 4/10 for improving health outcomes; 2/10 for \nimproving the health system capacity, 1/10 for graduating from dependence on international \nfinance, and 0/10 for ownership by the government and supporting their policies.” (Global KI)\nFinancing\nIn a context of plateauing DAH (34), the overall environment is marked by competition between \nGHI actors for funds, which drives expanding mandates to ensure continued relevance, for \nexample in the face of new threats such as COVID-19 – counterbalanced by long-standing \ninitiatives to improve alignment between GHIs (Figure 2). \nFigure 2 Creation of GHIs and some alignment initiatives, 2000-2023\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n27\nSource: Witter et al. 2023 (4) . Image credit: Claudia Molina\nGlobal KIs perceived competition for funding between GHIs and other global-level organisations, \ncreating a sense of a zero-sum game, where funds may also not align with the actual needs in \nterms of disease burden or the functional role of different organisations.  The competition for \nfunding from the same pot of money was perceived to be likely to contribute to a perceived \neagerness of GHIs to take on new roles and expand their mandate, as organisations jostle for roles \nand funding. The existing system of staggered replenishments by GHIs was perceived as \nchallenging for bilateral donors and governments of LMICs to manage (35–37) and there were \nconcerns regarding the overall financial sustainability of the repeated, increasing GHI requests \nfor replenishment.\nAt country level, dependence on GHI resources can lead to imbalances in relation to priority areas \nand loss of alignment. In Senegal, for example, despite low prevalence, HIV programmes continue \nto receive substantial funding, whereas non-communicable diseases, which are more prevalent, \nlack sufficient resources (KII and (38)). This was echoed in the South African case study, where \ndespite the high HIV prevalence concerns were raised that not enough finances were being \ndirected to non-communicable diseases and strengthening of primary health care.\nAt the country level, some GHIs wield considerable power, depending on their contribution to the \ncountry’s domestic funding.  GFATM and Gavi are important funders to governments, NGOs and \ncivil society. A comparison of WHO’s Global Health Expenditure Database (April 2023 update) \n(39)and OECD Creditor Reporting System (40) data indicates that Gavi and GFATM gross \ndisbursements accounted for a larger combined budget than domestic government funding in \nseven sub-Saharan African countries 2 in 2020, giving these two institutions considerable \ninfluence.  As an interesting contrast, in South Africa DAH constitutes less than 5% of total health \n2 Central African Republic, Democratic Republic of the Congo, Eritrea, Guinea, South Sudan, Uganda, \nZimbabwe\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n28\nexpenditure, with the GFTAM providing the largest share of funding for HIV and to a lesser degree \nTB and malaria.(39) KIs reported that this small contribution to the overall budget does limit \ntheir power at governmental level. As in other countries, GFATM and Gavi also work through a \nvariety of channels   and by empowering non-state actors or disease-specific programmes they \nare still capable of creating advocates for them.  Lack of transparency can also cause challenges \nfor managers at devolved levels:\n“In Ghana, in talking to district managers, they were so frustrated because these donors were \ncoming in, running their funding off budget and basically bypassing them… The district managers \nhave very little power in how these resources are allocated, but they're held accountable for \ndelivering within their districts. It's crazy, right? And there's so much frustration at that level. I \nthink from a governance side they should be very transparent.” (Global KI)\nThere are also imbalances within government, in that funds go disproportionately to some \nprogrammes (such as HIV/AIDS and malaria), which creates inequities and also vested interests \namongst some Ministry departments. For instance, in Mozambique, a KI reported that 80% of the \nfunding received is for HIV, which creates a set of vested interests out of balance with the rest of \nthe health system, and little incentive for these recipients to support a more integrated system.  \nThe ability to gain such disproportionate benefits from GHI funding, including as a result of the \nopaque mapping of funding to public expenditure, creates pockets of strong resistance to \nreforming the GHIs as they are currently functioning at country level.  \nBy contrast, GFF works through more a integrated funding mechanism, which raise different \nconcerns about fungibility.\n“Financing takes the form of budgetary support or trust funds, producing a substitution effect \nbetween donors and governments.  How can we explain the fact that while budget support is \nincreasing when, health expenditure and needs are not being covered?” (SEARO consultation KI)\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n29\nMoreover, provision of funding is perceived as not tied to a country plan led and owned by \nMinisters of Health and instead is tied to programmatic funding cycles of Gavi and GFATM, with \nan imperative to disburse funds rather than support national planning. This results in the \nprovision of fragmented ad hoc funding and exacerbates frustration within country \ngovernments at being powerless to direct funding or ensure accountability:\n“The power lies with GHIs so far. They send you the support but you do not have a say. If you do not \nhave a say, you do not have power” (Pakistan KI; see also Boxes 1-3)\nBox 1. Country Case Study: Pakistan\nDonor financing in Pakistan, inclusive of bilateral agencies, multilaterals and GHIs, has typically \nbeen less than 2 percent of the total health expenditure (74,75). Gavi finances vaccines, cold \nchain, advocacy and community outreach support for immunization-Polio eradication. GFATM \nextends the largest support to TB diagnostics, which includes integration of the private sector. It \nalso makes contributions towards malaria control and HIV prevention through community-\nbased outreach information systems strengthening, and awareness(76). Global Financing \nFacility has recently started contributing to Pakistan and will be contributing to maternal care \nas part of pooled financing with the World Bank (77)\nChallenges\nGovernance, coordination and alignment\n Competing technical assistance plans between donor agencies and GHIs, and between \ngovernment and GHIs, resulting in duplication of assistance and divergent priorities\n Weak country capacity for aid coordination, realistic target setting and planning but \nlittle investment in capacity building \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n30\n Lack of coordination between federal and provincial governments, exacerbated by \nfragmented projectized funding by GHIs, constrains cohesive country planning \n Leadership erosion with frequent leadership changes of health secretaries and disease \nmanagers \nHealth Systems Strengthening and integration\n Several ongoing local health reforms but GHI funding not integrated into reform \nplanning, hence constraining cohesion and sustainability\n Uneven capacity of disease planners and health system managers\n GHI prioritization of disease control programmes is insufficiently backed with local \nhealth systems strengthening support\n Disease control efforts are not framed within the larger ambit of Primary Health Care \n Large private sector but not effectively harnessed for disease control and PHC\nGHI financing\n Funding and disbursement is driven by donor-led burden of disease analysis with less \nconsideration of local health systems realities. \n Ad hoc use of external finance as standalone projects rather than integration into \nongoing initiatives for sustainability and efficiency\n Multiple parallel funding streams by GHI constrains oversight and coordination of \nexternal financing\n World Bank aspirations to leadership under pooled funding but lack of integration of \nlessons learned from past attempts at pooled funding \nMonitoring and performance accountability\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n31\n Overambitious targets set by GHIs, not contextualised to local health systems realities \nand opportunities within existing reforms\n Low political voice of the government to articulate accountability needs as well as weak \nsystems and staffing limits accountability and counter correction measures.  \n Accountability constrained by lack of a central accessible repository of funding / \nprojects data \nBox 2. Country Case Study: South Africa\nGHIs contribute less than 5% towards health financing in South Africa. PEPFAR and GFATM are the \nlargest donors. FIND, Unitaid and CEPI fund non-state actors. Gavi and GFF have no in-country \npresence. South Africa is a donor to Gavi and GFATM.\nChallenges\nGovernance, coordination, and alignment \n Lack of in-country alignment of GHIs’ priorities and activities with country health plans and \npriorities\n Separate in-country GHI coordination and resource mobilisation mechanisms\n GHIs tend to by-pass government structures and directly fund non-state actors\n Civil society not sufficiently active or strong to hold GHIs accountable for in-country activities\nHealth Systems Strengthening and integration\n Prioritized disease control programme by GHIs; lack of support for local health systems \nstrengthening (HSS) (e.g. Universal Health Care) reforms, resulting in fragmentation\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n32\n Duplication of systems (information, health financing, etc) resulting in increased burden on \nhealth managers\n Bypassing of local experts in favour of international ones who do not understand the local \ncontexts\nGHI financing\n Funding in areas of donor interest with little consideration of local health systems realities. \n Funding for implementation not always strategic or sustainable (e.g. use of funds for specific \nline items/activities)\n Bypassing of national government financing system/lack of reporting transparency; therefore, \ngovernment cannot account for all GHI funding\n Donor funding tend to have conditionalities or restrictions attached to them which may be at \nodds with country priorities\nMonitoring and performance accountability\n No formal in-country governance or accountability mechanisms that mandate that GHIs first \nreport findings and challenges to country before reporting to their stakeholders (e.g. Boards)\n Limited evidence of the real effect or impact of GHIs on health outcomes or whole-system \neffects.\n Large GHI datasets and multiple reporting systems undermines the country’s health \ninformation system processes; insufficient coordination, integration/alignment thereof\nBox 3. Country Case Study: Senegal\nAccording to the most recent National Health Accounts (NHA 2017-2021), donors finance almost \nas much as the state (22.7% vs. 25.7%) for all health expenditure, while households support \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n33\n43.5%. (38) However, the Senegalese government finances less than 10% of healthcare \nexpenditure for the three GFATM diseases. For malaria, USAID is also heavily involved in funding.  \nUnder the Global Financing Facility (GFF) investment plan, the government of Senegal was \nexpected to contributed 34% of funding by the end of 2021 (78).\nChallenges\nGovernance, coordination and alignment \n Lack of communication and coordination between the GHIs in Senegal \n Lack of comprehensive understanding of the overall landscape of GHIs by national \nstakeholders,\n National experts leave the civil service to become consultants to GHIs\n Coordination bodies and platforms are not dynamic and effective (“lethargy”)\n Global actors are far from the real world and population needs/lives\n Power imbalance in term of establishment of priorities\n Language barriers (almost exclusive use of English)\nHealth Systems Strengthening and integration\n Fragmentation of initiatives; program verticalization\n Implementation gap (delays in implementation of interventions)\n Insufficient investment and impact of GHIs on health system strengthening (HSS), despite \nrecent efforts\n Investments in specific diseases inadequately benefit the broader healthcare system\nGHI financing\n Cumbersome procedures\n Multiplicity of windows, interlocutors, and methods of financing\n Funding spread over activities instead of building sustainability\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n34\n Over-funding for certain sectors \nMonitoring and performance accountability \n Discrepancy between resources invested and impact\n Weak monitoring mechanisms and information systems\n Weak capacity for performance accountability (by GHIs and also more generally)\nSome countries have shown notable progress in adopting a more integrated approach – for \nexample, Malawi is currently making progress on greater integration  (41); additionally, Ethiopia, \nRwanda, Somalia, and certain provinces of South Africa have been recognised as enforcing a more \nharmonised approach across funders, including GHIs (42). There is scope for countries to shape \nGHI support, where will and capacity exist, but this is not always facilitated by the GHI \nrequirements. \nAccording to South African KIs, GHIs and larger donors often by-pass government, due to lack of \ntrust in government, and provide direct funding to NGOs, CSOs, Parliament and higher education \nand research institutions, undermining control and overview of central institutions such as the \nDepartment of Health and Treasury. Reportedly, approximately half of the GFATM funds are \nallocated to government recipients, but even among those, a significant portion remains off-\nbudget (40,43). In pursuit of their goal to channel 55% of funding through government systems \nby the end of 2021, Gavi has made strides in increasing the share. However, as of 2021, only 41% \nof the (non-commodity) funding had been directed through these systems. \nCountry KIs are also sceptical about the small proportion of funding that is expended within \ncountries. Only operational funds of country grants are actually spent in the country whereas \nthe bulk of the funding often comprises supplies which are internationally procured as local \nvendors are not pre-qualified for GHI procurement. There have been long-standing concerns of \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n35\nlack of international community support to boost the local industry for supplies production, \nwhich leads to a cycle of dependency on GHI funding. \n“Local vendors are not pre-qualified. Therefore, we send back 70% of the funding to the donors \nthrough international procurement and that at a very high cost compared to the local purchase”. – \n(Pakistan KI)\nDespite the focus on minimising fiduciary risks, there are also concerns that the GHIs (GFATM \nand Gavi in particular) may inadvertently contribute to or escalate corruption risks. This concern \nstems from the use of multiple independent bank accounts and off-budget systems, which can \ncreate opportunities for financial irregularities. Periodic crises have been linked to poor \naccounting practices and inadequate tracking of fund usage (44–48).  \nNarratives and framing\nPerformance narratives\nGHIs justify themselves in relation to results in their focal areas, but there is much contestation \nabout how those results are generated and whether they reflect others’ investments in the results \nchain. While the GHIs are recognised to have made substantial contributions to the results chain \nfor their focal areas, many global KIs and the literature (49–51) reported that some of them over-\nclaim results, especially on blunt indicators such as ‘lives saved’. Specifically, they are perceived \nto claim credit for the entire outcome of broader investments, which encompassed contributions \nfrom LMIC governments and from other funders. In some cases, reported results have been \nprimarily based on modelling, rather than comprehensive evaluations. \n“They get the receipts [for inputs], but they don’t really know what they are producing.” (Global KI)\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n36\nThe GFF has moved away from this model and reports on assessed contribution to \nnational/country results, with a clear line of sight to the nature and value add of the GFF \ncontributions, which made their reported results less questioned by KIs.  However, this was \nmentioned by some KIs as having weakened their case for impact in comparison to some other \nGHI claims.  This shows the pressure that GHIs are under to compete and ‘out claim’ one another \nin order to attract or maintain funding. \nIn response to concerns about health system impacts (52,53), there has been an increased focus \nin GHI policies on ‘HSS’ investments. However, with GFATM the classification of spending as \nsupporting resilient and sustainable systems for health (RSSH) was also questioned by global KIs, \nwho claim that what is counted as RSSH and what is seen as disease-specific does not follow a \nclear logic.  There has been ongoing debate and lack of clarity around how much money spent by \nGFATM and Gavi can be classified as actually strengthening the health system in a sustainable \nway (54).  Various attempts to classify expenditure have been made, ranging from 27% to 7% of \ninvestment (55,56).  \nSeveral KIs mentioned that the narrative is dominated by what they interpreted as powerful and \nvocal interests grouped around the GHIs at global level, which have strong interests in \nemphasising the strengths and successes of GHI activities, and have the resources to amplify this \nmessage.   This is in contrast to more critical voices at country level and globally, which are not \nable to project their views with such power. As was highlighted in the governance section, some \nBoard members also feel less able to speak out in the face of these power differentials. \nNarratives about capacity\nAt the national level, particularly in contexts of financial dependence, there can be a mutual blame \ngame, in which GHIs and other partners lament lack of national capacity and planning which \nforces them to play a dominant role, while national counterparts resent their lack of control, \nownership and independence, blaming GHIs for undermining these and not building their \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n37\ncapacity. Both sides have an element of justice and the behaviour on both sides can reinforce \ncontinued patterns of this nature.\n‘The government is meant to set targets but GHIs set priorities because the government is unable \nto define priorities. The country is thus pushed to achieve targets set elsewhere with the local \ncontext (e.g. economic climate, available resources, burden of disease, political realities) ignored. \nThis is because of very limited state capacities that is reflected in a weak national programme, a \nHealth Department with no vision or capacity, the absence of a public health approach, (realistic) \nhealth financing strategy or medium-term (five-year) and long-term (15-20 year) plans.” \n(Pakistan KI)\nPart of the challenge relates to the timeframe and institutional incentives of GHIs, which have \nrelatively short funding cycles, while building capacity takes longer and is harder to measure.\n“[GHIs are] top-down, selective, short-termist, and kind of have a bias towards delivering things \nthat can be measured. In a neglect of important things that need to be improved or strengthened. \nBut which can’t necessarily be measured in a way these initiatives tend to want to measure things \n– which is by counting things.” (Global KI)\n“So health systems work is by nature difficult. Part of what it achieves is preventing more bad \nthings from happening. That's always difficult to gauge and assess” (Global KI)\nSome of the divergence of discourse on the impact of GHIs relates to respondents focusing on \ndifferent outcomes – in particular, short-term gains in coverage in specific areas versus longer \nterm changes to how system operate. The fact that GHIs primarily fund inputs means that there \nis continuing dependence in the longer term.\n“We've done really well over 20 years in bringing down the incidence rate of HIV, saving people \nfrom dying of HIV with TB and malaria as well. But of course as soon as the money dries up, that all \nstarts to disappear, all those gains, and that's what we saw over COVID, right?” (Global KI)\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n38\nNarratives about risks\nIt is also important how risks are framed. The GHI systems are in many cases primarily designed \nto prioritise minimising fiduciary risk, which is crucial for donors. However, that may not be \ninherently more important than addressing programme and system risks, such as the risk of \nfailing to achieve progress, failing to strengthen programmes, or causing unintended harm to \nhealth systems. Enhancing effectiveness may involve increasing flexibility, even if it results in \nhigher fiduciary risk. This aspect becomes particularly significant in FCAS, where the \ncircumstances are dynamic and require adaptability. KIs point out that more work needs to be \ndone on balancing the costs of different approaches and using more context-adapted measures.\n“There is a problem with the financing flexibility. The Global Fund, for example, has very strict \nbudget lines and in conflict settings, it does not allow us to adapt according to the current \nsituation.” (EMRO consultation KI)\nNarratives about potential reforms\nThe data revealed divergent perspectives on the role and possible future path of the GHIs (see \nBox 4). Some implementers and funders were incrementalist in their approach to change, \nwhereas other country-level actors, multilaterals, and academics tended to be more radical. There \nis also a lot of variation within these groups. It is notable that there were surprisingly critical \nvoices from within the GHIs themselves, reflecting the divergent pressures that staff within them \nare having to manage. \nBox 4. Reform scenarios and narratives\nThree predominant reform narratives emerged from the interviews and consultations. These are \nsummarised here.\n1. Narrative of status quo – this narrative, predominantly emanating from some parts of the \nGHIs but also some of their implementing partners, focuses on the big benefits delivered by \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n39\nGHIs; it views the GHIs as one of the more adaptive, successful elements of the global health \ninfrastructure (‘why are you picking on us?’), on their successful mobilisation of funds (with \na threat of their withdrawal if GHIs were too radically altered), their focus on vulnerable \npopulations and innovative models of governance and financing. Problems that have arisen \nare presented as largely due to weaknesses of systems and capacity at country level. The GHIs \nshould therefore continue to operate broadly as they do, with minor adjustments.\n2. Narratives of radical reform – this narrative, which emanates from a range of respondents \n(academics, partners in multilaterals, also some GHI staff) highlights that GHIs have been \noverselling their success, as well as (in some cases) causing harms through fragmented, \ndistortionary funding, and not focusing on the need to build sustainable, integrated systems. \nFurther, they offer poor efficiency through input financing, are prolonging their own \nmandates beyond the original planned timespans, have low accountability to beneficiary \ngovernments, lack transparency on data, and have imposed high costs for governments and \nothers to access grants though complexity and lack of coordination between GHIs and other \nactors. An end date should be set for the GHIs, either very soon or in the foreseeable future.\n3. Narrative of moderate/iterative reform – in this view, which emanated from a range of \nrespondents including country partners and funders, these GHIs do make an important \ncontribution but their systems need to evolve to focus more on transition, capacity building, \nsustainability at country level, alongside the provision of global public goods, with recognition \nof the ongoing financial dependence for a smaller group of countries which are low income \nand/or fragile and conflict-affected. The focus of reforms should be on improving the \nfunctioning of the GHIs, which could include a range of actions from merger to shared \nfunctions, better alignment with country systems and one another, changed processes to \nreduce transaction costs for governments and implementers, and more support for integrated \nhealth systems.\nThe positive narrative about results noted above makes changes to the status quo more difficult. \nGHIs rely heavily on these narratives to make the case for their continued importance and \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n40\nexistence, providing information systems and data to support their positions. At the same time, \ncritical narratives emerged from our interviews, which support radical reforms. There is a \ndiscrepancy between these more radical voices and the official narratives within GHIs about \nreform, which weakens the possibility of agreement on the way forward.\nThe positive narrative about results noted above makes changes to the status quo more difficult. \nGHIs heavily rely on these narratives to make the case for their existence, and also use \ninformation systems and data to support it. At the same time, critical narratives emerged from \nour interviews, which support radical reforms. However, these narratives are not the official GHIs \nones and there is a discrepancy between the official and informal narratives within GHIs about \nreform, which weakens the possibility of agreement on the way forward.\nWhile reforming existing institutions is challenging, establishing new institutions appears to be \nan altogether easier route to plan to respond to new global challenges. Hence proliferation and \nfragmentation are perpetuated, impacting on recipient countries. Over the past few years, several \nnew global funds have been created, including the Global Oxygen Alliance (57), the Hepatitis Fund \n(58), Health4Life Fund (59), the Pandemic Fund (60), and the Health Impact Investment Platform \n(61). The relevance, functioning and unintended consequences of these new funds, largely \nsupported by the same bilateral donors, UN agencies and foundations, need to be evaluated. They \nadd a new layer of complexity and fragmentation to the global health architecture and at national \nlevel, where each initiative focuses on a specific field, such as sexual and reproductive health and \nrights, HIV, or innovation, and operates with its own programs, governance structures, \nmechanisms, and approaches. \n\"The mechanisms are fragmented, but the public health problems they tackle are not\" (Senegal KI)\nAnother potential reform that was mentioned is the expansion of mandates of existing GHIs. \nHowever, some interviewees, especially global KIs, expressed concern about what they perceived \nas constantly expanding mandates, particularly regarding the GFATM and Gavi. They pointed out \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n41\nthat these organisations have been expanding their roles and venturing into new areas, such as \nHSS (52,56). However, in their opinion, there is little evidence to suggest that GHIs are \nappropriately structured and technically equipped to handle these responsibilities effectively \n(South Africa KI; regional consultation).\nDiscussion\nIn this article, we examined the role of GHIs within the global and national health architecture \nfrom a political economy perspective in order to understand patterns, points of resistance and \npossibility for reforms. This work is original in that there have been many analyses of and \ncritiques of the GHIs but none which have looked with the lens of political economy, bringing in \nviews from a large range of global, regional and national experts.\nThe current arrangement, with its strengths and weaknesses, is not accidental but emerged from \na specific period which focused on reaching global goals on priority diseases, especially \ncommunicable ones (62), and when international funding was growing. Since then, the landscape \nhas changed, particularly in relation to the emergence of non-communicable diseases and the \nhealth impacts of climate change, and financing for international support is under strain. \nHowever, the structures which were established 20 years ago have created a path dependency, \nwith large, complex bureaucracies (in some cases; the scale is very varied across them) which \nhave momentum and can resist reforms, as well as a large network of clients (including \ngovernments, implementers, consultants, etc.) who are interested in maintaining the status-quo. \nReflecting on the lessons that KIs and literature highlighted in relation to previous efforts at \ncoordination and alignment, it is clear that individuals and organisations follow their own \nincentives, which need to be altered for behaviour change to follow. Voluntarist approaches to \nreform, which do not change rewards and sanctions are unlikely to gain traction (63,64).  \nThe actors involved are numerous, diverse, interconnected and have interests which largely \nfavour status quo or at most incremental reforms. These actors do not fit into neat categories and \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n42\neven at individual level can play multiple roles – for example, benefiting from being a consultant \nto GHIs at national level, while also holding a more critical perspective in a government role. \nThe GHIs themselves are also part of a wider network of DAH organisations, which interact with \nGHIs, with country health systems and with one another to influence outcomes, which makes \nreform highly complex. All are responsible and none are, which is a perfect setting for mutual \nblame and inaction on change.\nPower to bring about change is not evenly distributed – some actors have more power and \ninfluence, especially major funders and senior leadership in the GHIs, but they also have to create \nconsensus, work in coordinated ways and draw on wider legitimacy if they wish to enact reforms. \nFor that process, which started with the Lusaka Agenda, the ability to draw on powerful \nnarratives and clear accountability measures for reform will be significant.(65) Ultimately, all the \nelements of the political economy framework emerged as important here: the position and power \nof key actors, but also the context factors (financing flows and governance structures) which \naffect how GHIs function and how decisions are made, and the narratives and framing which \ninfluence both whether change is seen to be needed and what form it might take.\nIt is important to restate the differences between GHIs and note that the three country-facing \nGHIs exist on a continuum of integration with national systems, with the GFF most integrated \nthrough its provision of public financing, while Gavi is able to pool funds at national level and the \nGlobal Fund is least enabled to operate in that way. At national level, there are also variations in \nthe dynamics observed in this study; for example, countries with greater financial dependence on \nthe support of GHIs typically raised more concerns about their functioning, while better funded \nhealth systems (or sub-national components of health systems) were better able to use GHI \nsupport in ways that did not disrupt their operations.\nAs the GHIs continue to evolve in a dynamic global health environment, the deployment of \npolitical economy as a lens to understand what is possible, to understand change and its absence, \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n43\nand to strategise around building coalitions for reform (9) will continue to be very relevant from \nboth an academic and policy perspective. (66)\nConclusion\nThis paper has highlighted some of the key critiques and current dissatisfactions at national level with \nGHIs that are most active within country health systems.  It has also described how the GHIs are part \nof a wider complex and interdependent ecosystem and that their role has evolved in relation to other \nactors, all of which play a part in the patterns noted here. Reform of the GHIs will involve changes by \nthese wider actors, especially the funders, recipient countries, senior leaders in GHIs and influential \nNGOs, and will reflect shifting interests and narratives. Potential for change comes from the current \nperceptions of constrained resources and increasing threats, but this does not guarantee \nstrengthening of the role of GHIs unless consensus is reached around narratives of how the current \nsystem is working and options developed which serve the interests of key constituencies. Political \neconomy analysis can help to highlight these issues and point to strategies for managing them. \nAcknowledgements\nWe would like to acknowledge the support of the Government of Norway and the Wellcome Trust in \nfunding and guiding the original research. The opinions expressed here are however the responsibility \nof the authors alone. We would also like to thank all of our many key informants for their insights, and \nthank ReBUILD for Resilience research consortium for supporting publication of this article.\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n44\nReferences\n1. Institute For Health Metrics and Evaluation (IHME). Financing Global Health [Internet]. \nSeattle, WA; 2024 [cited 2024 Aug 10]. Available from: \nhttp://vizhub.healthdata.org/fgh/%E2%80%8B\n2. World Bank Group Development Finance. Understanding Trends in Proliferation and \nFragmentation for Aid Effectiveness during Crises [Internet]. 2022 Jul [cited 2024 Aug 10]. \nAvailable from: \nhttps://thedocs.worldbank.org/en/doc/ef73fb3d1d33e3bf0e2c23bdf49b4907-\n0060012022/understanding-trends-in-proliferation-and-fragmentation-for-aid-effectiveness-\nduring-crises?_gl=1*d1g7ww*_gcl_au*MjY4MjU3MjgxLjE3MjIzMTk1MzU.\n3. Mwangangi M, Røttingen JA. The Lusaka Agenda: Conclusions of the Future of Global Health \nInitiatives Process [Internet]. Lusaka; 2023 Dec [cited 2024 Aug 10]. Available from: \nhttps://futureofghis.org/final-outputs/lusaka-agenda/\n4. Witter S, Palmer N, James R, Zaidi S, Carillon S, English R, et al. Reimagining the Future of \nGlobal Health Initiatives, Research Report [Internet]. 2023 [cited 2024 Aug 10]. Available \nfrom: https://futureofghis.org/research-other-inputs/reimagining-the-future-of-global-\nhealth-initiatives-study/\n5. Sundewall J, McCoy D, Brugha R. Correspondence: Global health initiatives and country \nhealth systems. The Lancet. 2009 Oct 10;374:1237–8. \n6. Cohn J, Russell A, Baker B, Kayongo A, Wanjiku E, Davis P. Using global health initiatives to \nstrengthen health systems: a civil society perspective. Glob Public Health. 2011;6(7):687–702. \n7. Hanefeld J. How have global health initiatives impacted on health equity? Promot Educ. \n2008;15(1):19–23. \n8. Zicker F, Faid M, Reeder J, Aslanyan G. Building coherence and synergy among global health \ninitiatives. Health Res Policy Syst. 2015;13:1–8. \n9. Witter S, Sparkes S, Bertone MP, Ozcelik E. Political economy analysis for health financing: A \n‘how to’guide. World Health Organization; 2024. \n10. Fritz V, Levy B, Ort R. Problem-driven political economy analysis: The World Bank’s \nexperience. World Bank Publications; 2014. \n11. Bertone MP, Wurie H, Samai M, Witter S. The bumpy trajectory of performance-based \nfinancing for healthcare in Sierra Leone: agency, structure and frames shaping the policy \nprocess. Global Health. 2018;14:1–15. \n12. Harris D. Applied political economy analysis. A problem-driven framework London: Overseas \nDevelopment Institute. 2013;4. \n13. Varvasovszky Z, Brugha R. A stakeholder analysis. Health Policy Plan. 2000;15(3):338–45. \n14. Balane MA, Palafox B, Palileo-Villanueva LM, McKee M, Balabanova D. Enhancing the use of \nstakeholder analysis for policy implementation research: towards a novel framing and \noperationalised measures. BMJ Glob Health. 2020;5(11):e002661. \n15. Schmeer K. Stakeholder analysis guidelines. Policy toolkit for strengthening health sector \nreform. 1999;1:1–35. \n16. Hoffman SJ, Cole CB. Defining the global health system and systematically mapping its \nnetwork of actors. Global Health. 2018;14:1–19. \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n45\n17. Fergus CA. Power across the global health landscape: a network analysis of development \nassistance 1990–2015. Health Policy Plan. 2022;37(6):779–90. \n18. Victoria Fan and Eleni Smitham. Center For Global Development (blog post). 2023 [cited 2024 \nAug 10]. Is the New Pandemic Fund Where Vertical Can Finally Meet Horizontal? Available \nfrom: https://www.cgdev.org/blog/world-banks-new-pandemic-fund-where-vertical-can-\nfinally-meet-horizontal\n19. Micah AE, Bhangdia K, Cogswell IE, Lasher D, Lidral-Porter B, Maddison ER, et al. Global \ninvestments in pandemic preparedness and COVID-19: development assistance and domestic \nspending on health between 1990 and 2026. Lancet Glob Health. 2023;11(3):e385–413. \n20. Office of the Inspector General. The Global Fund. 2021 [cited 2024 Aug 10]. Evolving the \nTechnical Review Panel Model. Available from: \nhttps://www.theglobalfund.org/en/oig/updates/2021-11-04-evolving-the-technical-review-\npanel-model/\n21. de Bengy Puyvallée A. The rising authority and agency of public–private partnerships in global \nhealth governance. Policy Soc. 2024;43(1):25–40. \n22. Heywood M. South Africa’s treatment action campaign: combining law and social \nmobilization to realize the right to health. J Hum Rights Pract. 2009;1(1):14–36. \n23. Birn AE. Philanthrocapitalism, past and present: The Rockefeller Foundation, the Gates \nFoundation, and the setting (s) of the international/global health agenda. Hypothesis. \n2014;12(1):e8. \n24. McGoey L. Philanthrocapitalism and its critics. Poetics. 2012;40(2):185–99. \n25. Eckl J, Hanrieder T. The political economy of consulting firms in reform processes: the case of \nthe World Health Organization. Rev Int Polit Econ. 2023;30(6):2309–32. \n26. Eckl J, Hanrieder T. The political economy of consulting firms in reform processes: the case of \nthe World Health Organization. Rev Int Polit Econ. 2023;30(6):2309–32. \n27. Hanefeld J. The global fund to fight AIDS, tuberculosis and malaria: 10 years on. Clinical \nmedicine. 2014;14(1):54–7. \n28. Craveiro I, Dussault G. The impact of global health initiatives on the health system in Angola. \nGlob Public Health. 2016;11(4):475–95. \n29. Calain P, Abu Sa’Da C. Coincident polio and Ebola crises expose similar fault lines in the \ncurrent global health regime. Confl Health. 2015;9:1–7. \n30. Ooms G, Van Damme W, Baker BK, Zeitz P, Schrecker T. The’diagonal’approach to Global \nFund financing: a cure for the broader malaise of health systems? Global Health. 2008;4:1–7. \n31. Spicer N, Walsh A. 10 best resources on… the current effects of global health initiatives on \ncountry health systems. Health Policy Plan. 2012;27(3):265–9. \n32. Global Financing Facility (GFF). Global Financing Facility. 2024 [cited 2024 Aug 10]. Investors \nGroup (webpage). Available from: www.globalfinancingfacility.org/governance/investors-\ngroup\n33. English R, Dudley L, Daniels K, Hendricks L, Blows S, Pillay Y. Reimagining the Future of Global \nHealth Initiatives study: South Africa country case study [Internet]. 2023 [cited 2024 Aug 25]. \nAvailable from: https://d2nhv1us8wflpq.cloudfront.net/prod/uploads/2023/08/Appendix-9-\nSouth-Africa-country-case-study.pdf\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n46\n34. Chang AY, Cowling K, Micah AE, Chapin A, Chen CS, Ikilezi G, et al. Past, present, and future of \nglobal health financing: a review of development assistance, government, out-of-pocket, and \nother private spending on health for 195 countries, 1995–2050. The Lancet. \n2019;393(10187):2233–60. \n35. Clinton C, Sridhar D. Who pays for cooperation in global health? A comparative analysis of \nWHO, the World Bank, the Global Fund to Fight HIV/AIDS, Tuberculosis and Malaria, and \nGavi, the Vaccine Alliance. The Lancet. 2017;390(10091):324–32. \n36. Sridhar D, Woods N. Trojan multilateralism: global cooperation in health. Glob Policy. \n2013;4(4):325–35. \n37. The Lancet. The Global Fund: replenishment and future-proofing. The Lancet. 2022 \nSep;400(10355):787. \n38. Ministère de la Santé et de l’Action Sociale (République du Senégal). Rapport des comptes de \nla santé 2017/2021 [Internet]. Dakar; 2022 Dec [cited 2024 Sep 28]. Available from: \nhttps://p4h.world/app/uploads/2024/01/RAPPORT-COMPTES-DE-LA-SANTE-SENEGAL-2017-\n2021-.x23411.pdf\n39. World Health Organization (WHO). Global Health Expenditure Database [Internet]. [cited \n2024 Aug 10]. Available from: https://apps.who.int/nha/database\n40. Organisation for Economic Co-operation and Development (OECD). Creditor Reporting \nSystem (online database) [Internet]. [cited 2024 Aug 10]. Available from: https://www.oecd-\nilibrary.org/development/data/creditor-reporting-system_dev-cred-data-en\n41. Sharma L, Heung S, Twea P, Yoon I, Nyondo J, Laviwa D, et al. Donor coordination to support \nuniversal health coverage in Malawi. Health Policy Plan. 2024;39(Suppl 1):i118. \n42. Tadesse AW, Gurmu KK, Kebede ST, Habtemariam MK. Analyzing efforts to synergize the \nglobal health agenda of universal health coverage, health security and health promotion: a \ncase-study from Ethiopia. Global Health. 2021;17(1):53. \n43. World Bank Group Development Finance. A Changing Landscape: Trends in official financial \nflows and the aid architecture [Internet]. 2021 Nov [cited 2024 Aug 10]. Available from: \nhttps://thedocs.worldbank.org/en/ doc/9eb18daf0e574a0f106a6c74d7a1439e-\n0060012021/original/A-Changing-Landscape-Trends-inOfficial-Financial-Flows-and-the-Aid-\nArchitecture-November-2021.pdf\n44. Kohler JC, Bowra A. Exploring anti-corruption, transparency, and accountability in the world \nHealth organization, the United nations development programme, the world bank group, and \nthe global fund to fight AIDS, tuberculosis and malaria. Global Health. 2020;16:1–10. \n45. Gorodensky A, Bowra A, Saeed G, Kohler J. Anti-corruption in global health systems: using key \ninformant interviews to explore anti-corruption, accountability and transparency in \ninternational health organisations. BMJ Open. 2022;12(12):e064137. \n46. García PJ. Corruption in global health: the open secret. The Lancet. 2019;394(10214):2119–\n24. \n47. Mackey TK, Kohler JC, Savedoff WD, Vogl F, Lewis M, Sale J, et al. The disease of corruption: \nviews on how to fight corruption to advance 21 st century global health goals. BMC Med. \n2016;14:1–16. \n48. Chang Z, Rusu V, Kohler JC. The Global Fund: why anti-corruption, transparency and \naccountability matter. Global Health. 2021;17:1–11. \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n47\n49. McCoy D, Jensen N, Kranzer K, Ferrand RA, Korenromp EL. Methodological and policy \nlimitations of quantifying the saving of lives: a case study of the global fund’s approach. PLoS \nMed. 2013;10(10):e1001522. \n50. Friebel R, Silverman R, Glassman A, Chalkidou K. On results reporting and evidentiary \nstandards: spotlight on the Global Fund. The Lancet. 2019;393(10184):2006–8. \n51. Hanefeld J. How have global health initiatives impacted on health equity? Promot Educ. \n2008;15(1):19–23. \n52. Storeng KT. The GAVI Alliance and the ‘Gates approach’to health system strengthening. Glob \nPublic Health. 2014;9(8):865–79. \n53. Biesma RG, Brugha R, Harmer A, Walsh A, Spicer N, Walt G. The effects of global health \ninitiatives on country health systems: a review of the evidence from HIV/AIDS control. Health \nPolicy Plan. 2009 Jul 1;24(4):239–52. \n54. Bertone MP, Palmer N, Kruja K, Witter S, 1 HWG. How do we design and evaluate health \nsystem strengthening? Collaborative development of a set of health system process goals. Int \nJ Health Plann Manage. 2023;38(2):279–88. \n55. The Global Fund. Managing investments in Resilient and Sustainable Systems for Health \n[Internet]. Geneva; 2019 May [cited 2024 Sep 29]. Available from: \nhttps://www.theglobalfund.org/media/8441/oig_gf-oig-19-\n011_report_en.pdf#:~:text=The%20Office%20of%20the%20Inspector%20General%20(OIG)%\n20safeguards%20the%20assets\n56. The Global Fund Technical Evaluation Reference Group. Global Fund Mapping Health Systems \nStrengthening (HSS) Component of the Resilient and Sustainable Systems for Health (RSSH) \nInvestments: TERG Position Paper, Management Response, and Final Report [Internet]. 2023 \nJun [cited 2024 Aug 10]. Available from: \nhttps://www.theglobalfund.org/media/13115/terg_mapping-hss-component-\nrssh_report_en.pdf\n57. Global Oxygen Alliance (website) [Internet]. [cited 2024 Aug 10]. Available from: \nhttps://globaloxygenalliance.org/\n58. Thornton J. Hepatitis Fund aims to accelerate viral hepatitis elimination. The Lancet. \n2023;401(10386):1414–5. \n59. World Health Organization (WHO). Health4Life Fund. A global financing partnership on non-\ncommunicable diseases and mental health [Internet]. 2022 [cited 2024 Aug 10]. Available \nfrom: https://knowledge-action-portal.com/en/news_and_events/news/6559\n60. Boyce MR, Sorrell EM, Standley CJ. An early analysis of the World Bank’s Pandemic Fund: a \nnew fund for pandemic prevention, preparedness and response. BMJ Glob Health. 2023 \nJan;8(1):e011172. \n61. Samarasekera U. New US $1· 5 billion fund for primary care. The Lancet. \n2023;402(10397):174. \n62. Sachs JD, Binagwaho A, Birdsall N, Broekmans J, Chowdhury M, Garau P, et al. Investing in \nDevelopment A Practical Plan to Achieve the Millennium Development Goals: Overview. \nRoutledge; 2019. \n63. Labonte R, Marriott A. IHP+: little progress in accountability or just little progress? The \nLancet. 2010;375(9725):1505–7. \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n48\n64. Shorten T, Taylor M, Spicer N, Mounier-Jack S, McCoy D. The International Health Partnership \nPlus: rhetoric or real change? Results of a self-reported survey in the context of the 4 th high \nlevel forum on aid effectiveness in Busan. Global Health. 2012;8:1–13. \n65. Witter S, Baker P. Tracking Delivery on the Lusaka Agenda. Policy Paper 336 [Internet]. Center \nfor Global Development. Washington, DC; 2024 Sep [cited 2024 Sep 28]. Available from: \nhttps://www.cgdev.org/publication/tracking-delivery-lusaka-agenda\n66. Drake T, Khan M. Center for Global Development. 2024 [cited 2024 Sep 28]. A New Compact \nfor Health Financing: The Global Political Economy of Reform. Available from: \nhttps://www.cgdev.org/blog/new-compact-health-financing-global-political-economy-reform\n67. The Global Fund to Fight AIDS Tuberculosis and Malaria. Annual Financial  Report 2023 \n[Internet]. Geneva; 2024 Apr [cited 2024 Aug 10]. Available from: \nhttps://www.theglobalfund.org/media/14110/corporate_2023annualfinancial_report_en.pdf\n68. Gavi the vaccine alliance. webpage. 2024 [cited 2024 Aug 10]. Funding: Current period 2021–\n2025. Available from: https://www.gavi.org/investing-gavi/funding/current-period-2021-\n2025\n69. The Global Financing Facility (GFF). The Global Financing Facility in Support of Every Woman \nEvery Child: Executive Summary [Internet]. [cited 2024 Aug 10]. Available from: \nhttps://www.worldbank.org/content/dam/Worldbank/document/HDN/Health/GFF-\nExecutive-Summary_EN.pdf\n70. Unitaid. Proposed Operating Expenses Budget 2023 [Internet]. Geneva; 2022 Dec [cited 2024 \nAug 10]. Available from: https://unitaid.org/assets/UNITAID_EB41_2022_7_2023-Budget.pdf\n71. Unitaid. Unitaid’s Investment Case 2023-2027 [Internet]. Geneva; 2023 [cited 2024 Aug 10]. \nAvailable from: https://unitaid.org/assets/Unitaid_InvestmentCase_Report.pdf\n72. FIND. FIND Strategy 2021: Testing for healthy and safe lives [Internet]. Geneva; 2021 [cited \n2024 Aug 10]. Available from: https://www.finddx.org/about-us/strategy/\n73. CEPI. Funding and Expenditure (December 2018) [Internet]. London, UK; 2024 [cited 2024 \nAug 10]. Available from: https://static.cepi.net/downloads/2023-12/050319-Funding-and-\nExpenditure-Final_V3.pdf\n74. Pakistan Bureau of Statistics (PBS) G of P. National Health Accounts Pakistan 2019-20 \n[Internet]. Islamabad; 2022 [cited 2024 Aug 25]. Available from: \nhttps://www.pbs.gov.pk/publication/national-health-accounts-pakistan-2019-20\n75. WHO Regional Office for the Eastern Mediterranean. Country Private Health Sector Effective \nEngagement for Service Provision Pakistan. Cairo; 2019. \n76. The Global Fund Office of the Inspector General. Audit Report: Global Fund Grants in the \nIslamic Republic of Pakistan [Internet]. Geneva; 2020 Apr [cited 2024 Aug 25]. Available from: \nhttps://www.theglobalfund.org/media/9595/oig_gf-oig-20-012_report_en.pdf\n77. Global Financing Facility. Pakistan country profile: Global Financing Facility Data Portal. \n78. Direction de la Santé de la Mère et de l’Enfant. Ministère de la Santé et de l’Action Sociale \n(République du Senégal). Dossier d’investissement pour la réduction de la \nmortalité  maternelle, néo-natale,  infanto-juvénile, des adolescents et des jeunes [Internet]. \nDakar; 2019 Jun [cited 2024 Sep 29]. Available from: \nhttps://www.globalfinancingfacility.org/sites/gff_new/files/documents/Senegal-dossier-\ndinvestisement.pdf\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n49\n \n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint \n\n . CC-BY 4.0 International licenseIt is made available under a \nperpetuity. \n is the author/funder, who has granted medRxiv a license to display the preprint in(which was not certified by peer review)preprint \nThe copyright holder for thisthis version posted October 4, 2024. ; https://doi.org/10.1101/2024.10.04.24314895doi: medRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}