Equitable Partnerships in Global Health Research

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-11-11

By the end of this article you will be able to define what makes a global health research partnership equitable, walk through the steps of building one from first contact to final publication, audit an existing collaboration with a simple worked calculation, and recognize the mistakes that most often derail North-South projects. Nothing needs to be installed to read along, but to run the worked example yourself you will need Python 3 with no extra packages, plus your own partnership documents such as the memorandum of understanding, budget, authorship records, and ethics approvals.

What Equitable Partnerships in Global Health Research Mean

An equitable partnership in global health research is a collaboration between institutions, usually one in a high-income country (HIC) and one or more in low- and middle-income countries (LMICs), in which decision-making power, resources, authorship, data ownership, and benefits are shared fairly. A low- and middle-income country is a classification used by the World Bank for economies with lower gross national income per capita. The opposite of an equitable partnership is a transactional or extractive one, where the HIC partner sets the agenda, holds the funds, exports the data, and publishes the results with local contributors listed in minor authorship positions.

Three related terms anchor this field. Research ethics is the system of principles and review processes that protects participants and ensures scientific integrity. Responsible conduct of research covers the day-to-day practices that make research trustworthy, including honest data handling, fair authorship, and transparent reporting. Decolonization, in the global health context, refers to dismantling the colonial-era patterns that still determine who decides, leads, and benefits from research, a framing developed explicitly by Kumar and McCoy in their checklist for research funders [1].

The evidence that inequity persists is not anecdotal. A scoping review of 56 peer-reviewed studies published between 2011 and 2024 concluded that global health continues to perpetuate unequal power dynamics rooted in colonization and disparate resource access, even as acknowledgment of the problem grows across the field [2]. Firsthand testimonies from African researchers across multiple countries and decades document patterns of exploitation, marginalization, and systemic bias in how partnerships are conceived, funded, and executed, from the framing of funding priorities to control of finances and data to the sidelining of African intellectual leadership [3].

Why the Problem Is Structural, Not Interpersonal

It is tempting to treat unequal partnerships as a matter of individual bad behavior. The published record points to structures instead.

Funding architecture. Large donors defend autonomy over research priorities, and coordination among funders has historically been weak, which shapes which problems get studied and where [4]. On the receiving end, dependency runs deep. A bibliometric analysis of health research from Jharkhand, India, covering 2013 to 2023, found that 71 percent of funding for the 209 analyzed articles originated from international sources and 21 percent of corresponding authors came from international institutes [5].

Control points. The testimony-based commentary by Nadarajah and colleagues groups the inequities thematically: funders frame priorities externally, financial resources are controlled by HIC institutions, data flows out of the host country, accountability practices are distorted, and local intellectual leadership is marginalized [3].

Uneven outcomes. A survey across the Data Science Initiative for Africa network, analyzed with structural equation modeling, found that human capacity was the strongest positive predictor of collaboration effectiveness (β = 0.467, p = 0.008), while project and institutional practices showed a significant negative association (β = -0.464, p = 0.016). The study framed its findings through the lens of epistemic injustice, which means the unfair devaluation of some people's knowledge and ways of knowing [6].

Consequences for evidence itself. When local researchers cannot set agendas, locally generated evidence disappears. In pain research, for example, populations in LMICs carry a disproportionate burden of chronic pain yet remain underrepresented in high-quality studies, forcing clinicians and policymakers to rely on externally produced research that may not translate to local cultural, linguistic, or health-system contexts [7].

Understanding these structures matters because the fixes are structural too. Writing a friendlier email will not rebalance a budget.

The Frameworks You Can Use

Four frameworks appear repeatedly in the recent literature, and each answers a different question.

Framework	What it addresses	Best used for
, -	, -	, -
ESSENCE principles, operationalized through the RESPIRE collaboration	Moving from transactional, individual-training models to locally anchored, LMIC-led research ecosystems	Designing capacity strengthening programs [8]
Brocher Declaration (six principles)	Ethical standards for academic and NGO partnerships	Evaluating an existing or planned bilateral partnership [9]
Funder checklist (six strategic areas)	Structural and political dimensions of inequity in the funding ecosystem	Advocacy with funders, grant design [1]
Ecosystem approach to ethical reflexivity	Distributing responsibility for ethical conduct across the whole research system	Rethinking oversight, ethics committee relationships [10]

The Brocher Declaration's six principles are worth memorizing because they convert well into a memorandum of understanding: mutual partnership with bidirectional learning, an empowered host country and community that defines the needs, sustainable programs and capacity building, compliance with applicable laws and ethical standards, humility and cultural sensitivity, and accountability for actions [9]. A 20-year partnership between Lwala Community Alliance in western Kenya and Vanderbilt University in the United States was evaluated against these six principles across education, research, and service, engaging 57 graduate students in 33 discrete projects, and the case study demonstrates that long-term, framework-grounded collaboration is achievable [9].

The ESSENCE principles, applied in the RESPIRE collaboration across South and Southeast Asia, target a specific failure mode: capacity strengthening initiatives that remain externally driven, short-term, and focused on individual training rather than sustainable institutional ecosystems. RESPIRE's enabling mechanisms were local leadership, regional peer learning, governance, mentorship, and community engagement [8].

Equitable Partnerships in Global Health Research Step by Step

The following walkthrough takes you from first contact to a sustained collaboration. Adapt the sequence to whether you are starting fresh or auditing something already underway.

Step 1: Co-develop the research question before anything else

The host country and community define the needs. That is the Brocher Declaration's second principle, and it is also the step most often skipped [9]. In practice this means the LMIC partners identify the health problem, the research question, and the study design constraints before the HIC partner writes a grant. If the HIC institution has already secured funding and is searching for a site to implement its predefined protocol, the partnership starts from a deficit that later steps cannot fully repair.

A practical test: ask whether the local lead investigator can describe the project's rationale in their own words for a local audience. If the rationale only makes sense when recited from the funder's specific aims page, the question was not co-developed.

Step 2: Negotiate governance, budget, and money flow early

Put the money questions in writing before the first participant is enrolled. Key terms to negotiate:

Which institution holds the grant and how funds flow to the LMIC partner, including indirect cost recovery.
Whether the LMIC partner has signatory authority over its own budget lines.
Data ownership, storage location, and secondary use permissions.
Intellectual property arrangements.
A dispute resolution mechanism that does not default to the HIC institution's lawyers.

The funder checklist by Kumar and McCoy is useful ammunition here because it places the onus on funding agencies across six areas: shifting power within the health research ecosystem, prioritizing health equity in the portfolio, redressing historical underinvestment in health research systems, addressing conflicts of interest, equitably distributing research outputs, and reporting on outcomes and impact [1].

Step 3: Write authorship and data terms into the agreement

Authorship is where equity becomes visible to the outside world. Commit in the memorandum of understanding to criteria based on contribution, consistent with responsible conduct of research norms, and to target representation of LMIC-based investigators in first, last, and corresponding author positions, not only middle positions. Also specify who holds the data, where it is stored, and the conditions under which the HIC partner may analyze or share it.

For studies involving human participants, plan from the design stage to collect sociodemographic variables in a standardized way and to deliberately include the populations most affected by the condition under study, an approach highlighted in the context of equitable pain research [7].

Step 4: Build capacity that outlasts the grant

Individual training courses expire with the grant. Institutional ecosystems persist. The RESPIRE lessons emphasize local leadership, regional peer learning networks, governance structures, sustained mentorship, and community engagement as the mechanisms that convert short projects into durable research capacity [8]. The same logic applies at the level of institutions: public universities in LMICs produce the health workforce and national innovation, and equitable partnerships that foster brain circulation, meaning skilled people moving between settings and returning with skills and connections, strengthen the local knowledge ecosystem rather than draining it [11].

Step 5: Design engagement as partnership, not participation

The distinction matters. A study of women's involvement in climate and health research in Indonesia and Malaysia found that women were willing and able to contribute at all stages of the research cycle, provided safe, supportive spaces were created, including women-only environments that allowed open discussion. The study's framing was moving participants toward partners who shape agendas, methodologies, and outcomes [12]. Translating that finding into practice means budgeting for community engagement activities, creating the conditions under which marginalized group members can speak candidly, and giving community representatives real authority over decisions rather than advisory tokenism.

Step 6: Distribute the outputs and report honestly

Outputs include publications, datasets, conference presentations, and the practical benefits of the findings. The funder checklist explicitly names equitable distribution of outputs as one of its six strategic areas [1]. Operationally, that means pre-agreed authorship, data deposition under terms the LMIC partner controls, and results returned to the communities that contributed data in accessible local formats.

Step 7: Audit the partnership annually

Run the equity audit below every year and discuss the results jointly. An audit that only the HIC institution conducts reproduces the problem it measures.

A Worked Equity Audit You Can Run Yourself

The following calculation is an illustrative self-audit tool, not a validated index from the literature. It computes two parity ratios for a single collaboration: authorship parity and budget parity. A ratio of 1.0 means perfect parity. The absolute target you set is a partnership decision, and the point of the exercise is to make the numbers discussable.

# partnership_equity_audit.py
# An illustrative self-audit for a two-partner collaboration.
# Partner A = LMIC institution, Partner B = HIC institution.

counts = {
    "first_authors_A": 2,   # LMIC-based first authors, last 3 years
    "first_authors_B": 4,   # HIC-based first authors, last 3 years
    "last_authors_A": 1,    # senior authors from LMIC institution
    "last_authors_B": 5,    # senior authors from HIC institution
    "direct_budget_A": 180000,  # USD flowing to LMIC partner
    "direct_budget_B": 620000,  # USD retained by HIC partner
}

def parity(x, y):
    # Returns min/max so the ratio is symmetric, capped at 1.0.
    if max(x, y) == 0:
        return None
    return round(min(x, y) / max(x, y), 2)

first_ratio = parity(counts["first_authors_A"], counts["first_authors_B"])
last_ratio  = parity(counts["last_authors_A"],  counts["last_authors_B"])
budget_share_A = counts["direct_budget_A"] / (
    counts["direct_budget_A"] + counts["direct_budget_B"])

print("First author parity (A vs B):", first_ratio)
print("Senior author parity (A vs B):", last_ratio)
print("LMIC share of direct budget:", f"{budget_share_A:.0%}")

Expected output when you run the script:

First author parity (A vs B): 0.5
Senior author parity (A vs B): 0.2
LMIC share of direct budget: 22%

Reading the output: first authors are split 2 to 4, so parity is 0.5. Senior (last) authorship, which typically signals who led the science, is 1 to 5, a parity of 0.2, and only 22 percent of direct budget flows to the LMIC partner. Those three numbers together describe a collaboration where the local institution contributes sites, participants, and staff but neither leads the science nor controls the money. That pattern matches the themes documented across African researcher testimonies, including control of financial resources and marginalization of local intellectual leadership [3].

To check your result is right, verify two things. First, confirm every author on every paper from the collaboration appears in exactly one count, using the institutional affiliation listed on the publication. Second, reconcile budget figures against the actual subaward payments, not the proposal's planned budget, because the gap between planned and actual flow is itself a common finding.

A common error when running the audit is a KeyError such as KeyError: 'direct_budget_A', which means you renamed a key in the dictionary but not in the print statements. Fix it by keeping key names identical in both places. If you get a TypeError from the f-string on older Python versions, replace f"{budget_share_A:.0%}" with f"{budget_share_A:.2f}" or upgrade to Python 3.6 or later.

Common Mistakes and Limitations

These are the failures that recur across the literature, phrased as practical warnings.

Training individuals instead of building institutions. Short courses for a few individuals leave no governance, no mentorship pipeline, and no infrastructure behind when the grant ends. The RESPIRE collaboration identified this as the core failure mode of traditional capacity strengthening and countered it with local leadership, governance, and regional peer learning [8].

Treating capacity strengthening as a one-way transfer. Bidirectional learning is the first Brocher principle for a reason. HIC partners gain contextual knowledge, methodological insight, and access to understudied populations, and pretending otherwise distorts the relationship [9].

Starting with the funding already locked in. If the HIC partner designed the study to fit a call that LMIC partners never saw, every later equity gesture is decoration. Co-development has to happen before the proposal.

Confusing inclusion with partnership. Inviting local stakeholders to workshops, as the Indonesia and Malaysia study showed, only works when safe, supportive spaces are created and when participants can move into shaping agendas and methods rather than merely attending them [12].

Ignoring ethics committees or treating them as obstacles. A workshop report from the 2022 Global Forum on Bioethics in Research in Cape Town argued that ethical conduct is the responsibility of all stakeholders, from funders and governments to research teams and ethics committees, and called for collaborative, non-adversarial relationships between researchers and committees, with committees focusing on participant agency and the common good [10]. Partnerships that plan only for minimal regulatory compliance miss this broader duty, which sits at the heart of research ethics and responsible conduct of research.

Assuming good intentions guarantee good outcomes. The scoping review found that even with increased acknowledgment of decolonization across the field, unequal power dynamics persist because of structural conditions, including resource disparities [2]. Intentions are necessary and insufficient.

Skipping the local policy environment. National frameworks increasingly define what partnership should look like. India's Draft National Health Research Policy 2026, for example, emphasizes equitable capacity building, ethical governance, and stronger public-private partnerships across its research ecosystem [13]. Partners who ignore such national policies build projects that governments will not sustain.

A final limitation applies to any equity audit, including the worked example above. Ratios of authorship and budget are proxies. They can be gamed, and they cannot capture whether community members felt heard. Use the numbers to start conversations, and pair them with qualitative review, which is exactly how the Lwala-Vanderbilt case study was evaluated, through internal documentation, program records, and qualitative description across three pillars of education, research, and service [9].

Case Examples Worth Studying

RESPIRE, South and Southeast Asia. A respiratory health collaboration that treated equity as an operational problem with enabling mechanisms: local leadership of the agenda, regional peer learning between LMIC institutions, shared governance, long-term mentorship, and community engagement. Its published lessons are addressed directly to funders, institutions, and partnerships seeking to strengthen equitable research capacity in LMICs [8].

Lwala Community Alliance and Vanderbilt University, Kenya. Twenty years of collaboration across education, research, and service, evaluated against all six Brocher Declaration principles, demonstrating that a university and a community-based NGO can sustain a partnership grounded in an explicit ethical framework [9].

Global Bridges Oncology, sub-Saharan Africa. A funding program whose review committee was composed primarily of African oncology experts, which issued a call in English, French, and Portuguese, received 218 applications, and awarded 1.1 million US dollars to 17 projects in 10 African countries. By March 2026, 11 of the 17 projects were complete and 4,565 health care professionals had been engaged through direct training and dissemination. The design lesson is that equity was built into who reviewed the grants and in which languages the call was released [14].

DS-I Africa. A continental data science network whose cross-sectional survey quantified what drives collaboration effectiveness, finding human capacity to be the strongest positive predictor and pointing to governance, leadership, and institutional practices as areas where partnerships succeed or fail [6].

Cross-disease lessons from Parkinson's research. A review of international Parkinson's research collaboration catalogs the recurring barriers, including funding inequities, regulatory and data-sharing complexity, language barriers, unequal resource allocation, and regional mistrust, and the corresponding solutions, including shared databases, harmonized protocols, and cross-border funding mechanisms grounded in trust [15]. These barriers and solutions generalize well beyond neurology.

Training the Next Generation Equitably

If you teach or mentor, the pedagogy literature offers a specific framework. Embedding international research within global public health education can strengthen critical thinking, cross-cultural understanding, and evidence-informed decision making, but meaningful integration is constrained by institutional, financial, technological, ethical, and political barriers, particularly within LMICs. The proposed response is an equity-oriented framework with three components: progressive curriculum integration, equitable partnerships, and reciprocal capacity strengthening [16]. For students, this means the ethical questions of partnership should appear in coursework, not only in the field, and students from LMIC institutions should experience research collaboration as producers of knowledge rather than as participants in someone else's project.

Frequently Asked Questions

What does an equitable partnership in global health research look like in practice?

It looks like a collaboration where the LMIC partner sets or co-sets the research question, holds or shares budget authority, occupies first and senior author positions, controls local data, and has governance rights spelled out in a signed agreement, following frameworks such as the six Brocher Declaration principles [9].

What is decolonizing global health research?

It is the effort to dismantle colonial-era patterns that determine who decides, leads, and benefits from research. Kumar and McCoy structure it across three dimensions, colonialism within global health, colonization of global health, and colonialism through global health, and translate it into six actionable areas for funders [1].

How do I know if a partnership is exploitative before joining it?

Check who wrote the proposal, who controls the funds and data, what the authorship policy says, and whether local leadership can articulate the project's rationale independently. The documented warning signs include externally framed priorities, HIC-controlled finances, and marginalized local intellectual leadership [3].

What is epistemic injustice in research collaborations?

It is the unfair devaluation of a partner's knowledge and expertise, for example when local clinical or community knowledge is dismissed in favor of HIC-derived methods. Researchers analyzing North-South data science collaborations in Africa used this concept to explain persistent sustainability and equity problems [6].

Who is responsible for research ethics in an international collaboration?

Everyone in the research ecosystem. A workshop report argues that ethical conduct is a shared responsibility of funders, governments, institutions, research teams, and ethics committees, with distributed ethical scrutiny replacing oversight concentrated in a single body [10].

What should an LMIC institution ask for in a partnership agreement?

Ask for budget authority with signatory rights, data ownership and storage terms, authorship criteria and target authorship positions, intellectual property arrangements, a fair dispute mechanism, and capacity commitments that survive the current grant, consistent with the sustainability principle of the Brocher Declaration [9].

How can funders shift power in global health research?

The published funder checklist names six moves: shifting power within the research ecosystem, prioritizing health equity in the portfolio, redressing historical underinvestment in research systems, addressing conflicts of interest, equitably distributing outputs, and reporting on outcomes and impact [1].

How can an early-career researcher start building equitable collaboration skills?

Begin by learning the frameworks, auditing collaborations you are already part of with a simple parity calculation, and seeking programs structured around local leadership and regional peer learning such as those described in the RESPIRE lessons, where mentorship and governance are explicit design features rather than afterthoughts [8].