A grants administrator at a US university is staring at a submission deadline eleven days out. The call requires a named co-principal investigator in the country where the study will actually take place, a district in Malawi with a public hospital, a busy pediatric ward, and a physician the funder wants leading real scientific decisions, not just signing a letter of support.
The principal investigator has been to that hospital exactly once, three years ago, on a two-week site visit organized by a colleague who has since left academia. He does not have a current contact there. His department's usual approach, emailing whoever answered a similar request the last time, has not worked; the person who signed the last collaboration letter left the hospital for a different post eighteen months ago and nobody forwarded the message.
He is now eleven days from a deadline, with a study that genuinely needs a qualified local co-investigator with pediatric ICU access, real decision-making authority, and no way to verify that any name he might find through a cold search actually has the standing, the time, or the interest to do more than lend a signature.
He will very likely end up doing what happened three years ago: find someone willing to be listed, without the partnership behind the listing, and produce exactly the kind of paper the field now has a name for and a growing list of reasons to reject.
The physician who could actually co-lead this study, with the access, the standing, and the genuine interest, exists somewhere in that district right now. Nothing connects the two of them except luck and a deadline that does not allow for it.
The numbers behind the pattern
Start with how consistently this shows up across the literature, because it is not a single bad case, it is the default outcome.
Only 6.5% of papers in general medical journals have a coauthor from the country where the study population lives, a figure from a 2002 baseline that has anchored two decades of subsequent tracking. Move forward to 2016 and fewer than half of infectious-disease publications originating from Africa had an African first or last author. By 2026, in musculoskeletal-injury research specifically from Malawi and Tanzania, high-income-country institutions still held 59% to 69% of first and last authorships, more than fifty years into the modern era of international research collaboration.
The COVID pandemic, which should have been the clearest possible case for genuine local leadership given how directly it depended on local clinical and epidemiological knowledge, produced the same pattern at speed: one in five African COVID-19 papers had no African author at all, and 66% of authors on African COVID research were non-African. In some fields the skew is even sharper: publications from developed nations studying certain topics may have as low as 3% African authorship.
This is not confined to one region or one disease area. In enteric-virus research, the top five publishing countries produce 92.4% of the literature, while sub-Saharan Africa and Latin America together account for just 7.6%, despite carrying a disproportionate share of the actual disease burden the research is meant to address.
The consequences are now landing on the researchers who ignore this
For a long time this was framed almost entirely as an ethics problem, worth naming but without much operational teeth. That has changed. Journals have started converting it into a submission-blocking practical problem.
PLOS Medicine now requires local first or last authorship for papers drawing on data from the study population's country. The Lancet Global Health will reject papers with African data that fail to acknowledge African collaborators. EDCTP3, one of the major funders of European-African clinical research partnerships, mandates African leadership as a structural condition of funding, not a preference.
This turns what used to be an ethical lapse into an acute operational failure for the HIC investigator: a paper rejected for authorship inequity costs roughly a year, and a multi-country grant application that fails at partner identification loses its entire submission window. The financial stakes are not small. Fogarty, Wellcome, EDCTP3 and the Gates Foundation collectively fund several billion dollars of LMIC-sited research annually, and an increasing share of that funding now requires exactly the kind of named, real, verifiable local co-investigator relationship the grants administrator in the opening scene could not produce on eleven days' notice.
Why the current workflow keeps producing the same failure
The typical path to a collaborator, as it actually happens today, explains why the outcome is so consistent.
A US or European PI emails a former fellow, or a name that appears on a prior paper from the region, or reaches out through whichever personal contact happens to still be reachable. That contact is frequently a hospital administrator, or a single, already-overloaded academic, rather than the clinician actually positioned to lead the science. Diaspora clinicians, who trained in the target country and now practice in the HIC system, and who are structurally the strongest possible bridge between the two sides, are almost never asked. Local investigators, meanwhile, wait for an approach that only ever comes through prior HIC contacts, because they have no independent, discoverable channel of their own to signal interest, capacity, or availability.
And authorship gets negotiated, if it gets negotiated at all, at the point of submission, far too late for it to reflect genuine co-leadership in the design and conduct of the study rather than a courtesy credit attached after the fact.
The structural failure: no discoverable profile, no accountability for how partners were actually treated
Underneath the workflow failure is an infrastructure gap on both sides of the relationship.
LMIC clinicians have no discoverable, verified research profile. Institutional websites in many settings are sparse or outdated. ORCID adoption, the closest thing to a global researcher identifier, is uneven and skews toward researchers already embedded in international publishing networks, which is precisely the population least in need of being discovered. A qualified pediatric ICU physician in a district hospital in Malawi, exactly the person the grants administrator needed, may have no searchable digital footprint that would surface her to a stranger eleven time zones away on a deadline.
Trust, as a result, runs entirely through personal HIC-to-LMIC chains: whoever the PI happened to meet on a prior trip, whoever a departing colleague happened to introduce before leaving academia. And there is no record, anywhere, of which HIC partners actually honored their authorship and data-return commitments on a prior collaboration and which did not. Exploitation, where it happens, simply repeats, because the local investigator community has no shared, verified way to know in advance which HIC groups have a track record worth trusting, and HIC investigators have no way to signal, credibly, that they are one of the good-faith partners.
The diaspora bridge, physicians who trained in the target country and now practice in HIC systems, is structurally the strongest available solution and remains almost entirely unmapped. These are people with genuine standing on both sides: trusted by colleagues back home, credentialed and embedded in the HIC research system, and, in many cases, personally motivated to see equitable research done well in the country they trained in. Nobody has built the graph connecting them systematically to either side.
Why nobody owns this
ResearchGate and LinkedIn profiles are self-reported and skew heavily toward researchers already visible in HIC-dominated networks, which is the opposite of who this problem needs to surface.
ORCID indexes outputs, not willingness or track record. It can tell you a researcher has published. It cannot tell you they are available for a new collaboration, or that a prior HIC partner delivered on what was promised.
Funders hold partner lists, but only per grant, siloed to their own portfolio, not shared across the field in a way that would let a new PI search broadly.
Societies like ASTMH and CUGH are membership organizations, not matching engines. They connect people who already know to look for each other. They do not solve the discovery problem for the PI who does not yet know who to look for.
Nobody currently in this market is paid to build the two-sided verified trust graph: who in a given district has the clinical access and research capacity, and which HIC groups have a documented, honest track record of honoring authorship and data-return commitments. That absence is why the same pattern, a cold email to whoever is still reachable, followed by authorship negotiated too late, keeps recurring even as journals and funders raise the stakes for getting it wrong.
What would actually work
A verified LMIC clinician-investigator map, built for discoverability rather than self-promotion. Credentials, clinical access, research interest and specialty, verified rather than self-reported, so a PI eleven days from deadline can search rather than guess.
A partnership conduct record attached to HIC institutions and PIs, not just to individual grants. Whether a given group honored authorship commitments and returned data as promised on a prior collaboration is exactly the information a prospective LMIC partner needs, and currently has no way to check.
The diaspora bridge mapped deliberately, not left to chance. Physicians who trained in a given country and now practice in the US, UK or Canada are the highest-trust possible introduction on both sides, and are the most obviously underused resource in the entire problem.
A structured partnership covenant, agreed before data collection begins, not negotiated at submission. Authorship order, data-return terms and publication timeline settled up front, aligned with existing frameworks like the TRUST Code and BMJ Global Health's authorship guidance, rather than improvised under deadline pressure at the end.
An equity index that funders and journals can actually use. A per-institution, per-funder measure of authorship equity over time, built from both bibliometric data and member-reported partnership conduct, would give journals and funders a real signal instead of relying on self-reported reflexivity statements alone.
Built to work for community and public-hospital LMIC clinicians, not only academically affiliated ones. The majority of the target population practices in public hospitals with no research office at all, which is exactly the population least visible to any system built around institutional websites and ORCID profiles.
What you can do now
If you are an HIC investigator planning LMIC-sited research
Start the partner search a year before the grant deadline, not eleven days before. The single biggest driver of the pattern described here is time pressure forcing a PI back to whoever is already reachable, rather than who is actually best positioned to co-lead.
Ask your own diaspora colleagues directly whether they can introduce you to someone in their home country. This is the single highest-trust channel available and it is almost never used deliberately; it is treated as a personal favor rather than a standard part of study planning.
Settle authorship order and data-return terms in writing before data collection starts, not at the point of manuscript submission. Journals increasingly require this to be demonstrable, and doing it early is also simply the right thing to do.
If you are an LMIC clinician-researcher
Build whatever public research profile you can, even a minimal one, because the current discovery failure runs in both directions: HIC investigators frequently cannot find you even when they are looking in good faith.
Ask a prospective HIC partner, directly, about their track record on authorship and data return with prior local collaborators. There is currently no independent record to check, which means asking directly, and asking for names you can call, is the only verification tool available.
If you fund or govern global health research
Require a documented partnership covenant, settled before funding is released, not merely a letter of support at submission. EDCTP3's African-leadership mandate is a meaningful structural step; extending a similar requirement, with real verification, to other major funders would change incentives across the field quickly.
Fund a shared, cross-funder verified investigator map rather than each maintaining a siloed partner list per grant. The duplication of effort across Fogarty, Wellcome, Gates and EDCTP3 grantees each independently searching for the same small pool of known collaborators is itself a measurable waste this problem creates.
Frequently asked questions
What is helicopter research? Also called parachute research, it describes high-income-country investigators conducting studies in low- and middle-income countries with little or no meaningful local co-authorship or leadership, extracting data and samples without building lasting local research capacity or credit. Only 6.5% of general medical journal papers have had a coauthor from the study population's country, based on a widely cited 2002 baseline.
What percentage of African research papers have African authors? It varies sharply by field and era but is consistently low: fewer than half of infectious-disease publications from Africa had an African first or last author as of 2016, and one in five African COVID-19 papers had no African author at all, with 66% of authors on African COVID research being non-African.
How do I find a qualified research partner in a specific LMIC country? Currently, mainly through personal networks: former fellows, prior collaborators, or diaspora colleagues who trained in the target country. No verified, searchable directory of LMIC clinician-investigators currently exists, which is the structural gap this article describes.
What is the Lancet Global Health's policy on African authorship? Since 2018, the journal has stated it will reject papers using African data that fail to acknowledge African collaborators appropriately, part of a broader shift among major journals (including PLOS Medicine, which requires local first or last authorship) toward enforcing equitable authorship as a submission requirement rather than an ethical suggestion.
What does EDCTP3 require for African leadership in research? EDCTP3, a major funder of European-African clinical research partnerships, mandates African leadership as a structural condition of its funding, reflecting a broader trend among major global health funders (Fogarty, Wellcome, Gates) toward requiring named, verifiable LMIC principal investigators rather than accepting a local site as a data-collection venue only.
What is the TRUST Code? It is a set of published principles for equitable research partnerships between high-income and lower-income country institutions, covering fairness, respect, care and honesty in collaboration, referenced increasingly by funders and journals as a benchmark for what a genuine, non-exploitative partnership covenant should include.
The bottom line
The grants administrator in the opening scene is not choosing to run an exploitative study. He is eleven days from a deadline with no verified way to find the physician who is, right now, running the pediatric ICU his study needs, has real standing with her colleagues, and would very likely be genuinely interested in co-leading the science if anyone could reach her in time. The system he is operating inside gives him exactly two options: find someone willing to be listed, or miss the deadline. Neither option is the one journals and funders are now demanding.
The numbers show this failure is not an occasional lapse. It is close to the default outcome of cross-border research as currently practiced: 6.5% local coauthorship at baseline, under half of African infectious-disease research carrying African leadership as of 2016, one in five African COVID papers with no African author during the exact crisis that most needed local expertise. Journals and funders have started converting the ethical failure into a practical, deadline-costing one, which is real pressure. It has not yet been matched with the infrastructure that would let a good-faith investigator actually succeed under that pressure.
The diaspora physicians who could bridge both sides of this relationship, trusted at home, credentialed abroad, are sitting inside the same professional networks this entire series keeps finding underused. Somewhere in the district the grants administrator needs, a physician with exactly the access and standing his study requires is going about her evening shift, unaware that a study she would want to co-lead is eleven days from either finding her or finding someone else instead.
Part of a series on the missing professional infrastructure of healthcare. Previously: The Surgeon-Inventor Dead End
Evidence note: the 6.5% general-medical-journal coauthorship figure and the sub-50% African infectious-disease authorship figure are from a 2022 PLOS Medicine editorial synthesizing prior bibliometric work; the original underlying studies were not independently re-verified in this piece. The 2026 Malawi and Tanzania musculoskeletal-injury authorship figures (59% to 69% HIC first and last authorship) are from a single 2026 Bone and Joint Open study covering one clinical subfield in two countries, not a global sample. The one-in-five African COVID-authorless-papers and 66% non-African-authorship figures are from a 2024 Ghana Medical Journal analysis; the "as low as 3% African authorship" figure describes a specific worst-case subset rather than a typical rate. The enteric-virus publication concentration figure (92.4% from top five countries) is from a single 2025 Pathogens study of one disease area and should not be generalized to all global health research without caution. Funder budget figures for Fogarty, Wellcome, EDCTP3 and Gates are approximate, dossier-level aggregates rather than independently audited totals. This article does not describe or endorse any specific partnership platform; the TRUST Code and BMJ Global Health authorship guidance referenced are existing, independently published frameworks.
