Bridging Mankind: Planetary Health

Social responsibility does not stop at the local and the national. It is global.

The physician’s approach is from the scientific direction. ESG, meaning environmental, social, and governance criteria, has become the apparatus through which large organizations are held to account for their environmental footprint, their social impact, and the integrity of their decision-making. One Health and Planetary Health supply the science and the moral case. ESG supplies the existing reporting machinery, the players already obliged to enact it, and the standing that comes with both. Physicians who wish to see planetary health commitments enacted rather than affirmed will find that machinery is where the enacting happens.

The Institutional Landscape, and Its Landmark Reports

Global health is not an unorganized field, and the physician who enters it should know the architecture and the documents that built it.

The World Health Organization, founded in 1948, remains the normative center, and the Declaration of Alma-Ata of 1978 its most consequential statement: health as a fundamental human right, primary health care as the vehicle, and the recognition that attaining health requires action across many sectors. Health for All by the Year 2000 was not achieved, and the Declaration of Astana restated the commitment in 2018. The WHO Commission on Social Determinants of Health produced Closing the Gap in a Generation (2008), whose three recommendations - to improve daily living conditions, tackle the inequitable distribution of power, money, and resources, and measure the problem – constitute the most direct institutional endorsement of Virchow's proposition ever issued by a UN body.

The World Bank's Investing in Health (1993) introduced the disability-adjusted life year. The Lancet Commission's Global Health 2035 argued that a grand convergence in outcomes was achievable within a generation. And the Rockefeller Foundation-Lancet Commission gave planetary health its founding text in 2015. Alongside these sit the Sustainable Development Goals, UNICEF, UNAIDS, the Global Fund, Gavi, and CEPI, and at the professional level the World Medical Association, whose Declaration of Geneva is the modern restatement of the physician's oath.⁶⁴

One category of rule-making cuts across all of these institutions, but has no settled home in any of them. AI will demand rules that are technical, clinical, legal, and cultural at once. The hard question is which of them should be universal and which must remain responsive to culture and local norm, since the answer differs by domain. Standards of evidence and safety travel, while norms governing disclosure, family involvement, end-of-life decision-making, and the acceptable uses of a simulated presence do not, and imposing one jurisdiction's settlement globally is itself an equity failure.

No single body answers the questions that follow: what governs acceptable use; what standards apply across settings; what must remain local; and how accountability is assigned when a system crosses a border. Physicians have an obvious stake and an unusual qualification here, being the participants who see the consequences at the level of individual people, and they should be active participants in shaping universal and culture-appropriate rules for biomedical AI rather than recipients of rules written elsewhere.

Several themes recur across the aforementioned documents with striking consistency: health as a right; primary care as the highest-return organizing principle, repeatedly affirmed and repeatedly under-resourced; intersectoral action, since the determinants lie mostly outside the health sector; universal health coverage as the financing goal; equity within as well as between countries; country ownership over donor direction; and health security as a shared global good whose failure in one jurisdiction is a failure everywhere.

Where these institutions have fallen short is equally consistent. Financing is voluntary, earmarked, and unpredictable, so agencies are accountable to donors. Authority is normative instead of binding. Governance is fragmented across scores of initiatives with overlapping mandates. Vertical disease-specific programs have repeatedly outcompeted the horizontal primary care systems every declaration endorses. And the response cycle is reactive, with attention and money arriving with the emergency and receding before the capability is built.

The physician of 2050 can address these shortfalls in ways that are specific. Against donor-driven agendas, physicians constitute the one global professional community with a direct, non-financial line to patients in every country, and their societies can assert priorities funders have not chosen. Against fragmentation, clinicians experience the incoherence of parallel programs at the point of delivery and can document it credibly.

Against weak enforcement, the profession's contribution is evidentiary and reputational, which is Lown's instrument applied to compliance and transparency. Against the concentration of authority, the practical lever is the leadership of research and programs, which physicians control through the partnerships they enter. And against the chronic under-resourcing of primary care, the profession can stop treating generalism as a residual career and start treating it as the role that moves most in global health, which is what every one of these reports has said for nearly fifty years.

Two further commitments concern the point at which innovation and equity meet. The first is that equitable access requirements be embedded in global biomedical innovation frameworks from the earliest stages of discovery, not negotiated once a product exists and its price has been set. Access, licensing, affordability, and implementation would be designed together, since a technology that is available and unaffordable has not been distributed.

The second is that international governance systems be developed with the specific competence to oversee AI-enabled biomedical innovation, which crosses borders faster than any therapeutic and is at present governed almost entirely at national level. Universal access models capable of reducing disparities in advanced diagnostics and therapeutics worldwide are the stated goal. The test is whether the mechanisms for reaching them are written into the design of an innovation or appended to it afterwards.

Alongside both runs the proposition that care escapes geography and the physician goes global. Care without an address, remote monitoring, and networked expertise and data analytics extend care across borders, narrowing the gap in low- and middle-income countries and making equity and access foundational design questions.

Leapfrogging, Not Catching Up

The default assumption, that developing countries will traverse the same sequence the developed world traversed and arrive later at the same destination, is probably wrong, and the assumption itself is an obstacle. Infrastructure that was never built is more than an absence. It is the absence of legacy, and legacy is what makes incumbent systems slow. Telecommunications made the point already: much of Africa skipped copper landlines entirely and went to mobile, then skipped retail banking and went to mobile money.

Health is now producing the same pattern, and several of the moves are demonstrated rather than speculative.

Consider logistics over roads. Rwanda's national drone delivery of blood and vaccines cut delivery times from hours to minutes and sharply reduced wastage, and the enabling move was regulatory, a performance-based airspace framework adopted faster than any incumbent aviation authority could have managed, with Ghana, Nigeria, and Malawi following.

That pattern repeats across the other demonstrated moves. Countries with no installed base of proprietary hospital systems can adopt open, interoperable, patient-held records directly, avoiding the fragmentation that high-income systems will spend a generation undoing. Task-sharing designs care around community health workers supported by decision tools, which is not a stopgap but a more scalable model, already being re-imported as reverse innovation. Diagnostics at the point of need reach populations the facility-based pathway never would. And mobile money rails enable health financing and conditional transfers without a retail banking system, a health application riding on a leap already completed in another sector.

Several further moves are plausible and not yet realized, and they are where the next decade's advantage may lie. One is regulatory leapfrogging, in which small, agile jurisdictions become the evidence generators that larger systems follow. Another is population-scale genomics designed correctly on the first attempt, in populations of greater genetic diversity than those on which current reference data rest. A third is preemptive care as the default, since a system not already organized around episodic acute treatment has less to dismantle. The others are climate-adapted infrastructure built for the conditions of 2050, not retrofitted from those of 1970, and South-South transfer as the primary channel, in which innovations move directly between comparable systems.

Two conditions govern whether any of this materializes. The first is local validation, since a model trained in one epidemiological and demographic environment may fail silently in another, and the failure mode is the more dangerous for being invisible. The second is local ownership, of data, of intellectual property, and of the research agenda, without which leapfrogging becomes a new form of dependency conducted at higher technical sophistication.

Local ownership of intellectual property is the condition most often stated and least often operationalized, and the instruments for acting on it sit far upstream, where a university decides what to patent and how to license it. Barry Coller has argued that the equity lens belongs at the earliest phases of translation, in basic research and first-in-human work, because the terms agreed while a discovery is still preclinical determine what manufacturing, pricing, and distribution become possible a decade later. By the time a product exists and its price has been set, the decisive choices have already been made.

The levers are specific. Whether to file for patents in low- and middle-income countries at all, and whether to enforce them if filed. Whether to license exclusively or non-exclusively, and whether to reserve the right for academic and governmental investigators anywhere to use the technology at cost. Whether to write in tiered, at-cost, or royalty-free terms for territories the licensee does not intend to serve, and to require an affordable access plan naming those territories, as UCLA has done since 2020. Whether to tie continued exclusivity to development milestones, so that a shelved technology returns to the university. And whether to oblige a licensee to sublicense downstream to the Medicines Patent Pool, which reaches more manufacturers over larger territories than bilateral arrangements do. Coller sets out ten such questions for inventors to put to their technology transfer colleagues, and every one of them is answerable before there is anything to negotiate over.⁶⁵

The record shows how much distance remains. The Nine Points to Consider in Licensing University Technology, endorsed by more than a hundred institutions since 2007, measurably increased the reservation of research rights for nonprofit and governmental use, while the provisions bearing on public health and equitable global access changed licensing behavior far less. The United Nations High-Level Panel on Access to Medicines reached a parallel conclusion about publicly funded research in 2016, and the trilateral study of the World Health Organization, the World Intellectual Property Organization, and the World Trade Organization remains the standard account of how health, trade, and intellectual property policy intersect.

Manufacturing capacity is the second half of the problem and is being addressed slowly, since small molecules transfer readily while biologics and vaccines require know-how, quality documentation, and trained personnel to move with the license. For the physician of 2050 the implication is participatory. Faculty inventors and the clinical investigators who validate what they invent are the parties with standing to ask for these terms and the only ones in the room when the terms are set, which makes technology transfer policy a legitimate object of professional attention rather than an administrative matter settled elsewhere in the building.⁶⁶

The absence of legacy is what permits a jump, and it is also what removes the things imported technology quietly depends on: local validation data; an installed technical workforce; and the institutional habit of noticing when a tool has begun to fail. A health system with no legacy record architecture can adopt an open one directly, and it has no historical dataset against which to check whether a model performs on its own population.

Other factors are obviously to be considered. Technologies dependent on data, infrastructure, connectivity, and specialized personnel can widen inequity unless access is designed in from the start. Models trained in one environment may fail in another. High-income systems should not assume that innovation flows one way, since some of the most important advances in extending care will be invented under constraint and exported outward.