When the Iraq hospital we visited couldn't answer "what do you need?", the working assumption was that the gap was decision-making infrastructure — a framework problem. The conversation behind this case study suggests the gap may run deeper than that.

Earlier this year I had the chance to sit down with a humanitarian pediatrician who has spent more than two decades working in some of the hardest health environments in the world. He has worked across the Middle East — including in Gaza, Syria, and across the Syrian diaspora in Jordan, Lebanon, and Greece — as well as in Mexico, Haiti, Tanzania, Bangladesh, Bolivia, Colombia, and Honduras. Most of his career was spent as a pediatrician at a major public hospital in Chicago before he turned to humanitarian work full-time. He is a co-founder of an international medical relief organization that operates across more than ten countries.

He shared a number of insights that have stuck with me, and a few of them ended up reshaping how this framework is built. The case is worth laying out in his own framing.

Access is a system problem, not an infrastructure problem

His central point was simple. Access to care is not solved by building more hospitals. It is solved by fixing the underlying system that determines who gets care, where, when, and from whom.

His framing: a hospital can sit fully built and fully staffed and still fail to deliver care to the people who need it most, because the system around it routes patients incorrectly, reimburses providers for the wrong things, or fails to get the right specialty in the right place. Infrastructure improves the smoothness of care delivery once a patient is in the system. It does not, on its own, expand the population that the system actually reaches.

His example: treating fractured hips in Iraq. The infrastructure problem is real — orthopedic capacity is uneven and equipment is aging. But the system problem is bigger. Patients don't reliably reach the right facility. Specialists don't reliably reach the patients. Even where capacity exists, follow-up care, rehabilitation, and discharge planning are fragmented. A new orthopedic ward in the wrong hospital, attached to a system that can't route patients to it, doesn't change outcomes.

The DRG analogy

To illustrate how much policy and incentive design shape care delivery, he pointed to the introduction of Diagnostic Related Groups (DRGs) in the U.S. in the 1980s. Before DRGs, hospitals were paid per day of patient stay, which created a clear incentive to keep patients hospitalized longer than necessary. After DRGs introduced flat per-condition payments, the incentive flipped — hospitals were rewarded for efficiency and appropriate site of care. The shift moved enormous volumes of care from inpatient to outpatient settings, often with better outcomes and lower costs.

None of that change was about infrastructure. The hospitals were the same. The doctors were the same. What changed was the system that surrounded them — the rules about who got paid for what.

His implication: in any health system, the highest-leverage interventions often live at the policy and incentive layer, not the physical capacity layer. Infrastructure investment can amplify a well-designed system. It can also reinforce a poorly designed one.

Gaza — what infrastructure means when there is none

The conversation took its most pointed turn when he described his experience in Gaza, where he has worked twice during the current conflict.

By the time of his most recent visit, the infrastructure question had collapsed. There were, in his description, no hospitals left standing in any meaningful operational sense. Clinical decision-making was constrained not by what equipment was available but by what the environment permitted at all.

He gave one example that stuck. A child arrives with a serious limb wound. In a higher-functioning system, the standard of care is reconstruction — preserve the limb, manage infection risk through clean facilities and adequate antibiotics, restore function over months of follow-up. In Gaza, the standard of care collapses to amputation. Not because amputation is the right clinical choice, but because the environment cannot support reconstruction. Infection risk is too high. Follow-up care doesn't exist. The decision is made not on the merits but on what the environment permits.

The downstream cost of those amputations falls disproportionately on children. A growing child who loses a limb requires repeated prosthetic adjustments throughout development. Each adjustment requires a working orthopedic system, trained prosthetists, and access to components. None of that exists at scale in Gaza or for many of the children displaced from it.

His estimate: roughly 5,000 children in the region currently need prosthetic care. It is one of the largest single categorical medical needs in the region — and one of the most underserved.

The post-conflict amputation gap as an investment opportunity

This is where his critique flips into something concrete and actionable. The amputation problem isn't a technology problem. The components exist — including 3D-printed solutions that have come down dramatically in cost. Orthopedic surgeons exist. Prosthetists exist.

The constraint is integration. The pieces are scattered across organizations, geographies, and funding streams that don't talk to each other. A scaled, coordinated post-conflict prosthetic program for children would require less invention than orchestration. It is the kind of intervention that could be funded — by a family office, an impact fund, or a coordinated philanthropic effort — and produce measurable, life-changing outcomes within a defined population.

This is, in other words, an example of exactly the kind of investment the framework should be helping people identify. Not "build a hospital in a conflict zone." Build the connective tissue that lets existing capacity reach the children who need it.

The prioritization framework he uses

When asked how he decides where to focus, he pointed to a framework borrowed from the Coefficient Giving Group, which originated out of GiveWell's data-driven approach to philanthropy. It evaluates any potential intervention along three dimensions:

The post-conflict prosthetics example scores well on all three. It affects thousands of children, deeply, with lifelong consequences. It is markedly underfunded relative to its scale. And the constraints are coordination and capital, not science — which means money applied well can move the outcome.

Most large hospital construction in conflict zones, by his read, scores poorly on tractability. The constraint is not the building. The constraint is the system. Adding capacity to a system that cannot operate the capacity it has is unlikely to change outcomes.

What this means for the framework

This case study is the reason the framework now treats policy and incentive alignment as a first-order axis, alongside capital allocation and sophistication. The earlier two-axis version of this framework (capital × sophistication) couldn't distinguish between two countries with the same capital and sophistication but radically different incentive environments. That gap is what this conversation forced into the open.

The matrix is still useful for one thing: it tells you what kind of investment fits a country's context. A red-quadrant country needs different investment than an orange one. That part holds.

But every individual investment, regardless of cell, should pass a separate test — the one this conversation surfaced. Does this specific intervention improve quality of care delivery for actual patients? Not "does it add capacity," but "does the capacity it adds reach the people who need it, in a system that can use it well?"