A friend messaged me last Tuesday: “Cuba’s running out of insulin.” Not a headline. Just a fact, the way you’d say the milk is off. She was looking at fuel shortages shutting down hospital generators, waste piling up because collection trucks can’t run, epidemics spreading because there’s no diesel for ambulances. She wanted to know if I’d heard. I had not. What I heard instead was the sound of a system designed around one assumption — that supply chains continue — encountering the exact scenario where that assumption breaks.
The US oil blockade began in 1960 after the Cuban Revolution, when the US government sought to destabilize Fidel Castro’s regime through economic pressure. It has been in place, in various forms, for over sixty years. What changed in May 2026 was not policy but precision. A sudden halt in fuel shipments meant that diesel no longer arrived at ports. Hospitals lost power. Refrigeration failed. Medications requiring cold storage spoiled. Dialysis machines stopped mid-cycle. The cascade was predictable. It was also designed. An embargo is not a natural disaster. It is a policy choice with a body count that someone, somewhere, is willing to accept.
I have sat in hospital waiting rooms where the nurse says the scanner is down and we’ll have to reschedule. Mechanical failure, usually. Sometimes budget cuts. Once, during a heatwave, the whole building lost power and they sent everyone home. The assumption underneath: this is temporary. The machine will be fixed. The power will return. I have never sat in a waiting room where the assumption is: the machine will not be fixed because the fuel to run the hospital does not exist and will not exist tomorrow either.
There are two kinds of health system failure. The first is a gap — something breaks and gets repaired. A ventilator stops working. A new one arrives. The second is structural collapse. No ventilators arrive because the supplier no longer ships to your country. No generator starts because there is no diesel anywhere. The people who designed your health system did not design for this scenario because they assumed continuation. They built a system that works perfectly in a world where supply continues. That world no longer exists for people in Cuba.
This is not about disaster preparedness. Disaster preparedness assumes the disaster ends. What is happening in Cuba in 2026 is not a disaster. It is a permanent condition created by policy and sustained by the assumption that some populations are expendable inputs in someone else’s geopolitical calculation.
I keep thinking about a paper I read five years ago by Paul Farmer on structural violence. Farmer was a physician and anthropologist whose work examined how poverty and inequality are built into the systems we take for granted—how they kill not through direct force but through the ordinary functioning of institutions. He was writing about Haiti. The pattern was identical: a country under embargo, a healthcare system dependent on imported goods, chronic illness becoming acute crisis because the supply chain is a political tool. Structural violence means deliberate harm caused by social systems and institutions, not by direct force or individual action. The mechanism works like this: make a population dependent on external supply, then cut the supply. Call the resulting deaths inevitable. Farmer died in 2022, just as these patterns were accelerating globally. I wonder what he would have written about Cuba in 2026.
The chronically ill die first. Not because their conditions are more severe. Because chronic illness requires continuity. A diabetic needs insulin every day. Miss three days and you are in crisis. Miss a week and you are dead. The healthy person can survive a supply interruption. The chronically ill cannot. The system fails in order of dependence.
I have type 1 diabetes. I have never missed an insulin dose because I live in a country where pharmacies restock weekly and my prescription auto-refills. I also know exactly how many days I can survive without it. Four, maybe five if I stop eating carbohydrates entirely. After that, my blood sugar climbs past 20 millimoles per liter and my body starts breaking down fat for fuel, producing ketones, acidifying my blood. Diabetic ketoacidosis kills in hours if untreated. Treatment is insulin and IV fluids. Both require a functioning hospital.
In May 2026, Cuban hospitals were running generators on rationed diesel. Some had power four hours a day. Some had none. Insulin requires refrigeration between 2°C and 8°C. Above that, it degrades. It does not stop working immediately — it loses potency over days. You inject your usual dose and your blood sugar stays high. You increase the dose. It still does not work. You do not know if the insulin has failed or if your body has. By the time you know, you are already in crisis.
I asked my endocrinologist once what I should do if supply chains collapsed. She looked at me like I had asked what to do if gravity stopped working. “That won’t happen,” she said. She was right. It will not happen here. It is happening elsewhere, to people whose citizenship is the wrong one.
The autism researcher Simon Baron-Cohen published a study in 2003 arguing that autistic people have an empathy deficit. The study measured empathy by asking participants to identify emotions in photographs of eyes. It found that autistic people scored lower. What the study did not measure: whether autistic people could predict what another person needed to survive. That is a different kind of pattern recognition. It is also a different kind of empathy.
I can tell you exactly how many days a type 1 diabetic in Havana has left if the insulin supply stops. I can map the cascade: refrigeration fails, medication spoils, blood sugar spikes, ketoacidosis begins, hospitals cannot treat because there is no power for IV pumps. I can see the whole system from the inside because I live inside a version of it. The difference is that my version has not collapsed yet.
Baron-Cohen’s study treated empathy as the ability to read a face. It did not treat empathy as the ability to see a structure and know what happens when it fails. Those are not the same cognitive task. One is social. One is systemic. Autistic people are often better at the second one. We see patterns. We see dependencies. We see what breaks when the supply stops.
The humanitarian argument against the Cuba blockade is that it kills civilians. The utilitarian argument is that it does not achieve its stated political goals. Both are true. Neither addresses the mechanism: a health system is only as resilient as its least replaceable input. Make that input inaccessible and you do not need to bomb a hospital. You just wait.
There is a graph I think about often. It is from a 2019 paper on supply chain fragility in low- and middle-income countries. The x-axis is time. The y-axis is mortality. The graph shows two lines. The first line is mortality during a natural disaster — earthquake, hurricane, flood. Deaths spike, then decline as aid arrives. The second line is mortality during a sustained blockade. Deaths climb steadily. They do not decline. The line does not flatten. It just keeps rising because the condition is not temporary.
The paper did not include Cuba. It was published three years before the 2026 fuel crisis. But the mechanism was already visible. A blockade does not need to be total to be lethal. It just needs to target the one thing the system cannot replace.
I wonder how many people in Washington have insulin in their fridge right now.
This article was prompted by In Cuba, an unprecedented health crisis is deepening under US oil blockade from Le Monde English.