Twenty-nine weeks. Severe-range blood pressures she had not been carrying when she walked in. The cuff at triage had read in the high one-fifties; the cuff fifteen minutes later, after the nurse had her sit, had read two-twenty over one-fifteen. The headache she had been describing to her partner since the night before was still there. The corner-of-the-eye visual changes were still there. She had driven in. Her partner was in the chair the nurse had pulled up next to the gurney.
The nurse is on the second attempt at an IV. The first arm did not yield. The second arm is not yielding. The patient's vessels, at these pressures, are doing what vessels do at these pressures — they are clamping down. The peripheral vasculature in severe-range hypertension is constricted to the point where the blue lines a nurse can normally find by feel are not findable. The vein the ultrasound finds, the second time, blows when the catheter advances. The nurse pulls the line and looks at me.
I tell her to call anesthesia. I tell her, before she leaves to call, that I want IM mag in the room while we are waiting for them to come up. The nurse looks at me again. She is a good nurse. She has been on this unit for six years. She says, slowly, that she is not sure we have it. She says she is not sure she knows what IM mag is.
The four-gram bolus of magnesium sulfate that I order for a patient with severe preeclampsia is, in the standard order set, an intravenous bolus. We dissolve it in one hundred milliliters of normal saline and we push it over fifteen to twenty minutes through the line a nurse has placed. That is the order set I write every shift. It is the order set the nurses run on every shift. The intramuscular form of magnesium — five grams into each buttock, ten grams total — is what magnesium sulfate is when there is no line. It is the form the drug took in the original international trials. It is the form the drug takes in field hospitals, in rural clinics in countries with a fraction of our health expenditure per capita, in any setting where a peripheral IV is not something you can presume. It is the form that exists precisely because magnesium sulfate is not supposed to be a drug that depends on infrastructure. It is supposed to be a drug you can give anywhere there is a syringe and a body.
The reason mag sulfate is the global standard — the reason the Magpie Trial, with twenty-six countries and ten thousand women, was the trial that established it — is that it works in places where almost nothing else works. The trial was designed so that a hospital in a small city in Pakistan and a hospital in a small city in the United Kingdom could both administer the same drug, the same dose, the same way, and prevent the same seizures. The IM form is the form that made that universality possible. The IM form is what magnesium is for.
I tell the nurse to call pharmacy while she is calling anesthesia. The nurse calls. The patient's blood pressure cycles again. Two-eighteen over one-thirteen. She is closing her eyes against the headache. Her partner is holding her hand on the side of the bed where the line is not.
The pharmacist on the phone, when the nurse hands it to me, has been a pharmacist at this hospital for a long time. He is good. He is calm. He tells me that he is looking. He tells me, after a pause that is long enough that I count it, that he does not see magnesium sulfate in the intramuscular formulation in the building. He tells me he has not, in the time he has been the pharmacist here, ever had a request for it. He asks me if I can wait, on the line, while he calls the manufacturer's regional rep. He says — and this is the sentence that sits with me — "I'm not sure who in the country still makes the IM form. Let me find out."
I am not sure either. I am, in the operational moment, supposed to know. I have a patient with severe-range pressures who is at this minute closing her eyes against a headache that is the warning before the seizure. The intramuscular form of the drug that exists precisely so that her seizure can be prevented in any building on earth is, on this Wednesday afternoon, possibly not made by anyone in this country anymore. The pharmacist is, on the line, trying to find out. I do not have the time to wait for him to find out. I tell him, gently, that I do not have the time to wait. I tell him to keep looking and to page me. I hang up the phone.
Anesthesia walks in. The attending is someone I have worked with for a long time. She does not need to be told what is happening. She places an ultrasound-guided line, in the upper arm, in three minutes. The mag bolus runs. The labetalol runs. The headache, fifteen minutes later, is the smaller headache that comes when the cerebral edema starts to retreat. The patient does not seize. The transfer to the tertiary center, with the higher-level NICU, is arranged. The partner, still holding her bag, helps her into the ambulance an hour later.
The pharmacist pages me, much later, after the patient is gone. He tells me what he found. I will not put the specific manufacturer information in writing here. What he told me was a version of the sentence he had said on the phone, which was that the supply chain for the IM formulation of magnesium sulfate — the formulation that was, in the original trials, the form that made the drug universal — is, in the United States in 2026, a chain so thin that the pharmacist of a hospital that delivers thousands of babies a year has never had occasion to order it, and was not certain, until he called, whether anyone was still making it. The sentence sat with me. The sentence is still sitting with me.
I think a lot about Xi Jinping concluding, sometime in the last few years, that America is a declining power. I think about the podcast where a journalist explained that the read in Beijing now, the read in capitals where they pay close attention to us, is that the empire is on its way down. I do not, as a clinician, have a foreign policy. I do, as a clinician, have a list of things I have seen happen in American hospitals in the last five years that I do not believe could be happening in a country at the peak of itself. The intramuscular form of magnesium sulfate not being in the building is one of them. A fifteen-year-old being raised by the nurses of an emergency room because there is no psychiatric bed in the state is another one. A pregnant woman driving herself to triage because the outpatient clinic that should have caught her pressures a week ago no longer answers the phone is another one. The list is long. The list is, in the way of these things, the thing that makes the declining-power read feel right from inside the building.
It is an embarrassment that has the texture of a particular kind of embarrassment — the kind you feel when your mother hears that the priest came by the house and you did not offer him anything to drink. There is shame in it that is not exactly about you. It is about what your house is supposed to be. It is about what was supposed to be true in this house, in this country, in this hospital. The richest country in the world is supposed to have, in the building where it delivers its babies, the drug that the field hospitals have. That is the floor. We are below the floor.
What I can do, in the encounter, is what the encounter required, which is the page to anesthesia and the ultrasound-guided line and the bolus that ran when the line was finally in. What I cannot do, from inside the encounter, is fix the supply chain that decides whether the backup to the backup is in the building. The standard order set is the IV bolus because the standard order set presumes the line. The IM rescue is what the textbook teaches you to reach for when the line is not there. The textbook presumes the rescue is in the room. In the operational reality of 2026 the rescue is — in some hospitals, on some afternoons, including this one — not in the room.
The same logic, in a slightly different shape, is what we are also living through with oxytocin. When the oxytocin supply tightens, the operational consequence is not abstract: it is that we cannot induce labor on the schedule the clinical situation calls for. When you cannot induce a preeclamptic who needs to be delivered, you do not get to wait. You operate. The C-section rate goes up not because the obstetrics changed but because the pharmacy did. We do not want to live in a world where surgical delivery rates climb because the cheapest oral and intravenous induction agents in the formulary are not reliably in the building. We are, in some buildings, on some afternoons, in that world.
What I do, after the patient is gone, is the thing I do most weeks now. I write the case down. I tell my colleagues. I have a long conversation, that afternoon, with an emergency medicine attending who has been working in this region as long as I have. We talk for an hour. We are not, in the conversation, surprised. We are both, in the conversation, sad in the particular way that people who have given their adult lives to a system are sad when the system stops doing the thing it was built to do. He tells me, at the end of the conversation, that he has stopped being shocked. He tells me he stopped being shocked the year a fifteen-year-old lived in his emergency department for six months because there was nowhere for her to go. I tell him I have not stopped being shocked yet. I am not sure which of us is in the better position.
The point of this story is not the patient. The patient was, in the end, fine. She did not seize. Her transfer went smoothly. Her baby, born some days later, was in the NICU for a stretch and then went home. The point of this story is that her not seizing depended on the anesthesia attending walking through the door and finding a vein with an ultrasound — a workaround for a workaround that should not have been the second-line plan. The first-line plan is the line. The second-line plan is the muscle. The second-line plan is what is supposed to exist when the first-line plan fails. The second-line plan was, on this Wednesday afternoon, not in the building.
When the medicine has limits, the care does not. The care, this week, was the page to anesthesia and the bolus that eventually ran. The work — the work that does not happen at the bedside — is making sure that the next patient, on the next afternoon, in the next hospital, has the second-line plan in the room with her.