A decade in the country, fluent in the language, fully inside the system. That's the part of this one that stays with you.
Not a tourist. Not somebody who landed last month and got confused by the forms.
Right. A listener we're calling David, in Jerusalem, his name changed, his account checked before it reached us. He and his wife moved here about ten years ago. Citizens, Hebrew speakers, they know how the health system is supposed to work. And twice in two months, their sixteen-month-old son couldn't breathe properly, and twice they ended up somewhere that was supposed to catch him before the hospital.
The first time, the doctor had already given Ventolin. It got worse anyway. They went to one of the big nationwide urgent care chains, the son was examined, and from there he was referred on to pediatric emergency.
A month later, same thing. Faster breathing, labored, Ventolin not touching it. Back to the same place. A triage nurse puts a pulse oximeter on him, gets ninety percent, and tells the parents to give him oxygen. Their toddler is terrified and keeps tearing the mask off his face, and they can't do it alone, and nobody comes to help them. Twenty minutes, no doctor lays eyes on him. He vomits. And David's not asking us whether the staff were rude.
He's asking whether Israel's healthcare is starting to come apart at the level below the hospital emergency room. And if it is, why nobody seems alarmed.
One thing before we start, because the whole episode sits on it. The hospital care was good. He says so plainly. Competent, attentive, empathetic, and he's grateful. It's the layer underneath that failed him. So let's look at that layer.
Here's what that layer is supposed to be. Four health funds. Clalit, Maccabi, Meuhedet, Leumit. They're the insurers, but they're also your first point of contact for most things, primary care included. Each one runs a twenty-four-hour nurse-staffed telephone line, because the intention is that you call somebody before you go anywhere. Then there are after-hours clinics, and then there's the urgent care network, and the biggest of those is Terem.
Twenty-five branches, Nahariya down to Eilat.
Twenty-five, about a million patients a year. Their own language is that the point of the model is to reduce hospital loads. They claim it cuts emergency department visits by as much as forty percent. That's not a modest side benefit. That's the entire justification for the layer existing.
So it isn't a luxury. It's load-bearing. Take it out and the weight goes somewhere.
It goes to the ER, and the ER is the part everybody already agrees is crowded. But before we go any further, I want to be clear about the clinical side of David's night, because I practised pediatrics for a long time and I don't want this to read as a story about a long wait.
Meaning what.
A sixteen-month-old with oxygen saturation at ninety percent on room air is a child in respiratory distress. That's not a borderline number, that's not something you write down and keep an eye on. In any properly staffed pediatric setting, that reading triggers continuous monitoring, escalation to a physician, and assisted delivery of oxygen. The standard isn't 'let the parents figure out the mask.' Toddlers tear masks off. That is the single most predictable fact of pediatric respiratory care, and every children's ward in the world has a protocol for it, because small children fight the thing that's helping them.
So it's not bad manners. It's a specific failure at a specific point in a pathway.
Two specific failures, and David's right to separate them from how the doctor came across. One is clinical, one is human, and both matter, but they don't matter the same way.
Let's take them in order. What came first in the room.
First is the gap. A nurse documents ninety percent, and then no physician examines the child for twenty minutes. Second is that nobody is available to help two frightened parents get treatment into a struggling toddler. Third, and this is the one that troubles me most, is that there's apparently no monitoring and no escalation while he's getting worse. He vomited. That's a child deteriorating in front of people who are supposed to be watching him. David also describes the physician interaction, once it finally happened, as disengaged and begrudging, and he's careful to say that's his impression and not a judgment about the man's motives or what his shift was like. I'll take him at his word on that. But bedside manner and patient safety are different categories, and he's only reporting one of them as subjective.
So on the facts as described, what's the formal word for it.
Supervision failure. Basic safety failure. Not 'slow service.'
The word that keeps coming back to me is deserted. He described the place as practically empty. An urgent care center that's supposed to be taking pressure off an emergency room, empty, on a night a baby needed oxygen.
And that changes the question. If it's one burnt-out doctor, that's a Tuesday anywhere. If the rooms are empty and nobody's covering them, that's the model.
Take us through the second failure. The fund.
So they've decided the facility can't help them. They're driving to the hospital themselves, and on the way they call their health fund, hoping for a referral or an authorization that stops a large emergency department bill landing on them. And what they get is an automated phone tree that announces it isn't an emergency service, and then about fifteen minutes on hold before they hang up and keep driving.
Which is the exact call the fund exists to answer.
It is. The twenty-four-hour line staffed by nurses is one of the advertised functions of the fund. The closest thing I can find to corroboration in the data is that CBS's survey last year found twenty-nine percent of Israelis were bothered by waiting for referrals and financial approval, and more than eight hundred thousand people struggle to find information about their rights. That isn't proof of a phone-line collapse. It's a pattern with a shape, and the shape is that the administrative layer is where people get stuck.
And then failure three. They get to pediatric triage, their son is still working to breathe, and the triage nurse tells them about an outstanding bill of a few hundred dollars from the previous visit. Even though the fund later authorized reimbursement. So the authorization went through and the charge stayed open.
That's a coordination failure between two organizations that are supposed to be talking to each other. And it surfaces at triage, before clinical assessment, on the second worst night of these parents' year. Somebody's billing system did not know what somebody's authorization system had done.
If the fund and the hospital can't reconcile a payment, I have less confidence they can reconcile a patient.
That's the thing about billing failures. They're never only about money. They're a readout of how well the plumbing is connected, and the plumbing is what you need when your baby's struggling to breathe.
So that's three failures in one evening, from institutions that were all supposed to be the safety net. Now zoom out, because David's hunch is that something bigger is behind it. And on that, there's an actual study.
There is, and it landed in September. Prof. Itai Ater and Prof. Nittai Bergman at Tel Aviv University, with Doron Zamir. They tracked Israeli doctors who left the country for at least a year. Between 2023 and 2025, one thousand three hundred and sixty-nine of them.
Against what baseline.
Against seven hundred and five in 2013 to 2015. That's a ninety-four percent increase. The yearly shape is four hundred and nineteen in 2023, a record five hundred and thirty in 2024, and an estimated four hundred and twenty in 2025. And the number is more than the entire medical staff of a large hospital. Shaare Zedek runs about a thousand doctors. Soroka about nine hundred.
So in three years, the country lost a hospital's worth of physicians. Which hospital.
Any of them. That's the point of the comparison. But here's the part that matters more than the headline, and it's the part Ater himself keeps pressing on. Doctors over forty-five were about twenty percent of the people leaving, against fourteen percent a decade earlier. About two hundred and seventy senior doctors, where you'd have expected closer to a hundred.
One in five of the leavers is a department head or close to it.
Or somebody who trains the residents. Net loss, after counting the people who come back, averaged about two hundred and fifty-five doctors a year in 2023 and 2024. In the 2010s it averaged about fifty-four. Nearly five times higher.
Where does the loss land, by specialty.
Pediatrics lost ninety-four doctors. That's four point three percent of all pediatricians, and six point four percent of the pediatricians under sixty. Surgery lost a hundred and eighty-eight, which is six percent of all surgeons and nearly nine percent of those under sixty. Internal medicine, two hundred and fifteen. Family and community medicine, eighty-one. Intensive care and emergency medicine, thirty-nine.
Pediatrics is the one that isn't abstract for us tonight.
It isn't. And there's a second layer on top of it. Finance Ministry data show about twenty-four percent of specialists in public hospitals are approaching retirement age. So the senior cohort is thinning from both ends at once.
Prof. Shelly at Rambam said something about this that I keep turning over. He said losing a resident can be made up in a few years. Losing a veteran specialist also takes away the ability to train the next generation, and that gap closes in a decade, not two years.
That's the whole shape of the problem in one sentence. A hospital can absorb the loss of a resident. It cannot quickly absorb the loss of the person who makes residents. That's a capacity you can't hire back, because the person you're hiring back is the person who trains the replacement.
And Ater's line about it, that the service all of us receive will be worse. Not some other patient. All of us.
All of us, yes. Now, I want to be careful, because this is where the episode could overreach. We cannot claim a study links physician emigration to staffing levels at a specific urgent care chain. Nobody has measured that. What we can say is that a system losing pediatricians, emergency physicians and senior trainers is a system where the layer most dependent on adequate staffing is going to show strain first. That's a mechanism. It's a plausible one.
It's not a proven one.
It is not a proven one, and I'd rather say so than pretend we have a receipt.
Fair. Then there's the visible symptom, which is the waiting. And this got measured properly for the first time.
July, the Health Ministry published waiting times, and it's the first time they've done it, which is itself remarkable for a system this studied. Median MRI wait, fifty-five days nationally. But the national figure hides the story. Clalit members averaged about sixty-four days. Meuhedet members, thirty-five and a half. Nearly double, for the same scan, depending on which fund card is in your wallet.
Children.
Children aged zero to fourteen at Clalit waited an average of seventy-one days for an MRI.
Seventy-one days to look inside a child's head.
And the surgery disparities are worse than the scan disparities. Cataract surgery, median two hundred and two days at Carmel, twenty-two days at Laniado. Ear tubes, a hundred and twenty-nine days at Soroka, one week at Ichilov. Tonsils, a hundred and fifty-four days at Galilee Medical Center, one week at Ichilov.
So the same country, the same tax, and a gap of an order of magnitude depending on which building you walk into.
And on the survey side, over a million Israelis aged twenty and up waited more than a month for a specialist. More than three hundred thousand waited over three months. Dermatology was the worst, forty-four percent waiting more than a month.
Now give me the piece that argues the other way, because I know you have it.
I do. Israel still beats the OECD on the procedures they report. Knee replacement, seventy days here against a hundred and ninety-nine across the OECD. Hip replacement, fifty-seven against a hundred and ten. So it isn't a system that's uniformly failing. It's a system that is unequally straining, and the inequality is the finding.
The average is fine and the variance will kill you.
That's the whole thing in nine words.
So the workforce is thinning and the waits are lengthening, and one family we know got left alone in a room for twenty minutes. Here's the puzzle, though. If all of that is true, why do eighty-seven percent of Israelis say they trust the system?
Because it's the highest trust any institution gets in the country. The CBS social survey put the health system at eighty-seven percent, above the army, above the police, above the courts. And among people who'd actually been to an emergency room, seventy-five percent were satisfied with how staff treated them and sixty-seven percent with the physical conditions. Both of those are better than they were in 2015.
So people who use the system like the system.
They like the people in it. And then the same survey says sixty-one percent of emergency room visitors complained about waiting times. Twenty percent, about five hundred and forty thousand people, rated their ER care negatively. One point two million Israelis paid for private medicine in the past year, and thirty-eight percent of them did it to get an earlier appointment.
A million and a half people quietly buying their way out of the queue is not trust. That's an escape hatch.
It's the paradox in one line. The trust is in the clinicians and the hospitals. It isn't in the access. And David fits it perfectly. He trusts the hospital completely. What he's lost faith in is everything before it.
Which raises his hypothesis directly. He met another family in that waiting room with a kid with the same symptoms, and those parents said they don't bother with urgent care anymore. They go straight to the hospital. David wondered if that's happening quietly at scale, families abandoning the layer designed to keep them out of the ER.
I want to be honest here, because this is the one where I'd love to have a number and I don't. We found no study quantifying that. No report measuring how many people skip urgent care and go straight to an emergency department.
None.
None. What we have is suggestive and not more. Ben-Gurion University built an AI model that analyzes two point eight million emergency room visits, and the entire reason it exists is that ER crowding is a recognized problem. Sheba runs an AI emergency department, Rambam runs a system called Shaked. All of them are attempts to manage the flow. None of them tells us why the flow is what it is. David's other family is one conversation in a waiting room. It's not data.
But it's the right question.
It's the right question. And one out of every fourteen Israelis says the urgent care they're supposed to use doesn't exist where they live, which is the closest thing to a concrete answer we have.
Say that again.
Seven point one percent, over four hundred thousand people, told the CBS survey there are no urgent medical services in their area. In the West Bank it's twenty-nine percent. In the northern district, twenty-eight percent.
So it's not that they don't trust it. It's that it isn't there.
For four hundred thousand people, there's nothing to trust. And now put that next to the waiting times. Two hundred and two days for cataract surgery at Carmel. Twenty-two at Laniado. A hundred and twenty-nine days for ear tubes at Soroka, a week at Ichilov. The failures cluster in the periphery, and the people with the most private-care options to route around them are in the center. You live in Tel Aviv, you have choices. You live in the north, you have a queue.
David's in Jerusalem. Center, but not Tel Aviv.
Center, but not Tel Aviv, and that may be exactly what his night was. A system that works well in one place and strains everywhere else.
Then there's the war, and I think the war is the biggest single thing pushing on this and the thing least present in the conversation.
Clalit published a three-year report this month. Children's psychiatric emergency visits are up a hundred and twenty-one percent since October seventh, 2023. Anxiety diagnoses up forty-five and a half percent. PTSD diagnoses have doubled. Twenty percent of Israelis still report severe impairment of daily functioning.
Twenty percent. One in five.
And the PTSD numbers among the Gaza-border evacuees are the ones I can't get out of my head. Eighteen times the prewar rate in the first year. Eleven times in the second. Still six times in the third. It's coming down. It is nowhere near where it started. New high-cholesterol diagnoses up twenty percent overall, and sixty percent among northern evacuees. The Defense Ministry has more than thirty thousand wounded security personnel since October seventh, and about sixty-three percent of them with severe psychological trauma.
Demand going up while supply goes down.
That's the sentence. Dr. Yarden Levinsky at Clalit said it directly, that this burden is unlikely to return to prewar levels. So we're not waiting for a wave to pass. We're looking at a new baseline.
So the system is strained on the demand side permanently and losing capacity on the supply side. And with an election coming, David says he sees no alarm. I went and looked, and he's right.
Today's coverage, on the twelfth. Security stories, judicial stories, coalition stories, the Netanyahu campaign leaks, Eisenkot fighting to keep small parties out of the Knesset. No health story on the front page.
And I tried to find what the parties have actually pledged on health, and I couldn't. Not a list, not a page, nothing I could stand behind. So I'm not going to tell you what's in the manifestos, because I don't know.
That gap is itself a finding, and I'd rather report it as a gap than dress it up.
Then let's take the explanations seriously, because there are a few and I don't think any of them is stupid. First, eighty-seven percent trust. If people like the system, there's no constituency for a politician to run against it.
And second, the damage is slow. Ater and Shelly both stress the gap closes over a decade, not two years. There is no single morning when the waiting room is empty and it's on every front page. It's a deficit that compounds quietly. That's the kind of deterioration that never produces a crisis moment to be alarmed about.
Third, it's geographically concentrated. If the people bearing the worst of it are in the periphery and the people without private options, they're also politically weaker, and the people with options are already out of the public system. The pain is real and it's not where the votes concentrate.
And fourth, the war crowds out everything. Prof. Yasmin Maor at Wolfson put it as plainly as anyone. It's already being felt, but not yet at full intensity, and the question is what will be here in ten years. She also said the election results will have a major impact on the choices people make, and that she doesn't see that being done by the current government.
Not yet at full intensity. That's the phrase I'd put on the whole episode.
It is the whole episode. We can't say the system is collapsing. We can say the layer below the hospital is strained in measurable ways, skewed against the periphery, and getting no political oxygen in an election year. Whether that's collapse depends entirely on your time horizon.
And then there's the part of David's night that I haven't been able to put down. They're standing inside a healthcare facility, their baby has been assessed as needing oxygen, nobody's helping them, and the thing they turn to is a chatbot.
And the chatbot told them to go to an emergency department, which was the correct advice. Which is almost worse than if it had been wrong.
It's worse. Because it means the most reliable triage they had that night answered in a second and had no stake in the bill.
Here's the image I can't shake. Ben-Gurion University built a model to route people through emergency departments, Rambam built Shaked, Sheba built an AI emergency department. All that work exists to compensate for human capacity that isn't there. David reached for a general-purpose language model for the same reason. The system is building machines to do the routing because it can't staff the humans who used to do the routing. The workaround and the failure are the same phenomenon.
You're telling me the fix and the symptom have the same root.
I'm telling you they're not two stories. They're one story told twice.
My cousin's wife. Pediatric nurse. She worked an urgent care center up north for years.
Of course she did.
She kept a laminated card in her scrubs pocket. Direct line to a hospital pediatric ER. She got it from a paramedic she met at a wedding, and he wrote it on the back of a receipt for a ceiling fan.
A ceiling fan.
She still has the receipt. She produced it at a family dinner once to settle an argument about whether the number had changed. The store closed in 2019.
How is a receipt from a store that closed in 2019 still in her pocket.
I don't know. I've asked her. She gets annoyed. Anyway. She said the fund's triage line never once picked up before the twentieth ring. Twenty rings, then you hang up and you use the card.
A nurse, inside the system, was routing around the system's own phone line.
Everyone at that clinic knew. That's the part people miss. This isn't patients learning to distrust it. The staff got there first.
That's a small object, that card.
There's one more thing. I'd push back on one thing you two said. It isn't that urgent care is understaffed. It was never staffed for what it gets. She was trained for ear infections and sprained ankles. By the end she was doing triage on infants with saturations in the low nineties, because there was nowhere else for those families to go. You can't staff for ear infections and then hand people a baby who can't breathe. She once spent a whole shift unable to leave one room, because there was nobody to cover her for a bathroom break.
That's not a staffing crisis. That's a job that changed underneath the people doing it.
They were supposed to be the valve. Nobody resized the valve. And that ceiling fan was installed wrong. I've been meaning to bring it up with somebody for about a year.
Back to David's question, then, and what happens on a long enough horizon.
We can't say the system is collapsing. We can say the layer under the hospital is strained, on a record that's been measured for the first time, unevenly across the country, and politically invisible. And here's the thing I keep landing on. The eighty-seven percent isn't a refutation of David's night. It may be the reason his night never becomes an election issue.
That's the mechanism. The trust is real, it's earned by the people in the hospitals, and because it's real it hides the access layer underneath that's thinning. The physicians leaving are a ten-year problem, not a two-year one. The war demand isn't receding, it's settling into a new baseline. And the system is building AI to route patients around a human capacity problem it hasn't solved. So it isn't a question of whether the urgent care layer holds. It's what quietly replaces it when it doesn't.
One thing from the cutting room floor, because it's too good to waste. The Health Ministry shut down a private hospital in Tel Aviv this month over surgical hygiene, and the elective cases had to go into public hospital surgical departments that were already full. So the pressure doesn't arrive as a crisis. It arrives as a spillover from somewhere else.
The strain doesn't announce itself. It just moves.
Leave it there. David's question, unanswered and sharpened. If the people who study this are worried, and several of them said so by name, why is nobody campaigning on it?
Rereview if we've earned it, tell a friend. Everything's at my weird prompts dot com. And send us your own prompt on Telegram at t dot me slash MWP listener bot. Hilbert Flumingtop produces the show, as he has since before either of us could be trusted with a microphone.
If you enjoyed this one, try episode forty-five seventy-six, When a Society Starts to Creak; episode forty oh two, How to Spot Real Policy Research vs. Political Vibes; and episode thirty eighty-five, Why Jerusalem Feels Unsteered While Its Mayor Keeps Winning. This has been My Weird Prompts.
We'll be back soon.
