Homer wrote the Iliad about a war. Then he wrote a second epic, just as long, about one soldier trying to get home from that war. Ten years of fighting, ten years of return, and by the poet's own arithmetic, the journey back deserved as much attention as the siege of Troy itself.
Christopher Nolan has put that second story back in cinemas, and Israeli audiences are watching it through a lens nobody chose. For generations of readers, the Odyssey was a myth about monsters and patience. An Israeli watching it now, after nearly three years in which hundreds of thousands cycled between the front and the kitchen table, is watching a story about the longest part of a war, which is the part after it.
Jonathan Shay, the American veteran mental health psychiatrist whose reading of the Odyssey as a veteran's homecoming became foundational for military mental health, wrote about soldiers who came home once, for good.
That assumption quietly organizes most of what the world knows about combat trauma, and it does not describe this country. Of the roughly 22,000 wounded added to Israel's rehabilitation lists since October 7, 63 percent are reservists. For most of Israel's psychological casualties, home is a place the army calls you out of again.
The number that set a man on fire
On April 12, 2021, Itzik Saidyan, then 26, a Golani Brigade veteran of the 2014 Battle of Shejaia, set himself on fire in the entryway of the Defense Ministry's Rehabilitation Department in Petah Tikva. He survived, after 30 surgeries and 17 months at Sheba Medical Center, and he is a public advocate for wounded veterans today.
Self-immolation is not a legitimate form of protest, and nothing here treats it as one. A desperate act by a man in crisis nearly killed him, and it also forced onto the national agenda a question the system had spent decades not answering.
That question was the number. The ministry recognized Saidyan at 25% disability. He sought 50%. The official reasoning was that part of his condition traced to childhood trauma rather than to his service. Keep that dispute in view, because everything that follows is inside it: two parties negotiating percentage points of a mind, with no accepted, standardized method for computing them.
What the count actually counts
The official figures are imposing. The Rehabilitation Department caters to 82,400 people. Fifty-eight percent of the post-October 7 intake is treated for PTSD or other psychological conditions, some 1,500 new recognition claims arrive monthly, and the budget has grown 53% to NIS 8.3b., NIS 4.1b. of it for mental health, with forecasts passing 90,000 people this year and 100,000 by 2028.
Here is what those figures are not: epidemiology. When I put this project's original thesis to Prof. Brig.-Gen. (res.) Elon Glassberg, the former IDF surgeon-general, expecting him to confirm it, he took it apart instead, and the strongest part of his objection survives every check I have run since.
The caseload is a measure of recognition policy, he said, at least as much as of the underlying clinical picture. And he drew one line firmly: the systemic lens takes nothing away from the personal one. For the individual, he said, this condition is devastating, life-wrecking, and sad, and the question is whether the system treating it can tell what works.
After Saidyan, the state deliberately reversed the burden of proof. The One Soul reform ended the invasive background checks and adversarial committees that had forced veterans to litigate their own suffering, and after October 7, fast-track recognition followed. Those were defensible choices, arguably the only humane ones available. They also mean the curve measures policy as much as pathology. The United States ran the same experiment: in 2010 the US Department of Veterans Affairs dropped its requirement that veterans corroborate the specific stressor behind a PTSD claim, and claimants then surged for a decade.
One popular explanation, though, turns out to be wrong, and it matters. The rise is often blamed on psychiatry widening the diagnosis itself. The opposite happened: when the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition redrew PTSD's boundaries in 2013, population studies found prevalence came out slightly lower than under the old definition. The gates that moved were administrative, not clinical. Israel's count grew because Israel decided, for good reasons, to stop fighting its wounded at the door.
A family name, not a first name
Why can a state not simply measure this the way it measures everything else it buys? Because the thing itself resists the ruler. Under DSM-5 criteria, there are, by a 2013 calculation by Isaac Galatzer-Levy and Richard Bryant that has become famous in the field, exactly 636,120 distinct symptom combinations that qualify for a PTSD diagnosis, eight times more than under the previous definition.
Glassberg's formulation was blunter. PTSD, he said, is a family name: a general label for a condition carried by hundreds of thousands of people, presenting with a full spectrum of precipitating events and accompanying injuries. Each patient also has a first name, a specific clinical picture of their own, and the first names differ.
Even the vocabulary is unsettled: the defense system counts these people among its wounded. At the same time, medicine records a clinical diagnosis, and the gap between those two words is part of the story this piece is telling. There is no blood test, no scan, no validated biomarker for diagnosis or for recovery. The gold-standard instruments, which are structured interviews and symptom checklists, measure what a person reports about their own inner state.
They are serious tools, built and validated over decades. They are also, in the end, a conversation: the patient reports, the clinician records.
Physical injury is measured differently, and Glassberg's own career shows the difference. When he set out to reduce the share of wounded soldiers who die of their wounds, no ready-made instrument existed for that either. He chose one, the case fatality rate, which is the percentage of wounded soldiers who die, and made it the standard: the same calculation in every unit, independent of anyone's impression of how the care was going. The rate fell from roughly 15% at the start of this century to about 7%, and Israeli combat casualty care became a model other militaries study, because the number proved it worked. Psychological injury has no equivalent number. There is no figure that tells the system whether a treatment for PTSD works, and there never has been.
The market that ambiguity built
That absence shapes where money goes, in two opposite directions at once.
Investment that depends on proof of efficacy is scarce. The entire Israeli mental health technology sector raised $352m. in 2025, roughly a third of what the Defense Ministry alone spends on psychological care in a year. GrayMatters Health of Haifa, whose Prism device is the only one the FDA has cleared for PTSD, cleared as an adjunct to therapy through a pathway whose evidentiary bar provides for substantial equivalence rather than proof of efficacy, has raised under $20m. in its lifetime.
Empathy, an Israeli company that guides families through bereavement, has raised $162m. because a death arrives with a date, a claim, and an insurer already in the room, and six life insurers invested in it directly. An underwriter can price grief support. Nobody can underwrite an outcome that cannot be measured.
Revenue that does not depend on proof is plentiful. Neurostimulation helmets, wearables, retreats, protocols: a global industry sells trauma treatment profitably precisely because failure is unfalsifiable. The clearest demonstration arrived in August 2024, when the FDA rejected MDMA-assisted therapy for PTSD, the most celebrated trauma treatment of the decade.
Its trials showed large improvements on standard symptom scales. The advisory committee voted nine to two against efficacy anyway, because patients could tell whether they had received the drug, and a self-reported scale cannot survive that. A journal retracted three supporting studies the following day. The treatment may yet work. The point is different: after 20 years and hundreds of millions of dollars, the field's own measurement tools could not establish whether it did.
An economist would put it plainly. Where there is no hard endpoint, bad treatments are never exposed, marketing substitutes for evidence, and ambiguity itself becomes the moat. The same missing ruler that keeps the insurers out keeps the sellers in.
None of this is aimed at the wounded, for whom any one of these treatments may be a genuine lifeline, or at the clinicians and founders who want to help and often do. The missing ruler hurts them first. Without a standard, there is no individual diagnosis, no first name in Glassberg's language, and no professional way to match a specific treatment to the specific person in the room.
After reading a draft of this piece, Glassberg compressed it into one sentence: "Until we know how to measure, more miracle devices and miracle treatments, and more money poured into charities and therapies, will not add up to a systemic effect."
It does not have to be this way, and one country has proved it. England's National Health Service ties funding for its talking-therapies program to outcome measurement at every session, and captures standardized data on some 98% of the million-plus people it treats each year. The result is a national dashboard nobody else has, and what it shows is uncomfortable: overall recovery runs near 50%, but for PTSD specifically it has hovered around 40%, the weakest of the major conditions.
That discomfort is the entire lesson. Measurement's first product is bad news. A Haaretz investigation reported that when broader mental health screening was proposed inside the IDF during this war, senior officers resisted, fearing a Pandora's box that would leave units unmanned. They understood correctly what the English data confirms: an honest measurement exposes how much is not working. That is the case for building it anyway, because the alternative is not the absence of bad news. The alternative is 1,500 new claims a month adjudicated one percentage point at a time, against people who have already been hurt.
What the ruler would buy
Israel is unusually placed to build it. The Rehabilitation Department's 82,400-person caseload, with mental trauma its largest and fastest-growing share, is among the largest concentrated cohorts of war trauma in the Western world, growing monthly, under a single payer that already spends NIS 4.1b. a year. In every other domain where the Defense Ministry anchors demand, its specifications become the industry's product definitions.
If a fraction of that spend were conditioned on standardized, England-style outcome reporting, Israel would be doing for psychological injury what its casualty-care registry did for battlefield medicine: producing the dataset the world's clinicians, insurers, and regulators currently lack. Israeli combat care became export doctrine because its numbers could be audited. The first system that can say, with auditable numbers, what recovery from war trauma looks like will not just enter this market. It will define it.
A generation ago this country decided that six million was a number no mind could hold, and rebuilt Holocaust memory around individual names and faces, deliberately, as policy. That is the move the wounded now need: a general label is not enough, and the state that learned to give faces to the murdered has yet to give the ones who came back an individual, measurable diagnosis. Until it does, philanthropy is filling the gap: NATAL, an independent trauma nonprofit, carried about 100 patients before October 7 and has treated some 5,500 since, on donations.
The missing price
In April I said that the rising value of a human life has become a market force, that Western defense pays nearly any sum to keep a soldier alive, and that this moral premium is an engine rather than a burden. I stand by it, and I now see the clause I left out. We pay any price to keep a soldier alive. We have never agreed what it costs to bring one back, and we have built no instrument that could tell us.
Homer, who gave the journey home as many years and as many lines as the war, would have recognized the imbalance. Poseidon raises the storms, but the sea god does not account for the 10 years. One of them passes on Circe's island and seven on Calypso's, out of sight of everyone waiting in Ithaca. And when Odysseus finally lands on Ithaca, he is still not home. He spends three days in his swineherd's hut disguised as a beggar, then walks into his own palace, and the only one who knows him through the disguise is his old dog. His son, his old nurse and his wife each need proof, and it is days before the last of them is sure. The longest part of the journey home is the part nobody at home can see, and it continues after the soldier has walked through the door. That is the part no state has learned to measure.