Washington Wants You to Prove You Deserve a Bed. Jean Descamps Ran Out of Time.

Jean Descamps was 26 when a hospital in Milwaukie, Oregon decided he’d had enough of its help. It was winter. Someone at Providence Milwaukie Hospital wrote in his chart that he was “malingering” — hospital slang for faking it, for gaming the system to get attention he supposedly didn’t need. The hospital marked him stable and sent him out into the cold with nowhere to go. He was dead of an overdose within hours. His case is one of the ones anchoring an NPR investigation published this month into the years leading up to America’s overdose deaths.

NPR’s reporters found that two-thirds of the people who died of an overdose in 2024 had already been in front of the healthcare system in some documented way — an ER visit, a hospitalization, a note in a chart — before they went on to die anyway. More than 40 million Americans live with a substance use disorder, a diagnosable, treatable medical condition involving compulsive drug use despite harm, not a character flaw, and more than 80 percent of them get no treatment at all, according to the same investigation.

A hospital decided Jean Descamps hadn’t earned a bed. Washington just decided the same thing about everyone else.

Days after NPR’s investigation ran, the Department of Health and Human Services and the Department of Housing and Urban Development unveiled a new “Treatment First” toolkit — federal guidance instructing cities and clinics on how to run addiction and homelessness programs — that pushes local governments toward the opposite of what a decade of housing data has shown: give someone a stable place to live first, without conditions, and their odds of surviving and eventually stabilizing go up. The toolkit tells them to require proof of change before offering shelter instead, according to STAT News’s reporting on the rollout.

A Parking Lot Is Not a Discharge Plan

If you’ve ever sat in an emergency room hoping a nurse would believe you were actually in danger, you don’t need “malingering” translated for you. You already know the look. Jean Descamps got that look from a hospital that had every tool available to help him and used none of them. He wasn’t an abstraction or a data point before he was a data point — he was a 26-year-old man who walked into a hospital because he needed something, and walked out of it because the people inside decided he didn’t.

That decision is not rare, and it is not new. NPR’s investigation found that only 1 in 5 physicians are willing to treat opioid use disorder at all, and that 43 percent of healthcare workers surveyed wrongly believe that medications for opioid use disorder — buprenorphine and methadone, the daily medications that curb cravings and withdrawal and are the closest thing addiction medicine has to a gold standard — just “substitute one drug for another,” according to the same NPR reporting. That belief isn’t a footnote. It’s the reason a hospital can look at a person in active withdrawal, decide he’s exaggerating, and discharge him to a parking lot instead of starting him on a medication that might have kept him alive until morning.

This is the part of the story that rarely makes it past the hospital doors, because it doesn’t happen in a press conference. It happens in triage, in a five-minute conversation between a tired resident and a patient who’s been written off before the chart is even finished. Two-thirds of 2024’s fatal overdose victims crossed paths with a doctor or nurse first. Two-thirds. That’s not a system with a few bad actors in it. That’s a system whose default setting, at the bedside, is suspicion.

It happens in triage, in a five-minute conversation between a tired resident and a patient who’s been written off before the chart is even finished.

Here’s the concrete part, if you’re the clinician reading this rather than the person living it: this week, before your next discharge for a suspected overdose or opioid withdrawal, pull up your unit’s actual protocol and check one line — does it require staff to offer buprenorphine induction, a fast-acting medication that can be started in the ER within the same visit, or a warm handoff to a same-week treatment appointment before a patient can be marked “stable” and sent out the door? If the answer is no, that’s the gap to close, not the patient’s willpower.

Washington Just Rewrote the Rules on Meeting People Where They Are

The federal shift didn’t happen in one hospital. It happened in the toolkit HHS and HUD released together this month, which leans hard on abstinence — staying off all substances entirely, as the sole definition of success — plus faith-based intervention and contingency management, a program model where people are paid or rewarded for testing negative on drug screens. The toolkit barely mentions methadone or buprenorphine at all, and it explicitly rejects “unconditional” housing — meaning housing offered without first requiring sobriety or program compliance — in favor of housing that has to be earned, per STAT’s review of the document.

HUD Secretary Scott Turner and HHS Secretary Robert F. Kennedy Jr. have both put their names behind it publicly. Kennedy went further, touring California on August 13 to sell the approach directly, arguing that the state’s encampments and open drug use are the visible cost of a housing-first strategy that let people stay unwell indefinitely rather than pushing them toward treatment, according to STAT’s coverage of the trip. Turner has made a similar case for HUD’s side of the toolkit: that housing without expectations isn’t compassion, it’s abandonment dressed up as help.

Here’s the part of the toolkit’s own paperwork that undercuts the argument: it cites a single study on veteran housing and overdose risk as its central piece of justification, while acknowledging — in the same document — that Houston’s housing-first program successfully housed roughly 25,000 people over the past decade, as STAT reported. Twenty-five thousand people is not a rounding error. It is the largest, longest-running counter-example available in the entire country, sitting inside the government’s own toolkit, footnoted rather than followed. You don’t cite a decade of success and then recommend the opposite unless the conclusion was decided before the evidence was reviewed.

It’s also worth sitting with the timing. The CDC’s most recent provisional data shows 68,641 overdose deaths in the 12 months ending in February 2026, a 12.1 percent drop from the year before, per the CDC’s National Vital Statistics System — a real, sustained decline that occurred under the exact mixed approach, harm reduction plus treatment plus housing support, that the new toolkit is now moving away from. Something has been working. The toolkit doesn’t explain why the answer is to change direction rather than keep going.

We’ve Made This Mistake Before, and We Know How It Ends

This isn’t the first time the federal government has decided that people who use drugs need to earn help before receiving it, and it isn’t the first time that decision cost lives on a mass scale. In 1988, at the height of the AIDS epidemic, Congress banned federal funding for syringe exchange programs — the practice of giving people who inject drugs clean needles to use instead of shared ones, which cuts HIV and hepatitis transmission dramatically. The ban held for more than two decades while HIV spread through communities of people who inject drugs, lifted only briefly in 2009 before being reinstated again from 2011 to 2016. The public health case for exchanges was settled science well before the money finally followed it. Ideology held the line anyway, and people died of a preventable, treatable virus while it did.

Ideology held the line anyway, and people died of a preventable, treatable virus while it did.

That history isn’t a metaphor for what’s happening now — it’s the same policy lever, pulled again. SAMHSA, the federal agency that funds most addiction and harm reduction programs nationally, has already restricted funding this year for fentanyl test strips and syringes, tools that let people check what’s actually in their drugs and use them more safely, according to reporting from the National Association of Counties. County health departments that relied on that funding are now covering the gap themselves or cutting the programs outright.

The same week the Treatment First toolkit landed, Filter Magazine — a harm reduction news outlet that had covered this exact beat for eight years — announced it was shutting down, citing “an increasingly difficult funding landscape” for the kind of journalism that takes people who use drugs seriously as sources rather than cautionary tales, in the outlet’s own closing statement. That’s not a coincidence of scheduling. When the federal government decides a whole approach is out of favor, the institutions built around that approach — clinics, county programs, even the press that covers them — lose funding and standing at the same time, on the same week, for the same reason.

Same Failure, Two Faces

Here’s the thing both failures have in common: neither one is really about evidence. A resident who writes “malingering” in a chart isn’t citing a study. A toolkit that footnotes 25,000 successfully housed people and then recommends against unconditional housing isn’t citing a study either, not really — it’s citing a conclusion it already believed. Clinical stigma at the bedside and ideological stigma at the policy level are the same instinct, just operating at different altitudes: the belief that a person has to prove they’re worth saving before the system will spend effort saving them.

That instinct kills the same way both times. It kills Jean Descamps in a parking lot because a nurse decided he was faking a medical emergency. It will kill people in cities that adopt the new toolkit’s model, because a shelter bed with conditions attached is a shelter bed that a person in withdrawal, in psychosis, in crisis, often cannot meet the conditions for — not because they don’t want to, but because active substance use disorder makes “just comply first” close to impossible for the people who need help most urgently. The toolkit’s own evidence says housing first works. It recommends against it anyway. That is the whole story, compressed into one document.

If you’re the one in recovery reading this, or trying to get there, or watching someone you love try to get there: none of this is a verdict on you. A system built to demand proof before offering help was never measuring your worth. It was measuring its own willingness to trust you, and it failed that test, not you.

Naloxone — the medication that reverses an opioid overdose in minutes — is still legal, still federally funded, and still available without a prescription at pharmacies and health departments across the country. None of what happened this month changed that. It’s the one piece of ground that hasn’t moved.

Jean Descamps didn’t need a philosophy. He needed someone in that hospital to open a door instead of pointing him toward one. Somewhere tonight, another 26-year-old is sitting in an ER chair, sober enough to be blamed and sick enough to need help, waiting to find out which door he gets.