Washington Will Pay to Reverse Your Overdose. It Won't Pay to Help You Avoid One.

Ty Hughes died in his bed in Oklahoma on May 11, 2020. He was 22. He had taken a pill he thought was oxycodone, alongside a former high school classmate, and it was not oxycodone. His sister, Shelby Salazar, spent the six years since building a nonprofit called Talk About Ty and helping keep a naloxone vending machine stocked in Lawton — a machine she told Oklahoma Watch reporter Raynee Howell she had to refill “quite a bit, maybe a few times a week.”

The machine is gone. So is the rest of Oklahoma’s naloxone vending program, once the largest in the country: 40-plus machines bought, 18 of which never left the warehouse, $14 million spent, canceled in late 2024 after roughly fifteen months. Three of the surviving machines now sit with the East Central Oklahoma Family Health Center, whose CEO Donna Dyer told Howell the state handed over the first load of supplies and after that, “we’ll pay for those out of our program income.” On July 1, the state law that let harm reduction organizations register and operate without criminal exposure expired, because House Bill 2012 passed the Oklahoma House and the Senate never brought it to a vote.

Salazar’s summary of how her brother died is the whole argument in one sentence: “Since Ty did not know that the pill he was about to take was going to kill him, he was just alone in his bed.”

Now hold that next to what the federal government is doing this week.

SAMHSA will pay to bring you back and has made it a disqualifying offense to warn you first.

On July 6, HHS announced more than $281 million across 15 SAMHSA grant programs — $68.2 million for medication treatment of opioid use disorder, $34.7 million for first responder overdose-reversal training, $11 million for community overdose prevention. Applications close Monday, July 27. And every one of those 2026 notices carries the same line: applications must align with SAMHSA’s strategic priorities, and “the application and budget narrative must not support harm reduction.”

The Agency Kept the Money and Deleted the Category

Start with the plain-language version, because the jargon here is doing real work. “Harm reduction” is not a philosophy course. It’s the set of things that keep a person who is using drugs alive and uninfected until whatever happens next happens — sterile syringes so you don’t get hepatitis C, a paper strip you dip in a few drops of dissolved drug to see whether fentanyl is in it, a phone line you call so someone knows to send help if you stop answering, sterile water and a bit of vitamin C to dissolve the powder without abscessing your arm. Cheap, unglamorous, mechanical.

SAMHSA has spent a year taking those apart one at a time. Under a July 2025 executive order, syringes and safer-smoking supplies went off the fundable list, but fentanyl and xylazine test strips survived — the agency explicitly called them “life-saving overdose prevention tools.” Then on April 24, 2026, SAMHSA’s acting leader Chris Carroll sent two Dear Colleague letters that reversed that. Filter’s Kastalia Medrano reported that the new letter banned by name fentanyl strips, xylazine strips, the newer medetomidine strips, and “any other substance test kits … intended for use by people using drugs.” Overdose hotlines went too, on the grounds that their “primary function” is providing a telephonic companion to someone in active use. The National Association of Counties’ summary adds sterile water, saline, and ascorbic acid to the prohibited column. STAT quoted the agency describing its own move as a “clear shift away from harm reduction and practices that facilitate illicit drug use.”

What survived is instructive. Naloxone and nalmefene stay fundable. So do lock boxes, disposal kits, overdose-reversal training, wound care, and HIV and hepatitis testing and treatment. Methadone and buprenorphine stay fundable — the $68.2 million program requires grantees to guarantee methadone access and offer at least one medication.

Methadone and buprenorphine stay fundable — the $68.2 million program requires grantees to guarantee methadone access and offer at least one medication.

So the money didn’t disappear. The word did. SAMHSA has reclassified naloxone and addiction medication as something other than harm reduction, which is a little like reclassifying seatbelts as something other than car safety. Medrano put the contradiction precisely: the agency insists addiction is “a lifelong, chronic, relapsing medical disease,” yet requires grantees to bundle medication with counseling in a way no one demands of insulin or chemotherapy. “If HHS was putting all this money toward what it viewed as a purely medical problem,” she wrote, “you’d think the solution would be viewed the same way.”

The second April 24 letter is the one clinicians should read twice. It instructs that “SAMHSA funding should be used to provide comprehensive treatment and recovery support services rather than medication-only models,” and leans toward eventually stopping medication rather than maintaining it. That is a clinical position, issued by a funding agency, pointing away from the single intervention with the strongest mortality evidence in this entire field.

HHS Secretary Robert F. Kennedy Jr. said the quiet part in Clinton Township, Michigan, on June 17, at a $700 million behavioral health announcement: “We know what doesn’t work. Ignoring addiction doesn’t work. Harm reduction doesn’t work.” He told Bridge Michigan that methadone and Suboxone are part of the president’s recovery initiative but that needle exchanges and safe injection sites “probably will not be getting federal money.”

The Cheapest Item in the Kit Is the One They Took

A fentanyl test strip costs about a dollar. That is the entire budget line under dispute.

In Ohio, the statewide nonprofit Harm Reduction Ohio distributed more than a million strips last year. Executive director AmandaLynn Reese told the Ohio Newsroom she expects to move about a quarter of that this year. “Everyone from your school nurse to your person who is injecting substances, living in an encampment, wants these supplies,” she said. “And now, at a dollar a piece, how is that sustainable for a nonprofit?” Tonja Catron, who runs the statewide SOAR Initiative, was blunter: “We can’t buy anything that keeps people who use drugs safe.” Then she said the thing that ought to end the argument. “I do not think that it’s gonna stop drug use in Ohio. If their long-term goal is recovery — if that’s what the government wants — that person is never gonna be alive long enough to make it there.”

Ohio State researcher Janet Childerhose named the theory of the case being tested here: “I understand the inclination to assume that if we take harm reduction supplies away from people who are using that we’re going to dissuade them from using. What we know from the evidence and from other kinds of public health interventions is that simply is not true.” Her description of the current supply is the reason the strips matter at all: “People are ingesting a supply that is unknown to them. It’s not regulated. It is adulterated with lots of different things. They don’t know what the amount is. Basically, people are sitting ducks.”

Let me be straight with you about what the strip evidence does and doesn’t show, because overselling it is how harm reduction loses arguments it should win. In a JAMA Network Open study published May 1, 2025, Rachel Vickers-Smith and colleagues followed 732 people who use drugs — 414 who used test strips, 318 who didn’t. Strip users scored meaningfully higher on a daily composite of overdose risk-reduction behaviors: 7.37 versus 6.51, a difference of 0.86 points. Over the 28-day follow-up, there was no statistically detectable difference in self-reported nonfatal overdoses between the groups. Strips change what people do. A four-week window in a few hundred people was never going to resolve whether that translates into fewer deaths, and the honest read is that the behavior signal is real and the mortality question is open.

Strip users scored meaningfully higher on a daily composite of overdose risk-reduction behaviors: 7.37 versus 6.51, a difference of 0.86 points.

Ask the people who actually use them and the picture sharpens. In 2021, researchers led by Megan Reed interviewed 29 people who use drugs in Philadelphia about test strips. They titled the resulting paper in the Journal of Substance Abuse Treatment with a participant’s own words: “If I had them, I would use them every time.” Most participants found the strips practical and folded them into daily routine. What they asked for was more supply, not less — strips in packs of 20, distributed with the water and cookers needed to run a test, available not just at harm reduction storefronts but at emergency departments, shelters, and food banks. That is what demand looks like from the demand side. SAMHSA’s response was to make the strips ineligible.

We Already Ran This Experiment, and We Counted the Bodies

This is the syringe fight again, with a different consumable.

In 1988, at the height of the AIDS epidemic, Congress banned federal money for needle exchange — a ban largely authored by Senator Jesse Helms on the theory that sterile syringes signaled approval of drug use. The ban held for most of three decades. Cities like New York and San Francisco built exchanges anyway on state, local, and private money, and HIV incidence among people who inject drugs collapsed where they operated. In places that waited for permission, it didn’t. Congress finally loosened the restriction in 2016, after an HIV outbreak tore through Scott County, Indiana — 235 infections in a rural county of 4,200 people, in a state whose governor had opposed exchanges until the outbreak forced his hand. Then in February 2022, a manufactured panic over “crack pipes” in a federal harm reduction grant sent HHS scrambling to clarify that pipes were never funded, and the political appetite for the whole category never recovered. The $30 million Congress authorized for harm reduction in 2021 expired at the end of 2024 and was not renewed.

We know the shape of the bill because researchers just modeled it. On June 18, 2026, Kirk Fetters, Pranav Padmanabhan, Kristina Yamkovoy and colleagues at the University of Colorado Anschutz and the University of Miami published a microsimulation in JAMA Network Open — a computer model that follows a simulated population of people who inject drugs and tracks what happens to them under different funding scenarios. Under a sustained 80% cut to syringe service program funding over five years, the model projects 39,600 excess deaths from all causes, including 15,600 excess overdose deaths. Under a milder 11% cut held for five years, it’s 5,400 excess deaths. The authors’ conclusion is one line: “Eliminating federal support for SSPs may result in excess overdose deaths and reverse recent progress in this field.” They did not model HIV or hepatitis C transmission at all, and said so — those infections sit outside the count entirely.

Regina LaBelle, who ran the White House drug policy office and now directs Georgetown’s Addiction and Public Policy Initiative, described to Roll Call how this plays out at ground level: “Syringe services will have to make some hard choices, like reducing services or hours.” Paul LaKosky of the Dave Purchase Project, who runs the Tacoma exchange, explained why that’s worse than it sounds: “While syringe exchange is the basis of what we do, it’s the least of what good syringe exchange does.” The syringe is the reason someone walks in the door. Wound care, HIV testing, a referral to a buprenorphine prescriber, a naloxone kit, someone who knows your name — that’s what happens after they’re inside.

The Line Is Finally Bending, Which Is a Strange Moment to Break the Tools

Here is what makes this genuinely reckless rather than merely wrong. Overdose deaths are falling, hard. CDC’s National Center for Health Statistics reported on May 13 that an estimated 69,973 people died of overdose in the United States in 2025, down almost 14% from 81,313 in 2024 — the third consecutive annual decline. Opioid-involved deaths fell from 55,296 to 44,564. Cocaine and methamphetamine deaths fell too. Rhode Island, New York, North Carolina, Alabama and Vermont each posted declines of 25% or more.

Rhode Island, New York, North Carolina, Alabama and Vermont each posted declines of 25% or more.

Nobody can cleanly apportion credit among naloxone saturation, buprenorphine expansion, changes in the fentanyl supply, and the grim arithmetic of a shrinking population of people at highest risk. But three states went the other direction by 10% or more: New Mexico, Arizona, and Colorado. The gains are uneven, provisional, and reversible. The last four years are the strongest argument ever assembled for keeping every tool on the table, and the policy response has been to sort the tools into the ones that look like medicine and the ones that look like permission — and throw out the second pile.

Oklahoma shows what the second pile is worth. Fentanyl poisoning deaths there went from 127 in 2020 to 730 in 2023, then down to 487 in 2024 — during the fifteen months the vending machines were running. Ryane Draper, assistant director of campus wellness at Oklahoma State, told Oklahoma Watch what the loss means in a rural state: “Education can only go so far, especially in rural communities where their access to everything is already so limited.” And then the part that isn’t about supply chains at all: “People don’t feel as safe, and they don’t feel as supported.”

If you use drugs, or you love someone who does, here is what is still standing tonight and is not going anywhere. Naloxone remains explicitly fundable under every one of these SAMHSA grants, including the $34.7 million first-responder program. It is available over the counter without a prescription. NEXT Distro mails it free, nationwide, to people who can’t get it locally. Oklahoma’s “OK I’m Ready” mail program still ships it. Methadone and buprenorphine are not only still funded, they are the single largest line item in this entire $281 million package. Nothing in the April letters made it illegal to buy a test strip with your own money, or a nonprofit’s money, or a state’s money — it made it ineligible for one specific federal wallet. Several states are already spending opioid settlement dollars to cover the gap.

But understand the trade that was made on your behalf. The federal government looked at a person who has just swallowed something they can’t identify and decided it would pay for the ambulance, the naloxone, the emergency department, and the months of buprenorphine afterward — and would not pay the one dollar it costs to tell that person, before they swallow it, what is in their hand. Every dollar of that is more expensive than the dollar it refused.

Ty Hughes took a pill he believed was oxycodone in May 2020, before most people outside harm reduction had heard of a test strip. Six years later, the strips exist, cost about a dollar, and there is now a federal grant clause specifically ensuring that the $281 million announced this month cannot buy a single one. Applications are due Monday. Somewhere tonight a nonprofit finance director is stripping test strips out of a budget narrative so the rest of the application survives review, and that deletion will not appear in any report, any dashboard, any press release. It will show up somewhere else, later, in a bedroom, alone.