A Thousand People Watched Someone Overdose. Peers With Naloxone Saved 83 Percent of Them.

Somewhere in England, Wales, Scotland, or Sweden, sometime in the past two years, more than a thousand people were handed a naloxone kit and sent back into their lives. Researchers wanted to know what happened the next time one of them watched a friend stop breathing. The answer, published this month, is that a stranger’s training doesn’t save most people who survive an overdose — a peer with a kit does.

That’s the finding from NalPORS — the Naloxone Prospective Outcome Research Study — published in Addiction on July 12, led by Nadine Metrebian and colleagues. It’s the kind of study harm-reduction workers have wanted for years: not a retrospective survey asking people to recall, months later, whether naloxone “worked,” but a prospective cohort followed forward in time from the moment they received take-home naloxone.

What the study actually did

Researchers enrolled 1,025 to 1,028 people who use opioids at 22 drug treatment and harm-reduction sites across the UK and Sweden between 2021 and 2024, each receiving take-home naloxone — either the injectable prefilled-syringe formulation (2mg) or one of two nasal spray doses (2×1.8mg or 2×1.26mg) — at enrollment. Of those, 594 were followed for a full six months, with researchers tracking not survey recall but real-time reports of what happened when participants actually witnessed an overdose.

In plain terms: this wasn’t asking “do you think naloxone helps people.” It was asking “the last time you were standing next to someone who stopped breathing, what did you do, and did it work.”

The number

Among the 160 participants who witnessed at least one opioid overdose during follow-up — 278 witnessed overdoses in total across six months — naloxone was administered in 83% of cases (95% CI 74–85%), and it was administered safely by people with no formal medical training. Medscape’s coverage frames the headline finding correctly: this is confirmation, at scale and in real time, that ordinary people who use drugs are not just capable of reversing overdoses — they are, in practice, the people who actually do it, because they are the people who are there when it happens.

That last point matters more than it sounds. Emergency responders are, almost by definition, never present at the moment an overdose starts. The person who is present is whoever’s in the room — a partner, a friend, someone using the same batch. NalPORS measured what happens when that person has been handed the tool in advance. Four out of five times, it worked.

Why this lands differently this month

This study did not arrive in a vacuum. As we reported today, a $281 million SAMHSA grant cycle closed July 27 with language barring applicants from proposing harm reduction — the exact category of work, including naloxone distribution, that NalPORS just validated at 83% real-world effectiveness. A JAMA Network Open modeling study published in June projected up to 39,600 additional deaths over five years from the related defunding of syringe services, which function as one of the main channels through which take-home naloxone actually reaches people. Overdose deaths are, encouragingly, falling nationally — down 13.2% year-over-year in the latest CDC data — and peer-distributed naloxone is one of the interventions credited with that decline.

What the study doesn’t tell you

Calibrated confidence means naming the limits along with the finding. NalPORS is observational, not a randomized trial — participants weren’t compared against a group without take-home naloxone, so the study can’t isolate naloxone’s effect from other factors present at these 22 well-resourced treatment and harm-reduction sites. The population is European, and formulations, drug-supply composition (UK and Swedish opioid supplies differ meaningfully from the U.S., where fentanyl and its analogues dominate), and health-system context don’t transfer perfectly to an American reader. And the 83% figure describes successful naloxone administration, not perfect outcomes in every case — a small number of witnessed overdoses in the study still ended without survival, a reminder that naloxone is necessary and highly effective, not infallible.

None of that weakens the headline finding. It sharpens it: the strongest available real-world evidence says that giving naloxone to the people most likely to be standing next to an overdose, with no medical training required, saves the overwhelming majority of the lives it’s given the chance to reach. The open question isn’t whether this intervention works. It’s whether the funding structure that gets naloxone into those hands survives long enough to keep proving it.

It’s whether the funding structure that gets naloxone into those hands survives long enough to keep proving it.