Overdose Deaths Are Down 12%. That Is Not the Same As Fixed.

A number can be true and still be the wrong thing to celebrate alone. CDC’s newest provisional data puts the 12-month overdose death total ending in March 2026 at 67,798 — a 12.3% decrease from the same period a year earlier. That’s real. It’s also the kind of number that, read in isolation, invites exactly the wrong response: relief that turns into funding cuts, attention that moves somewhere else, a sense that the emergency phase is over.

Look one line down in the same dataset and the story gets more honest. Nonfatal overdoses — the ones that don’t make a death certificate but still mean an ambulance, an ER bed, and a naloxone reversal — are declining too, but more slowly: suspected overdoses accounted for 54.7 of every 10,000 emergency department visits in the same window, down only 2.6% year over year. Fatal overdoses falling roughly five times faster than nonfatal ones isn’t a contradiction — it’s what naloxone saturation looks like. More overdoses are being reversed before they become deaths. That’s a genuine public health win, worth naming plainly. It is a different thing entirely from fewer people overdosing in the first place, and conflating the two is how a real gain gets used to justify pulling the naloxone funding, the syringe services, and the outreach workers that are the actual reason the death number moved.

Provisional data carries its own caveat worth naming, not hiding: these are early counts that CDC itself expects to shift as death certificates finish processing, sometimes for months. A 12.3% decline reported today has, historically, moved by a few points either direction once the numbers are finalized. That’s not a reason to distrust the trend — multiple independent data streams have now shown declines for two straight years — it’s a reason to read “12.3%” as “meaningfully better, direction confirmed, exact number pending,” which is a less satisfying headline and a more honest one.

If you’re the person these numbers are actually about, here’s the version that matters more than any percentage: naloxone works better today than it ever has, more of it is in more hands, and the reason the death count is falling faster than the overdose count is that people are being brought back who wouldn’t have been five years ago. That’s not abstraction. That’s someone’s Tuesday. The number worth watching next isn’t this one — it’s whether the agencies that made this decline possible keep their funding through the year it takes to find out if it holds.