Overdose Deaths Nearly Halved, and We Still Can't Count the People Who Know Why
The national drug survey misses people in shelters and jails, the exact people the supply shifts hit first.
Picture the person a federal drug survey cannot reach: someone sleeping in a shelter, or on a county jail bunk, who knows exactly what the dope changed into this month and has never been asked. Three researchers made the same point this week in STAT, and it is the most useful sentence in the whole debate over why overdose deaths fell.
The country is bragging about a drop it cannot explain because it never built the instrument that would explain it.
J. Travis Donahoe, David Zhu, and Michael Barnett write that the National Survey on Drug Use and Health misses people who are homeless or incarcerated, asks almost nothing about illicit drug prices, and added a question on illicit fentanyl only in 2022, after deaths had already peaked. Researchers, they say, have been “largely limiting our research focus to areas with data,” and they compare it to searching for lost keys under a streetlight: insurance claims and treatment records are plentiful, so that is where the looking happens. Meanwhile the likeliest cause, a shock to the fentanyl supply that a Science paper in January described, runs through the one place the data are thinnest. The DEA no longer publishes regular purity and price figures, which that paper’s authors flagged as a limit.
The authors’ fix is concrete. Build a survey that reaches shelters and jails, asks about prices, and tests drug samples, the way CDC’s HIV risk surveillance and Australia’s long-running drug trends surveys already work. They also point out that the DEA’s shipment database was disclosed in 2019 through litigation, about eight years after prescribing peaked, which is a long time to be flying blind.
For providers, the data gap has a practical edge today. You are closer to the supply than any federal dataset. A case manager who writes down “client says the dope is weaker, price doubled, new sickness on mornings” and sends it to the county health department is doing surveillance. If your county runs drug checking, ask what they are seeing and tell them what you hear.
For the person who uses, that is also your leverage. Telling the truth about what is in front of you is the one data source no one can fund their way around.
When no one is counting you, the first thing to do is ask them to.
Sources Cited
- 01.B
- 02.B
Filed Under
policytrendssocial-culturalGovernment DataOverdose
Keep up with the reporting.
One email each morning with the stories that put days like this in context.
Continue reading
More from this section
Nobody Can Say Why You Stopped Dying. Until They Can, Don't Bet on It Lasting.
A JAMA analysis of 2.5 million EMS overdose responses finds naloxone use did not rise as deaths fell. The likeliest cause, a weaker fentanyl supply, can reverse. Here is what that means this week.
The Crisis, By the NumbersOverdose Deaths Fell 12.3%. The Numbers Came From the Same Office That Wants Credit.
CDC provisional data show 67,798 overdose deaths in the 12 months ending March 2026, down 12.3%. HHS officials claim credit as a SAMHSA nominee awaits confirmation.
The Crisis, By the NumbersHHS Is Claiming the Overdose Decline. Arizona, Where Deaths Rose, Isn't Invited to the Victory Lap.
Two HHS officials say overdose deaths are falling and the administration is continuing the momentum. The CDC-based reporting says the cause is unclear and that Arizona went the other way.