Emerging drug threats · Overdose surveillance
Population-based emerging threat analysis
Population-based emerging threat analysis is continuous detection of changes in a local drug supply, such as carfentanil reappearing, using toxicology, overdose events and street-vocabulary signals corroborated against each other. The output is a warning for a county, a corridor or a set of ZIP codes, never for an individual.
Detecting a change in what is killing people in a jurisdiction early enough to warn the people it will kill next.
Why does emerging drug threat analysis matter?
The current lethal supply is not the one any prescribing policy was designed for. CDC surveillance across 45 states and the District of Columbia estimated that during 2023, approximately 72,000 drug overdose deaths — nearly seven in ten — involved illegally manufactured fentanyls. Across 2021–2023, illegally manufactured fentanyls were detected in 188,082 of 251,089 overdose deaths, 74.9%.
And the composition of that supply keeps moving. Carfentanil, which has no approved human therapeutic use, produced an outbreak of 1,181 deaths in Florida across 2016–2017 and then largely receded, before reappearing: 29 deaths in the first half of 2023 and 238 in the first half of 2024. CDC and DEA both describe carfentanil as far more potent than fentanyl — the published estimates are roughly 100 times fentanyl, or 10,000 times morphine — and both hedge those figures explicitly, because analog potency has not been evaluated in humans. We reproduce the hedge rather than dropping it.
The interval between a change like that entering a local supply and a jurisdiction knowing about it is measured in months. That interval is where people die who did not have to.
How does emerging threat analysis detect a change in the drug supply?
- Toxicology and overdose data assembled continuously rather than in reporting cycles, so a change in composition is visible as a change rather than as a revised annual figure.
- Corroboration across independent sources — toxicology, overdose events, and the vocabulary and geography signals from supply signal analysis. A single source is a hypothesis.
- Geographic resolution at the level a response can act on — a county, a corridor, a set of ZIP codes — never an individual.
- Output is a warning with a recommended action: what has changed, where, what it is being called locally, and what services and clinicians should do about it.
Does emerging threat analysis track individuals?
It is not individual-level. The unit of analysis is a place and a substance, and outputs are aggregate. This distinction is structural, not a policy setting.
It is not enforcement targeting. The purpose is to warn people who use drugs and the services around them. An agency intending to use it for interdiction is using a public-health instrument for something it was not built for, and should expect us to raise it.
It is not a prediction of who will overdose. It is a description of what is in a supply.
An honesty note on the classification. CDC states that where there was insufficient evidence to classify detected fentanyl as illegal or prescription — 9.9% of such deaths — it was classified as illegal on the grounds that most fentanyl overdose deaths involve illegal fentanyl. Part of the illicit share is therefore an assumption. It is a reasonable one, and it is still an assumption.
Frequently asked questions
What is carfentanil, and why is it tracked?
Carfentanil has no approved human therapeutic use. It caused 1,181 deaths in Florida across 2016 and 2017, largely receded, then reappeared: 29 deaths in the first half of 2023 and 238 in the first half of 2024. CDC and DEA estimate it at roughly 100 times the potency of fentanyl, a figure both agencies hedge because analog potency has not been evaluated in humans.
How many overdose deaths involve fentanyl?
CDC surveillance across 45 states and the District of Columbia estimated that about 72,000 overdose deaths in 2023, nearly seven in ten, involved illegally manufactured fentanyls. Across 2021 to 2023 they were detected in 74.9% of overdose deaths. Part of that share is an assumption: where the evidence could not tell illegal from prescription fentanyl, 9.9% of such deaths, CDC classified it as illegal.
What does an emerging threat warning include?
What has changed in the supply, where it has changed, what it is being called locally, and what services and clinicians should do about it. The warning covers a county, a corridor or a set of ZIP codes, never an individual, and it goes out only after toxicology, overdose events and street-vocabulary signals corroborate one another.
Can emerging threat analysis predict who will overdose?
No. It describes what is in a supply, not who will be harmed by it. The unit of analysis is a place and a substance, and every output is aggregate. That is structural, not a policy setting. An agency that wants to use it for interdiction is using a public-health instrument for something it was not built for, and we will raise it.
Sources
Every figure on this page is traceable to the source listed here.
- Tanz LJ, Stewart A, Gladden RM, Ko JY, Owens L, O’Donnell J. Detection of illegally manufactured fentanyls and carfentanil in drug overdose deaths — United States, 2021–2024. MMWR Morb Mortal Wkly Rep. 2024;73(48):1099–1105. PMID 39636782. View source.
- Delcher C, Wang Y, Vega RS, et al. Carfentanil outbreak — Florida, 2016–2017. MMWR Morb Mortal Wkly Rep. 2020;69(5):125–129. PMID 32027630. View source.
- O’Donnell JK, Halpin J, Mattson CL, Goldberger BA, Gladden RM. Deaths involving fentanyl, fentanyl analogs, and U-47700 — 10 states, July–December 2016. MMWR Morb Mortal Wkly Rep. 2017;66(43):1197–1202. PMID 29095804. View source.
- Drug Enforcement Administration, Diversion Control Division, Drug & Chemical Evaluation Section. Fentanyl. December 2025. View source.