Drug supply monitoring · Street names
Drug-supply signal analysis
Drug-supply signal analysis tracks the street names and geography of substances in a region from openly published material, in aggregate, and checks each signal against toxicology and overdose data before anyone acts on it. The subject is the drug supply, never a person, and the output goes to warnings, not enforcement.
What substances are circulating in a region, under what street names, and where — assembled from public and aggregate sources. This is analysis of a drug market, not of people.
Why does drug-supply signal analysis matter?
Read this first, because it is the point of the page. The subject of this analysis is a drug supply. It is not a patient, a caseload or a population of individuals. Nothing in this component identifies a person, and it is not designed to.
The reason it exists is that the substances killing people change faster than any official reporting cycle. CDC surveillance found carfentanil-involved deaths rising roughly sevenfold in a year, from 29 in the first half of 2023 to 238 in the first half of 2024. An adulterant can enter a local supply, kill people for months and appear in a published dataset a year later. By then the pattern has moved.
The vocabulary problem is real too. What a substance is called locally changes constantly, and it changes precisely so that it is not recognized. A harm-reduction worker, an emergency clinician and a public health analyst in the same county can be describing the same thing with three different words and not know it.
How does drug-supply signal analysis work?
- Public, aggregate sources. Openly posted content, in aggregate, analyzed for vocabulary and geography. No account-level analysis, no identification of posters, no correlation of posts with patients.
- Vocabulary tracking. New street names for substances, and shifts in what existing names refer to — which is often the first observable sign that a supply has changed composition.
- Geographic aggregation. Output is at the level of a region, not a person. A heat map of a county is a public-health instrument. A map of an individual would be something else entirely, and this program does not produce one.
- Corroborated before it is used. A signal from open sources is a hypothesis. It is checked against toxicology and overdose data — see emerging threat analysis — before anybody acts on it.
- Output goes to warning, not enforcement. The product is “this is in the supply here, this is what it is being called, warn people” — routed to public health communication, harm-reduction services and clinicians.
Does drug-supply signal analysis identify people?
It is not monitoring of patients’ social media. Individuals enrolled in the program are not searched for, matched to accounts, or analyzed. If that is what an agency wants, this is not the product and we will say so.
It is not identification of sellers or buyers. The output is vocabulary and geography in aggregate. The program does not attempt attribution to individuals and does not build a capability to.
It is not covert collection. Only openly published material is analyzed. No accounts are created to gain access to closed spaces, and no private communications are accessed.
It is not an intelligence product for law enforcement. It exists so that people who use drugs can be warned about what is in the supply before it kills them.
If reading this page still leaves you uneasy, that is a reasonable response to the category, and the right thing for an agency to do is put it to counsel and to whatever community oversight structure exists locally. We would rather a deployment be narrowed than have it do something this page says it does not.
Questions this raises
Does IntellaRx monitor patients’ social media accounts?
No. This component analyzes publicly posted material in aggregate for drug-market vocabulary and geography. Individuals in the program are not searched for, not matched to accounts and not analyzed. Building that capability is outside the scope of this program. See what supply signal analysis actually looks at.
Who receives the output?
Public health communication, harm-reduction services and clinicians — so that people who use drugs can be warned about what is circulating. The product is a warning, not an intelligence lead. See what we do and do not do.
Why can’t agencies rely on official overdose reports alone?
The substances killing people change faster than any official reporting cycle. An adulterant can enter a local supply, kill people for months and appear in a published dataset a year later, and by then the pattern has moved. Signal analysis looks for the change while it is happening, then confirms it against toxicology and overdose data before anyone acts on it.
Why do street names for drugs matter?
What a substance is called locally changes constantly, and it changes so that it is not recognized. A new street name, or a shift in what an existing name refers to, is often the first observable sign that a supply has changed composition. Tracking that vocabulary lets a harm-reduction worker, an emergency clinician and a public health analyst in the same county see they are describing the same thing.
What sources does drug-supply signal analysis use?
Openly published material only, analyzed in aggregate for vocabulary and geography. No accounts are created to reach closed spaces, no private communications are accessed, and no poster is identified or matched to a patient. A signal from these sources is treated as a hypothesis until toxicology and overdose data confirm it.
Does the analysis produce maps?
Yes, at the level of a region, never a person. A heat map of a county is a public-health instrument that shows where a substance or a new name for it is circulating. A map of an individual would be something else entirely, and this program does not produce one.
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.