COLUMBUS, Ohio – The largest study to date that quantifies and compares medical and illegal use of fentanyl in the United States suggests that illicit use is more than twice as common as medically based use.
The analysis also shows the risk of a nonfatal overdose is up to three-fold higher in illicit users compared to patients exposed to fentanyl in a medical setting or through a prescription.
Overall, the study found that illicit-source fentanyl exposure is linked to a significantly higher risk of opioid-related harmful outcomes than medical-source exposure.
Almost 300,000 people from a dataset of 300 million patients were divided into two groups based primarily on urine drug test results, with illicit-source users constituting 71.1% of the study cohort, compared to 28.9% medical-source users.
"Fentanyl is a big problem, but we don't actually know how serious illicit drug use is compared to medical use," said co-lead author Ping Zhang , professor of computer science and engineering and biomedical informatics at The Ohio State University. "Finding that illicit use has reached 70% among all fentanyl users is a surprise. And this is also a contribution of this paper. To make something that's been a black hole visible to the healthcare system is significant."
The analysis also showed that among all states, illicit fentanyl use is most common in Ohio.
The study was published Aug. 14, 2026, in eClinicalMedicine .
Data on patients screened for fentanyl exposure from 2015 to 2023 came from Epic Cosmos, a large U.S. electronic health record dataset. First author Seungyeon Lee , a PhD student in Zhang's lab, conducted the analysis linking urine drug test results with medication records to group patients by medical or illicit use.
"Inferring or characterizing individual-level illicit fentanyl exposure poses significant challenges because illicit use often occurs outside the clinical setting and therefore is not directly observed in clinical data, so it's difficult to measure the scale and impact," Lee said. "This study focused on identifying illicit-source fentanyl use and its risk for opioid-related harmful outcomes."
Among patients in the Epic dataset who met study criteria, people who tested positive for fentanyl and had received a prescription for the drug within 72 hours before the urine test were categorized as medical-source patients (85,535), and people who tested positive but had no recent fentanyl prescription composed the illicit-source use cohort (210,193).
Fentanyl is a synthetic opioid that's up to 50 times stronger than heroin and 100 times stronger than morphine. Its main clinical use is to treat severe pain, especially after surgery and for advanced cancers. Fentanyl is among the synthetic opioids that contribute to nearly 70% of overdose deaths, according to the U.S. Centers for Disease Control and Prevention.
In addition to categorizing fentanyl users, Lee conducted analyses to examine patterns of adverse outcomes in both illicit and medical-source users that emerged over time from 2015 to 2023 and to estimate short-term risks within 30 days of fentanyl initiation.
The negative outcomes linked to fentanyl use included nonfatal overdose, abuse or misuse, and dependence that leads to withdrawal symptoms when use stops. Fatal overdoses were not included in this work because causes of death were not part of the Epic dataset.
The temporal analysis showed illicit fentanyl users had a higher prevalence of all three negative outcomes compared to medical users, and by 2023, the prevalence of nonfatal overdose was four-fold higher for illicit users than for medical-source fentanyl patients, at 18.6% of illicit users compared to 4.3% of medical users.
In the short-term risk analysis, after adjusting for a variety of health conditions and demographics, results showed that illicit-source fentanyl initiation was associated with a two- to three-fold higher risk of nonfatal overdose, abuse and dependence within 30 days compared to medical-source initiation.
"These patterns indicate that illicit-source fentanyl remains a significant and growing threat to public health nationwide," the authors wrote.
The analysis also showed that, compared to medical exposure to fentanyl, illicit fentanyl users were younger and less likely to be married, had a higher concentration of fentanyl detected in their urine drug test, and were more likely to have a mental health or substance use disorder.
Medical-source users were more likely than illicit-source users to have a chronic condition, and patients with chronic conditions were more likely to have more frequent healthcare utilization and greater exposure to opioid medications.
Geographically, Ohio had the highest proportion of illicit-source fentanyl users – 12.2% – followed by Massachusetts (10.5%) and California (8.8%). Medical-source exposure was most common in North Carolina (10.6%), Ohio (10.5%) and California (10.3%).
Opioid regulations have been focused on medical-source opioid drugs, including fentanyl, but the researchers say the availability of new data on the widespread illicit use of fentanyl and related health dangers highlights a need for additional policy action.
For example, the examination of patient data revealed a transition from medical-source fentanyl use to illicit-source use among some patients. The Zhang lab's next step will be analyzing patterns related to this transition, with hopes that the findings could assist with policymaking strategies.
This work was funded by the National Institute on Drug Abuse and the U.S. National Science Foundation.
Zhang is also a core faculty member in the Translational Data Analytics Institute at Ohio State. Additional co-authors were Richard Urman of Ohio State and Wenyu Song and David Bates of Harvard Medical School.
#
