A nationwide analysis of nearly 300,000 patients suggests that illicit fentanyl exposure has become substantially more common than medically supervised use, offering one of the clearest comparisons yet between the two sources of exposure in the United States. The study, published in eClinicalMedicine, found that 71.1% of the patients identified as fentanyl users had no recent prescription for the drug, while 28.9% had been exposed through medical care or a documented prescription. The findings also indicate that illicit-source users face markedly greater risks of nonfatal overdose, fentanyl misuse and opioid dependence than patients exposed to fentanyl in clinical settings.
Researchers at The Ohio State University and Harvard Medical School analyzed electronic health records from Epic Cosmos, a large U.S. database containing information from approximately 300 million patients. Their study focused on individuals who underwent urine drug testing between 2015 and 2023 and tested positive for fentanyl. By linking toxicology results with prescription and medication records, the team created two broad categories of exposure. Patients who tested positive and had received fentanyl within the preceding 72 hours were classified as medical-source users. Those who tested positive but had no recent prescription were categorized as illicit-source users.
That classification provides an indirect way to detect fentanyl exposure that may otherwise remain invisible to healthcare systems. Illicit fentanyl use frequently occurs outside hospitals, clinics and pharmacies, meaning it is rarely recorded in conventional medical histories. A positive urine test, when considered alongside prescription records, can therefore reveal exposure that is not documented through routine prescribing data. The researchers caution that the approach cannot determine precisely how a person obtained fentanyl, but it offers a large-scale method for distinguishing likely medical exposure from exposure that occurred outside a recent treatment context.
Of the 295,728 patients included in the analysis, 210,193 were placed in the illicit-source group and 85,535 in the medical-source group. The difference was more than a statistical curiosity: the two populations also displayed substantially different clinical profiles. Illicit-source users tended to be younger, less likely to be married and more likely to have a diagnosed mental health or substance use disorder. They also had higher fentanyl concentrations in urine samples, a finding that may reflect differences in dose, frequency of use, drug potency or the timing of exposure before testing.
Medical-source patients, by contrast, were more likely to have chronic health conditions and to use healthcare services frequently. This pattern is consistent with fentanyl’s legitimate medical applications, including anesthesia, treatment of severe acute pain and management of pain associated with advanced cancer. In hospitals, fentanyl is administered in controlled doses and monitored by trained professionals, although medical exposure can still carry risks. The contrast between the groups highlights how the same pharmaceutical compound can produce very different dangers depending on dose, purity, context and access to emergency care.
The most striking differences appeared in the study’s analysis of harmful outcomes over time. By 2023, 18.6% of illicit-source users had experienced a documented nonfatal overdose, compared with 4.3% of medical-source patients. That represents more than a four-fold difference in prevalence. Illicit-source users also showed higher rates of recorded fentanyl abuse or misuse and dependence accompanied by withdrawal symptoms. Fatal overdoses could not be evaluated because the Epic Cosmos records did not include complete causes of death, meaning the study’s estimates describe only harmful outcomes that were captured in clinical data.
The researchers also examined what happened during the first 30 days after a patient’s initial identified fentanyl exposure. After adjusting for demographic characteristics and existing health conditions, illicit-source initiation was associated with a two- to three-fold higher risk of nonfatal overdose, abuse and dependence than medical-source initiation. This short-term analysis is important because fentanyl can produce rapid changes in tolerance and dependence, while illicit products may vary unpredictably in strength and composition. A person may also be unaware that fentanyl is present in a drug obtained outside the medical system, making accidental exposure and overdose more likely.
Geographic differences added another dimension to the findings. Ohio had the highest proportion of illicit-source fentanyl users among the states represented in the analysis, at 12.2%, followed by Massachusetts at 10.5% and California at 8.8%. Medical-source exposure was most common in North Carolina, at 10.6%, followed by Ohio at 10.5% and California at 10.3%. These figures describe the distribution within the study population rather than the overall prevalence of fentanyl use in each state, but they nevertheless point to important regional differences in exposure patterns, healthcare access and the changing geography of the opioid crisis.
The study also detected evidence that some patients moved from medical-source exposure to illicit-source use. The researchers say this transition deserves closer investigation because it may reveal how dependence, inadequate pain treatment, disrupted access to care or exposure to illicit drug markets can interact. The findings suggest that policies aimed primarily at regulating prescribed opioids may not be sufficient to address the current fentanyl threat. More effective public-health strategies could require expanded toxicology screening, earlier identification of substance use disorders, wider access to naloxone and treatment, and surveillance systems capable of recognizing illicit exposure before it results in a fatal overdose. The authors describe illicit-source fentanyl as a growing nationwide danger and plan to study the pathways that lead patients from medical exposure into illicit use.
Subject of Research: Medical versus illicit fentanyl exposure and associated opioid-related harms in the United States.
Article Title: Longitudinal patterns of fentanyl utilisation (medical versus illicit sources) in the United States 2015–2023: a retrospective cohort study
News Publication Date: 14-Aug-2026
Web References: https://doi.org/10.1016/j.eclinm.2026.104141
References: Lee S, Zhang P, Urman R, Song W, Bates D. “Longitudinal patterns of fentanyl utilisation (medical versus illicit sources) in the United States 2015–2023: a retrospective cohort study.” EClinicalMedicine. 2026; DOI: 10.1016/j.eclinm.2026.104141.
Keywords
Fentanyl, illicit drug use, opioid overdose, nonfatal overdose, opioid dependence, substance use disorder, electronic health records, Epic Cosmos, public health, synthetic opioids, United States, toxicology screening
Tags: analysis of electronic health records for fentanyl exposureclinical vs. illicit sources of fentanylcomparison of medical and illicit fentanyl exposurefentanyl misuse and overdose statisticsfentanyl testing and toxicology dataillicit fentanyl use prevalenceimpact of illicit fentanyl on public healthnationwide study on fentanyl exposureopioid dependence from illicit fentanylopioid epidemic and drug misuse trendsrisks of nonfatal fentanyl overdosetrends in fentanyl use in the United States
