On a cold December night in 2023, an ambulance brought a 26-year-old man named Jean Descamps to Providence Milwaukie Hospital in Oregon, after sheriff's deputies found him at a light-rail station. He was homeless and emaciated. Staff evaluated him, gave him Narcan, and prepared to send him back out. On body camera footage later released by the Milwaukie Police Department, a hospital staff member can be heard saying he had no medical problems and that his condition was "all behavioral. Yeah, hundred percent behavioral." NPR, which reported the case this month, could not identify who was speaking. The officers on the scene said on tape that they were not comfortable with the discharge, and one of them said, "That guy does not need to be dead right now." Descamps left the hospital at 10:49 that night and was dead by 11:31, in police custody during a transport to another facility, of what the medical examiner's preliminary findings described as likely related to a drug overdose with contributing natural causes. The hospital later acknowledged that it "fell short of our goal of providing safe, reliable, compassionate care to our patient."
I have spent more than thirty years in this field, and I want to stay with one word from that recording, because it is the word that does the damage: behavioral.
What Clinicians Mean by "Behavioral," and Why It Is Wrong Here
In hospital shorthand, calling a presentation behavioral says the problem is not in the body, and therefore not really the hospital's problem. It is a triage judgment dressed up as a diagnosis. Applied to a person in withdrawal or intoxication, it carries a second meaning nobody writes in the chart: this person is doing this to himself, and he could stop.
That second meaning is not supported by the imaging. In 2023, a team led by Jacob Stubbs and Michael Fox published an analysis in Nature Mental Health pooling 144 neuroimaging studies that covered 9,047 participants. Findings across different substances had never lined up cleanly, and that inconsistency was used for years as an argument that the neuroscience was soft. What Stubbs and colleagues showed is that the scattered abnormalities converge on a single brain network connected to the anterior cingulate cortex, the insula, the dorsolateral prefrontal cortex, and the thalamus. Brain lesions that cause people to spontaneously stop smoking intersect that same network. I would not call this proof that substance use causes the differences, because the underlying studies are cross-sectional. It does establish that there is something physical and consistent to find.
The structural work points the same direction. The ENIGMA Addiction Working Group pooled scans from 23 laboratories in 14 countries, covering 3,240 people, 2,140 of them with substance dependence, and found lower gray matter volume and cortical thickness in the dependent group relative to non-dependent controls, with the insula and medial orbitofrontal cortex implicated regardless of which substance the person used. Those regions handle interoception, valuation, and decision-making, and the convergence is the one I described in the universal brain signature of addiction.
The Control Problem Is Physical
Nora Volkow, who directs the National Institute on Drug Abuse, and Carlos Blanco laid out the mechanism in World Psychiatry in 2023. Every drug with addictive potential raises dopamine in the nucleus accumbens. With repeated use, dopamine neurons shift toward firing in response to cues that predict the drug rather than to the drug itself, which is why a street corner or a particular song can produce craving years into recovery. Repeated use also tends to produce prefrontal impairments that weaken self-control and increase impulsivity.
Read those two changes together. The system that generates wanting has been made more sensitive, and the system that overrides wanting has been made weaker. Asking a person in that state to simply decide differently is asking them to do something their neural hardware has been altered to make harder. The brain disease model is not unanimous in my field. Marc Lewis has argued that it over-medicalizes what is better understood as learning and can leave patients feeling that change is impossible, and Carl Hart has argued that the model props up discriminatory drug policy. I take both critiques seriously, but my argument here is narrower: whatever else addiction is, it is not adequately described by the word behavioral.
We Have Treatments. We Do Not Give Them.
Here is the part that should bother every clinician reading this. A meta-analysis by Thomas Santo and colleagues in JAMA Psychiatry, covering 36 cohort studies and 749,634 people with opioid dependence, found that during periods when people were receiving methadone or buprenorphine, all-cause mortality was less than half what it was during periods off treatment, and drug-related death was roughly 59 percent lower. In a separate study of 40,885 Americans aged 16 and older with opioid use disorder, Sarah Wakeman and colleagues found that only buprenorphine and methadone were associated with reduced overdose. Inpatient detoxification and behavioral health treatment alone showed no measurable protective effect, and neither did naltrexone.
These are observational data, and confounding by indication inflates the apparent benefit to some degree. Withholding an agonist medication in a randomized trial would be unethical, so this is the strongest evidence we are going to get, and it is strong enough. Yet the 2025 National Survey on Drug Use and Health found that of the 47.2 million Americans aged 12 and older who needed substance use treatment, about 16 percent received any. More than eight in ten got nothing.
Stigma Is Measurable, and It Has Been Measured
Judy Chertok, who treats patients and teaches addiction medicine at the University of Pennsylvania, put it to NPR this way: "We would not allow, as a society, people with heart attacks to come to a hospital and not get appropriate treatment. This is as serious as that."
That is not only an impression. Catherine Parish and colleagues surveyed 1,240 primary care physicians, emergency physicians, and dentists drawn from a nationally representative sample, in work published in Addiction in 2025. Stigma scores toward opioid use disorder averaged 35.6 and toward stimulant use disorder 36.3, against 26.2 for depression, 25.8 for HIV, and 23.2 for type 2 diabetes. More than 30 percent said they would prefer not to work with patients who have opioid or stimulant use disorder, compared with 2 percent for diabetes, and 22 percent agreed that there is little they can do to help patients like this. Robert DeForde of Shatterproof, which works to get clinicians to treat opioid addiction, told NPR that this combination of stigma and disbelief in FDA-approved medications is "a fatal flaw in our system."
What This Means for You or Someone You Love
If you are the patient or the family member, the most useful thing you can do is decline the framing. When a clinician calls a presentation behavioral, it is fair to ask what was ruled out and whether a toxicology screen was run. Ask directly whether buprenorphine or methadone is on the table, because those two carry the strongest mortality evidence and the alternatives have not been shown to match it. If the answer is that this facility does not do that, that is an answer about the facility, not about you.
For clinicians, the implication is smaller than it sounds. You do not need to become an addiction specialist. You need to stop treating a diagnosis of substance use disorder as a reason to stop looking.
Our Rescue From Rehab program exists because the conventional system keeps repeating the error captured on that body camera. We start from imaging, biomarkers, and nutritional status rather than from an assumption about willpower, and we treat metabolic and nutritional repair as part of the work rather than an afterthought, which is why brain nutrition through Action Potential Supplements is built into the protocol. That is a statement about how we practice, not a claim that nutrition substitutes for medication. It does not.
Jean Descamps was 26 years old. He was not faking, and he was not choosing. He had a treatable disorder of a brain network researchers can now map, and he was sent out the door because someone decided his suffering was a matter of conduct. Every patient we meet with dignity instead is a correction of that error.
Illustration created with AI image tools.
This article was drafted with the assistance of AI writing tools, then reviewed, edited, and approved by Dr. Drew W. Edwards, Ed.D., who holds full editorial responsibility for its content.
Sources: Mann B. Hospital staff said he was faking it and released him. He died from an overdose soon after. NPR. August 6, 2026. npr.org · Stubbs JL, et al. Heterogeneous neuroimaging findings across substance use disorders localize to a common brain network. Nature Mental Health. 2023. DOI: 10.1038/s44220-023-00128-7 · Mackey S, et al. Mega-analysis of gray matter volume in substance dependence. American Journal of Psychiatry. 2019;176(2):119-128. DOI: 10.1176/appi.ajp.2018.17040415 · Volkow ND, Blanco C. Substance use disorders: a comprehensive update. World Psychiatry. 2023;22(2):203-229. DOI: 10.1002/wps.21073 · Santo T Jr, et al. Association of opioid agonist treatment with all-cause mortality. JAMA Psychiatry. 2021. DOI: 10.1001/jamapsychiatry.2021.0976 · Wakeman SE, et al. Comparative effectiveness of different treatment pathways for opioid use disorder. JAMA Network Open. 2020;3(2):e1920622. DOI: 10.1001/jamanetworkopen.2019.20622 · Parish C, et al. Stigma toward substance use disorders among US clinicians. Addiction. 2025. DOI: 10.1111/add.70122 · SAMHSA. 2025 National Survey on Drug Use and Health. July 2026.