The call came at 2 a.m. Not about a new patient. About one who had left.
She had been in my department three days earlier. Eighteen years old. Came in with chest tightness after a night out. Tox screen negative. Discharged. Three days later, fentanyl overdose. Dead. I have replayed that visit more times than I can count.
New data published in Emergency Medicine News puts a number on how many families are living that same story. In a study combining King County Medical Examiner records with University of Washington medical records, researchers analyzed 386 overdose deaths in adolescents and young adults between 2018 and 2024. One in five, 21 percent, had been seen at a health care facility in the 7 days before they died, and 17 percent had been seen within 72 hours.
We had them. And in most cases, we sent them home with nothing that could have changed the outcome.
Real physician voices, twice a week
Free, and one click to unsubscribe.
Only 27 percent received take-home naloxone at their last visit. Only 2.5 percent received buprenorphine. These numbers reflect a national pattern: Adolescents are the age group least likely to receive bystander naloxone in the event of an overdose, and least likely to be prescribed buprenorphine relative to any other age group.
This is not a resource gap. Fentanyl was present in 77 percent of deaths in this cohort. Most of these young people were not long-time opioid users. They were taking what looked like a normal pill, something bought at a party, pressed to look like Adderall or Xanax, laced with a dose of fentanyl high enough to stop breathing. The drug supply has shifted beneath our feet. Our clinical reflexes have not kept up.
Here is what keeping up looks like at the bedside.
Any adolescent who arrives intoxicated with any substance, not just opioids, is a naloxone candidate. Not a referral. Not a pamphlet. A kit, with a live demonstration, before the door closes. The research on this is clear. The barriers are not clinical; they are time, habit, and an underestimation of who is actually at risk.
Any adolescent using stimulants, benzodiazepines, or anything in pill or powder form deserves fentanyl test strips. This is harm reduction in the most literal sense. Giving a teenager the ability to test their supply does not endorse use. It acknowledges the reality of contamination and gives them a tool to survive long enough to make a different choice.
Any adolescent with a pattern of opioid use causing harm deserves a real conversation about buprenorphine, not a referral to an outpatient program they will not attend, but a prescription, initiated during the ED visit, before they leave. ED-initiated buprenorphine in adolescents is safe. The evidence supports it. The barriers are discomfort and a persistent belief that addiction treatment belongs in someone else’s clinic.
There is no someone else’s clinic for many of these patients. There is us, and there is a waiting room, and there is a window that closes when they walk out the door.
The Seattle data are not an indictment of any particular system. They are a mirror. Emergency departments, urgent care centers, primary care offices, anywhere an adolescent shows up, are handling the patient in front of them and missing the risk behind them.
I cannot change what happened three days after that eighteen-year-old left my department. I can change what happens before the next one leaves.
We have the tools. Naloxone reverses overdoses. Buprenorphine prevents them. Fentanyl test strips buy time. None of them require a referral. None of them require a specialist. All of them can happen before discharge.
The question the data force is simple: What did we do with the visit?
Stephen M. Sandelich is a pediatric emergency physician.
