--- title: "Opioid deaths and overdose calls in Hamilton: why the two counts moved in opposite directions" kind: "Article" topic: "Drugs & public health" date: "2026-10-10" url: "https://canopticon.com/articles/opioid-deaths-and-overdose-calls-in-hamilton" author: "Chris Telford" license: "CC BY 4.0" --- On Hamilton's streets, people sit folded forward for hours — not unconscious, not awake. Outreach workers have a name for it. The official measure of the overdose crisis, deaths, has been falling since 2021. This article argues that both are true, and that the falling number is the misleading one. ## Deaths fell while overdose calls more than doubled Hamilton's own figures show the split. Opioid deaths went 167 (2021) → 166 (2022) → 150 (2023) → 129 (2024). Over the same stretch, paramedic calls for suspected opioid overdoses went from an average 61 (May 2024 – March 2025) to an average 134 (May 2025 – March 2026). The city's medical officer of health said it plainly: "Many overdoses go unreported due to the prevalence of the overdose-reversal drug naloxone, meaning the real number is likely higher." The provincial picture matches. Ontario opioid deaths fell from 2,639 in 2023 to 2,231 in 2024 — a 15% drop, and the chief coroner has said the reason is uncertain: possibly a slightly less toxic supply, possibly naloxone saturation, possibly a shift to inhalation. Hamilton's own rate runs high — roughly 40–45% above the Ontario provincial average for years ## Street fentanyl is now mostly cut with something else The drug has changed. In one Toronto testing fortnight in 2025, the share of fentanyl samples containing a benzodiazepine reached **43%** benzo-positive. The deaths show it too: Ontario's Office of the Chief Coroner reports that non-prescription benzodiazepines were detected in 62% of opioid toxicity deaths in 2024. The mix moves fast. Toronto's checking service found, in one earlier fortnight, 15% of samples sold as fentanyl contained no fentanyl at all, and later in the year the supply had been dominated by medetomidine (a veterinary tranquilizer) at 67% while benzos dropped to ~14%, then benzos rebounded. British Columbia's testing shows the same churn: 44.7% of all opioid-down samples in December 2025 contained a non-fentanyl active ingredient. At the national peak, "Overall NMB detection peaked in December 2023 with nearly 60% of opioids testing positive." Fentanyl itself is unusually unforgiving on the street. The gap between a dose that gets someone high and one that kills is narrow; the active dose (a few hundred micrograms) and the fatal dose (1–3 milligrams in an opioid-naive person, much higher in someone tolerant) are separated by only about a factor of ten, and street mixing is uneven: A "hot spot" in a poorly-mixed batch — a single grain extra — can be the difference between getting high and dying. ## Naloxone reverses the opioid, not the sedative Naloxone, the overdose-reversal spray, works on opioids only. Bromazolam, like all benzodiazepines, does not respond to naloxone — naloxone reverses the opioid component, restoring breathing, but the user remains profoundly sedated for hours. The same is true of the veterinary sedative mixed in alongside it: Like benzodiazepines, xylazine does not respond to naloxone. That sedation is what outreach workers call "the bend" or "tranq lean": users in a forward-folded posture, not quite unconscious, sometimes for many hours, vulnerable to robbery and sexual assault. So a reversal that once ended an overdose now starts a long, dangerous wait. The post-naloxone period is therefore far longer and clinically messier than in the heroin era. The sedatives are added on purpose. One drug-checking project explains that xylazine extends the 'legs' or duration of a high. The research calls it: A market-rational adulteration, not random poisoning. The benzodiazepines extend the "high" by 6–12 hours rather than fentanyl's 1–2, users effectively spend much more of their day immobile. The services built for overdose were not built for this: supervised consumption services (SCS) and overdose prevention sites (OPS) have not been designed to provide ongoing support to people experiencing prolonged sedation ## What people on the street are actually seeing The visible decline that gets called "zombie" is several things at once. The research names the convergence of four pharmacologically and socially distinct processes. The wounds come from xylazine, which activates peripheral alpha-2b receptors on arteriolar smooth muscle and starves the skin of blood; crucially, the effect does not depend on injecting: ulcerations can occur remotely from an injection site. A Toronto toxicologist's clinical advice is that "Naloxone reverses opioid toxicity but has no effect on the sedating properties of xylazine, which may persist." Stimulants supply the rest of the picture. In Ontario in 2024, stimulants were detected in 69% of opioid toxicity deaths. Cycling between the sedatives and methamphetamine produces the gaunt, agitated, prematurely-aged appearance. Housing finishes it: Hamilton's actively homeless population has exceeded 1,500 since December 2021, and a wound that might heal in a housed patient becomes chronic and limb-threatening in someone sleeping in a tent in February. The research is explicit that this is new rather than a perception problem: The current picture — dissociated catatonia, large necrotic wounds, methamphetamine-driven agitation, profound sleep deprivation, and rapid physical aging — is genuinely new > **Objection: Fewer deaths is still good news** > > It is. Naloxone is saving lives — BC pioneered a take-home naloxone program (now over 400,000 kits annually) that has clearly saved lives. The problem is not the falling deaths but treating them as the whole story. A measure built for the heroin era counts the one outcome naloxone now prevents, and misses the ones it cannot. > > The federal health agency's own analysis credits it most: expanded naloxone distribution as the single most likely factor in the modest 2024 mortality decline, with regions distributing more kits experiencing larger drops in deaths. ## Deaths are the top of a pyramid Ontario's chief medical officer of health had a name for this years ago: "opioid injury pyramid" — for every death there are exponentially more non-fatal overdoses, ED visits, and chronic injuries. The base of the pyramid has been growing. In Ontario, "paramedic use of those who call 911 ... transport to the ER for emergency care have increased nearly fourfold." The research ties the paradox to the pharmacology in a sentence: Naloxone is preventing acute death but not the slow, semi-conscious, wound-accumulating decline. ## The system that watched the supply was cut back The drug-checking service that produced the numbers above lost its collection points when Ontario closed consumption sites. Its own report recorded a **50% reduction in samples collected** in the first half of April vs. March — in the same fortnight the benzodiazepine share hit its high. The research puts it simply: the surveillance system was decapitated at the moment the supply got worse. The closure was part of a provincial change. Hamilton's site closed under a law that banned consumption sites within 200 metres of schools or daycares, and the replacement hubs explicitly prohibits needle exchange, drug checking, and supervised consumption. The province's auditor found the closures were "made without proper planning, impact analysis or public consultations." The hubs were not ready when the sites closed; a deputy minister told a committee that "all of the HART hubs were not ready. There are 14 of them that are up and running with a full suite of services. Now, there are some that have some services." The research's reading of what changed on the street is direct: The drug-poisoned supply has not changed; what changed is that drug use has been pushed out of supervised indoor settings and into parks, alleys, libraries, and bus shelters. The city library's main branch considered temporarily closing the branch because of overdoses inside it. ## What this explains It explains how a city can report fewer deaths while its paramedics report more calls than they have seen in years: naloxone turns deaths into survivals, and the sedatives it cannot touch turn survivals into hours of danger. It explains why the people folded over on the street are not a sign of the crisis ending but of the drug changing. And it explains why the death count, the number politicians cite, is the wrong instrument for this supply. A crisis that has learned to stop killing as often has not stopped. It has stopped showing up where we look.