On 24 September 2024, the Government of Alberta announced the overdose prevention site closure in Red Deer Overdose Prevention Site would close. The site had been operating since 2018, offering a supervised environment where people could use illicit drugs without immediate risk of dying from an unwitnessed overdose. Six months later, on 31 March 2025, it shut its doors for the last time.

The announcement was met with fierce opposition. Advocates warned that without the site, people would die. Overdose deaths would rise. Healthcare systems would buckle. The closure, they argued, was ideologically driven and evidence-free, a political act dressed as a policy decision.

Those are serious claims. They deserve serious scrutiny.

A research team at Recovery Alberta responded with exactly that. Using linked provincial administrative health data, they tracked 381 identifiable Red Deer OPS clients before and after the overdose prevention site closure, measuring what actually happened to real people across a range of health outcomes. Their findings, published in the peer-reviewed journal Addiction, do not fit the script that either side of this debate had prepared.

What they found was this: when the site closed, more people entered treatment. The predicted surge in overdose deaths did not materialise. And the evidence, carefully read, raises questions about supervised consumption services that the harm-reduction lobby has been slow to answer.

This commentary examines what the study found, what it means, and why it matters for how we think about drug policy, recovery, and the people at the centre of both.

The Conversation We Are Not Allowed to Have

There is a story that dominates public debate on drug policy. It goes like this: people who use drugs are trapped, powerless, and beyond the reach of recovery until the conditions are perfectly right. The most we can do is keep them alive while they wait. Any suggestion that people can, and do, choose to exit drug use is dismissed as moralising, stigmatising, or politically motivated.

This story is not just incomplete. In important ways, it is cruel.

It denies agency to the very people it claims to protect. It treats recovery as an exceptional outcome rather than a reasonable aspiration. And it has been used to build a policy infrastructure (supervised consumption sites, managed use frameworks, harm-reduction-as-endpoint ideology) that in some configurations makes it easier to keep using than to stop.

A new study published in Addiction has quietly produced evidence that cuts against this story. It deserves far more attention than it has received, not because it settles the debate, but because buried inside its data is something the harm-reduction establishment has been conspicuously reluctant to celebrate.

Some people, when the supervised use environment was removed, chose recovery.

That is worth saying out loud. Loudly.

What the Study Found

Before the closure announcement, around 9.9% of Red Deer OPS clients were receiving opioid agonist therapy (OAT) in any given week. After the overdose prevention site closure, that figure rose to 16.1%. A 63% relative increase. Statistically significant. Sustained across the entire six-month post-closure follow-up period. Significantly higher than the 14.4% observed at the comparison site in Lethbridge, where the OPS remained open and OAT uptake did not move.

When the supervised consumption environment was removed, more people entered evidence-based treatment for opioid dependence. Not fewer. More. OAT (buprenorphine-naloxone, methadone, extended-release buprenorphine) is among the most rigorously validated interventions in addiction medicine. Engagement with it reduces overdose risk substantially, supports physical stabilisation, and creates the conditions in which recovery becomes possible. More people accessing it is not a neutral data point. It is people moving toward better lives.

The catastrophe advocates predicted did not arrive. Emergency department visits: no statistically significant change. Suspected opioid-related EMS callouts: no statistically significant change. Mortality: five deaths in the 26 weeks after closure, compared with six in the 26 weeks before. In a community whose drug supply was becoming measurably more toxic, with carfentanil presence rising sharply in regional toxicology data, the predicted surge in preventable deaths did not occur.

The authors are appropriately careful. Six months is a limited window and statistical power for rare outcomes was constrained. These are genuine qualifications. But the direction of the evidence is unambiguous, and it points somewhere harm-reduction discourse rarely looks.

The Ideology That Has Been Substituted for Evidence

Harm reduction, in its original form, was pragmatic and defensible. For people in active addiction not yet ready to stop, interventions that reduce immediate risk can prevent death and preserve the possibility of future recovery. The evidence for some of those interventions is solid.

But that is not what harm reduction is anymore.

Over decades it has transformed from a clinical tool into a political movement, one that treats supervised consumption as a permanent right, frames recovery aspiration as coercive, and has built an entire service infrastructure organised around the ongoing management of drug use rather than its resolution. The goal of getting better has been quietly replaced by the goal of using more safely. For many of its loudest advocates, that is not a compromise. It is the point.

Ideology shapes funding, and funding shapes what gets built. When harm reduction dominates the policy conversation, money flows toward services that keep people alive in their addiction rather than toward services that help them out of it. Treatment waitlists grow. Recovery support is underfunded. The people who most need a pathway out are instead offered a more comfortable place to stay.

The bridge-to-treatment claim has been the movement’s central justification: that sites draw people in, build trust, and gradually connect them to recovery. It is the story that secured government contracts, academic endorsement, and media sympathy for two decades. The Red Deer data exposes it. If these sites reliably functioned as bridges, OAT uptake should not rise when the bridge is removed. It rose by 63%. That is not a bridge collapsing. That is evidence the bridge was not being used.

The evidentiary record underneath the advocacy has always been thinner than it appeared. A RAND Corporation assessment found the causal case for favourable supervised consumption outcomes to be minimal, with most studies limited by the absence of adequate control groups. Researchers Rehm and Fischer noted that weak study designs leave the entire evidence base open to alternative interpretations. The literature is large and almost unanimous in its conclusions, but it is also almost entirely observational, self-referential, and concentrated on a handful of sites whose evaluators were often ideologically aligned with the intervention.

The movement also rarely acknowledges that most people who use drugs and enter treatment want to stop. Survey data consistently show abstinence is the stated goal of the majority of people seeking help. Harm reduction has reframed that preference as naive, recasting the desire to be free of addiction as false consciousness imposed by a moralising society. In doing so, it has stopped listening to the people it claims to speak for.

Beyond the evidence, there is the normalisation effect. Supervised consumption environments send a signal to users, communities, and policymakers that ongoing drug use is a manageable and indefinitely sustainable state. That signal reduces the urgency of recovery. It makes staying where you are feel like a supported choice rather than a crisis requiring resolution.

The harm-reduction lobby will call this stigma. It is not. It is the same standard applied to every other health intervention: does this move people toward recovery, or does it make it easier to remain in harm?

Harm reduction, as currently practised, has become a system more comfortable with people continuing to use than with people getting better. For some of its architects, that is not a failure. It is the design.

Recovery Is Not a Dirty Word

Somewhere along the way, in the vocabulary of progressive drug policy, recovery became suspect. It acquired associations with abstinence ideology, with coercion, with the judgement of people in crisis. Advocates began treating the word itself as a form of stigma.

This is a profound mistake, and it has done real damage to real people.

Recovery is what the people in the Red Deer data moved toward when their circumstances changed. Recovery is what OAT supports. Recovery is what the shift from 9.9% to 16.1% represents: human beings taking a step toward health, autonomy, and a life not defined by dependence.

They deserve to be celebrated. Not managed. Not indefinitely maintained. Celebrated.

The people who choose recovery, often after years of trauma, failed attempts, and profound loss, are demonstrating something extraordinary. They are choosing their lives back. Every policy, every service, every dollar of public funding should be oriented around making that choice more available, more supported, and more likely to succeed.

A policy framework that treats the goal of recovery as optional, or as one legitimate outcome among many equivalent options, is not a neutral position. It is a failure of ambition on behalf of the people it claims to serve.

What the Evidence Demands from Policymakers

The Red Deer overdose prevention site closure has direct implications for how services are designed and funded.

Treatment infrastructure must precede service transitions. Alberta’s expansion of low-barrier OAT access (virtual prescribing, same-day initiation) appears to have made it possible for people displaced from the OPS to enter treatment relatively quickly. That is a lesson about sequencing, not a licence for reckless closures. Removing harm-reduction services without building robust, accessible treatment alternatives leaves vulnerable people with nothing. The obligation is to build the pathway first.

Services must be evaluated honestly against recovery outcomes. OAT uptake, treatment retention, mortality, functional recovery: these are the measures that matter. Services that cannot demonstrate movement toward these outcomes over time must be scrutinised, not protected from scrutiny by ideological ring-fencing. The people using those services deserve better than that.

Long-term follow-up is not optional. Six months of data is the beginning of an evidence base, not the end of one. The researchers are correct that continued monitoring is essential. Policymakers must fund that monitoring and commit to following the evidence wherever it leads, including toward conclusions that complicate existing service models.

The carfentanil finding matters. Red Deer’s drug supply was measurably more dangerous than Lethbridge’s during the study period. The absence of a mortality surge under those conditions does not mean closures are risk-free. It does mean the catastrophist predictions that preceded this supervised consumption site closure were not borne out by the available evidence.

Conclusion

The Red Deer OPS study does not resolve the supervised consumption debate. Six months of administrative data cannot do that. What it does is demonstrate, with genuine methodological rigour, something evidence-based prevention has long maintained: when people are given access to real treatment, many of them will choose it.

That choice is a person deciding their life is worth more than their next hit. Someone walking into a pharmacy for a script instead of into a site to use. A family getting someone back. A human being reclaiming their future.

Harm reduction has a role in keeping people alive long enough to reach that point. But survival is not the goal. Recovery is. Every service, every framework, every dollar of public funding must be measured against whether it moves people closer to that outcome or makes it easier to avoid it.

The Red Deer data suggests that for a significant proportion of one cohort, at one site, the supervised use environment was doing the latter.

That demands honesty about what harm reduction can do, what it cannot do, and what we owe the people who deserve more than managed decline.

Recovery is possible. It happens every day. It deserves to be the goal.

Reference: Day N, Kaufmann K, Devoe DJA, et al. Healthcare utilization and mortality after overdose prevention site closure: A linked cohort analysis using segmented difference-in-differences time series. Addiction. 2025.