Supervised consumption rooms: what the evidence says

A woman enters a small, clean, well-lit room, picks up sterile material from the counter and sits down in one of the booths. A nurse watches from a distance. Minutes later, the woman becomes drowsy, her breathing slows and her lips begin to turn bluish. The nurse arrives with oxygen and naloxone, and the woman responds again before anyone needs to call an ambulance. The scene is fictional and was written for teaching.

This text deals with supervised consumption rooms, also called supervised injection sites or overdose prevention centers: spaces where a person uses drugs obtained elsewhere in the presence of a team trained to intervene if something goes wrong. The goal is to look at what studies show about overdoses, mortality, infections, crime in the surrounding area and access to care. If you play the case of the day on NEXO, you will recognize the type of reasoning: weighing different outcomes, separating what was measured from what was estimated and understanding the context before concluding.

Key points

• The rooms are spaces where a person uses drugs obtained outside, with a team ready to respond to an overdose.
• The model emerged in Europe in the 1980s and today exists on several continents.
• In Vancouver, frequent use of the room was associated with lower overall mortality.
• Studies of the surrounding area found trafficking and assaults stable after the opening.
• Population effects vary according to the outcome measured and the location.
• Safety, trust and social inclusion appear as the path through which the rooms work.

What a supervised consumption room is

A realist review with 391 documents defined overdose prevention centers as non-residential spaces where people use illicit drugs obtained elsewhere, in the presence of a team able to prevent and manage overdoses (Stevens et al., 2024). The room offers supervision; the drug comes from outside. The centers can receive people who inject or who inhale, and also offer guidance on injection technique, wound and infection care, drug content checking, primary care and referral to housing and addiction treatment (Stevens et al., 2024).

These spaces are usually open to people of any social background, but they are used mainly by those living different forms of exclusion, such as homelessness or unstable housing (Stevens et al., 2024). This profile helps explain why the outcomes studied go beyond overdose and include infections, housing and connection with care.

In Vancouver, authorized rooms operate with nursing supervision, and the person finds sterile injection material there, immediate response to overdose and referral to other programs, inside and outside the service (Kerr et al., 2017). An addiction counselor and nursing care were also available on site at the first room in North America (Wood et al., 2006).

The name varies from one place to another. Studies speak of supervised injection room, supervised consumption service, overdose prevention site and overdose prevention center. In British Columbia, overdose prevention sites are usually operated by peers, people with lived experience of drug use who offer first aid to prevent and respond to overdoses (Dogherty et al., 2022).

For the clinician, the definition matters because each model has a different team, schedule and audience. This logic connects with harm reduction beyond all or nothing: the immediate goal is to keep the person alive and in contact with the service, without requiring them to stop using before being cared for.

Overdose prevention centers began operating in Switzerland and Germany in the mid-1980s, spread across continental Europe and later reached Australia, Canada, Mexico, Colombia, Iceland, Scotland and the United States (Stevens et al., 2024).

In Canada, the road was long. In Vancouver, in the late 1990s, the annual incidence of HIV infection reached 19 per 100 people who injected drugs, and the province recorded more than 300 fatal overdoses, which led the health authority to declare a public health emergency (Kerr et al., 2017). Before the official room, an unauthorized site called Back Alley received more than 100 people per night, according to attendance lists (Kerr et al., 2017).

The municipal strategy of 2000 and 2001 was based on Western European models that sought to balance prevention, enforcement, treatment and harm reduction (Kerr et al., 2017). The authorized room opened in 2003 and spent more than a decade with pilot project status, until political changes created a more favorable environment and other cities began opening their services (Kerr et al., 2017). Along the way, the room's continuation ended up in the courts, and the Supreme Court of Canada ruled 9 to 0 in favor of keeping the service (Kerr et al., 2017).

Starting in 2016, more than 30 low-threshold overdose prevention sites, authorized by the province, were set up in Canada (Kennedy et al., 2019). In British Columbia, the public health emergency declaration in April 2016 allowed these sites to open in several parts of the province (Panagiotoglou, 2022).

What studies show about overdose and mortality

The most direct effect of a room is the response to overdose at the moment it happens. Between March 2004 and July 2008, the Vancouver room recorded 1004 overdoses, of which 453 met the definition of potentially fatal, because they required naloxone, an emergency call or an ambulance (Milloy et al., 2008). Based on these data, the authors estimated between 8 and 51 deaths avoided in the period, equivalent to 1.9 to 11.7 per year (Milloy et al., 2008).

This estimate depends on assumptions. Because the literature reported very different rates of nonfatal overdose, the authors tested three scenarios, and the number of deaths avoided varied considerably among them (Milloy et al., 2008). Even so, by design the room received about 5 out of every 100 daily injections in the neighborhood (Milloy et al., 2008).

The next question is about all-cause mortality. In a Vancouver cohort of 811 people who injected drugs, 432 reported frequent use of the room, at least weekly (Kennedy et al., 2019). The crude mortality rate was 22.7 deaths per 1000 person-years, and frequent use of the room was associated with a lower risk of death from any cause, with an adjusted hazard ratio of 0.46 (Kennedy et al., 2019). At the start of follow-up, the median age was 39 years, and 278 participants (34.3 per 100) were women (Kennedy et al., 2019). The association held after adjustment for age, sex, HIV, unstable housing, daily cocaine injection, public injection, incarceration and addiction treatment (Kennedy et al., 2019).

In Victoria, also in Canada, professionals and users interviewed pointed to the absence of overdose deaths as the main impact and success indicator of the prevention sites (Pauly et al., 2020). There were 15 professionals and 12 users interviewed (Pauly et al., 2020).

Population-level effects are harder to measure. In British Columbia, 25 sites opened in 14 of the 89 local health areas between 2015 and 2017, and the analysis with control areas showed a drop in paramedic calls and opioid-related emergency department visits (Panagiotoglou, 2022). In the same analysis, local hospitalization and mortality rates showed no detectable change (Panagiotoglou, 2022). In paramedic calls, the drop was preceded by an initial increase right after opening, which the author considered expected, because the original guidance recommended calling emergency services for every overdose reversed with naloxone (Panagiotoglou, 2022). The detail shows how an operational rule can change an indicator without the real risk having changed.

This contrast recalls a basic rule for reading studies: the outcome measured changes the conclusion. The reasoning is similar to thinking about the base rate in diagnosis, since a real effect inside the room can be diluted when you look at an entire region where most people use elsewhere.

Infections, crime and public order

Injection-related skin infections, such as abscesses and cellulitis, are common and preventable. In a cohort of 1065 participants from the Vancouver site, the proportion with a current infection stayed below 10 per 100 during the study (Lloyd-Smith et al., 2008). Female sex (adjusted odds ratio 1.68), unstable housing (1.49), borrowing a used syringe (1.60), needing help to inject (1.42) and injecting cocaine daily (1.41) were associated with higher risk of infection (Lloyd-Smith et al., 2008).

These factors tell the clinician something useful. The person who needs help to inject and the person who shares equipment are precisely the ones who benefit most from a space with sterile equipment and trained staff. The realist review also listed reductions in blood-borne viruses, infections and injection-related wounds among the possible outcomes, depending on the context (Stevens et al., 2024).

One of the biggest concerns before sites open is increased crime in the surrounding area. In Vancouver, drug trafficking charges were stable when comparing the periods before and after opening (average of 124 versus 116 per month), as were assaults and robberies (174 versus 180) (Wood et al., 2006). Vehicle thefts and break-ins fell from 302 to 227 in the same period (Wood et al., 2006).

The relationship with the police also appeared in the data. Among 1090 site clients, 182 (16.7 per 100) reported having been referred by the police when injecting in public (DeBeck et al., 2008). At study entry, 22 participants (2.0 per 100) said they had learned about the site from the police (DeBeck et al., 2008). Sex work (adjusted odds ratio 1.80) and daily cocaine injection (1.54) were associated with more police referral (DeBeck et al., 2008). The authors note that the criminalization of sex work exposes these people to more contact with the police, which may explain part of the association (DeBeck et al., 2008). The data suggest that the way police act on the street can bring people closer to or further from the health service.

How the sites work, according to the realist review

The review of 391 documents proposed a causal path: the feeling of safety and the immediate outcome of not dying are the conditions for a person to build trust and experience social inclusion (Stevens et al., 2024). From this combination, and depending on the context, less risky use practices, fewer infections and access to housing may emerge (Stevens et al., 2024).

These outcomes depend on the political and legal environment, which weighs differently on women and on people from racialized minorities (Stevens et al., 2024). The contrast appears in two examples cited by the review: New York, where two centers opened in majority Hispanic neighborhoods, with many people living on the street, HIV and a market saturated with fentanyl, and Geneva, where the Quai 9 center has operated for years within a network with social housing, opioid agonist treatment and drug checking (Stevens et al., 2024). The authors note that most of the literature comes from high-income countries, and that the conclusions may apply above all to those places (Stevens et al., 2024).

At overdose prevention sites in British Columbia, the team usually brings together workers with lived experience of drug use and harm reduction workers without that experience (Pauly et al., 2020). In Victoria, interviewees described other changes: the shift from a crisis response to overdose prevention, less trauma, and the replacement of using alone in bathrooms or in public with a space that has time to test the drug and inject (Pauly et al., 2020). The presence of drug checking inside the room is one example: at Insite, 1411 checks were done, and 1121 were positive for fentanyl (Karamouzian et al., 2018).

Trust appears in almost every study. For the clinician, it opens space for a visit to use to become a conversation about wounds, testing, treatment or housing. This point connects with interprofessional clinical reasoning, because the first contact is usually with a peer or a technician, not with the physician.

The table summarizes the main findings of the studies reviewed.

Outcome	What was observed	Where	Source
Overdoses attended	1004 overdoses and 453 potentially fatal ones in just over four years	Vancouver	(Milloy et al., 2008)
Overall mortality	Adjusted hazard ratio of 0.46 with weekly use of the room	Vancouver	(Kennedy et al., 2019)
Emergency visits	Drop in paramedic calls and emergency department visits	British Columbia	(Panagiotoglou, 2022)
Hospitalization and local mortality	Stable, with no detectable change	British Columbia	(Panagiotoglou, 2022)
Crime in the surrounding area	Drug dealing and assaults stable; vehicle thefts falling	Vancouver	(Wood et al., 2006)
Skin infections	Proportion below 10 per 100	Vancouver	(Lloyd-Smith et al., 2008)

Beyond injection and beyond the street

Most of the research on rooms deals with injection, and people who use by other routes also run the risk of overdose and face barriers to reaching health services (Speed et al., 2020). To map these services, the authors screened 22,882 titles and included 39 articles (Speed et al., 2020). The review identified 48 services that allow other routes of use, oral, nasal or inhaled, most of them in Germany (Speed et al., 2020). The differences included ventilated rooms or outdoor areas for people who smoke and shorter time limits for use by routes that do not need a needle (Speed et al., 2020).

The model also entered the hospital. In Vancouver, a hospital opened in 2021 a nurse-led overdose prevention room, open from 10 a.m. to 8 p.m., with two nurses per shift and accessible to inpatients, emergency department patients and outpatients (Dogherty et al., 2022). In nine months, the service recorded 1612 visits for injection, on average 42 per week, and 46 overdoses (Dogherty et al., 2022). Of the overdoses, 37 (80 per 100) required naloxone and 12 (26 per 100) required calling a code blue (Dogherty et al., 2022).

Four steps came before the opening: identifying the problem, seeking ethical guidance, adapting policies and practices, and supporting and training the team (Dogherty et al., 2022). The hospital serves a central area of the city, with a focus on historically marginalized populations (Dogherty et al., 2022). The authors grouped the lessons into six categories: engagement, communication, access, team training and support, data collection and safety (Dogherty et al., 2022). For those working on a ward, this finding is a reminder that drug use during hospitalization exists, and that the response can be organized rather than improvised.

What changes for the clinician

The supervised consumption room seems distant from the routine of many doctors. Even so, the studies offer direct lessons for the office, the emergency department and the ward.

Some points stand out clearly:

• Ask where and with whom the person uses, since using alone increases the risk of an overdose with no one to help (Pauly et al., 2020).
• Look into skin infections in people who need help injecting or share equipment (Lloyd-Smith et al., 2008).
• Remember that use during hospitalization happens, and that an organized response saves lives (Dogherty et al., 2022).
• Separate individual and population outcomes when reading studies about the rooms (Panagiotoglou, 2022).
• Value trust as the starting point for further care (Stevens et al., 2024).

This approach fits with what we discussed about clinical decision-making: the right course of action depends on understanding the person's context and what is possible at that moment. It also helps with prognostic reasoning, because mortality among people who inject drugs is high, and every contact with the service is a chance to change the trajectory.

For those who teach, the studies on the rooms offer a good critical reading exercise. They mix model-based estimates, cohorts, time series and realist reviews, each with its own limits. Discussing these designs in morbidity and mortality sessions or in reading groups helps avoid both unfounded enthusiasm and rejection without analysis. Fictional cases built from these data work well in case-based learning, for example when discussing the person who arrives at the emergency department with cellulitis on the arm and a history of injection. The same applies to dealing with diagnostic uncertainty, since the history of use is rarely complete at the first visit.

How NEXO helps train this reasoning

In NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues, and you propose the diagnosis by searching for the code in ICD-10.

The limits on using the game are described on the NEXO Economy page. If you research the teaching of clinical reasoning, the Researcher page explains how to get in touch. Professionals who write or review cases find information on the Curator and Creators pages, and institutions can learn about the proposal on the partnerships page. Training with differential diagnosis and clues and knowing how to find the right ICD-10 code complete this practice.

Frequently asked questions

What happens inside a supervised consumption room?

The person uses drugs obtained outside in the presence of a team that can prevent and manage overdoses (Stevens et al., 2024), and finds sterile equipment and referral to other services (Kerr et al., 2017).

Do the rooms reduce deaths?

In Vancouver, weekly use of the room was associated with lower overall mortality, with an adjusted hazard ratio of 0.46 (Kennedy et al., 2019), and the authors of another study estimated between 8 and 51 deaths averted in just over four years (Milloy et al., 2008).

Does crime increase near the rooms?

In Vancouver, drug dealing, assaults and robberies remained stable after the opening, and vehicle thefts fell (Wood et al., 2006).

Do the rooms serve only people who inject?

At least 48 described services accept oral, nasal or inhaled use, most of them in Germany (Speed et al., 2020).

Who works in the rooms?

In British Columbia, overdose prevention sites usually have peers with lived experience of drug use who provide first aid (Dogherty et al., 2022), alongside harm reduction workers (Pauly et al., 2020).

Can a room operate inside a hospital?

A Vancouver hospital maintained a nurse-led room, with 1612 visits and 46 overdoses handled in nine months (Dogherty et al., 2022).

Take-home messages

• The room offers supervision, sterile supplies and immediate overdose response for drugs obtained outside.
• The model began in Switzerland and Germany and today exists on several continents.
• In Vancouver, frequent use of the room was associated with lower overall mortality.
• Effects inside the room and effects on the population are different questions, with different answers.
• Crime in the surrounding area remained stable in the Vancouver studies.
• Safety and trust open the way to other care.

Want to practice your reasoning with progressive clues? Play the case of the day.

References

1. DeBeck K, Wood E, Zhang R, et al. Police and public health partnerships: evidence from the evaluation of Vancouver's supervised injection facility. Subst Abuse Treat Prev Policy. 2008. PMID 18462491.
2. Dogherty E, Patterson C, Gagnon M, et al. Implementation of a nurse-led overdose prevention site in a hospital setting: lessons learned from St. Paul's Hospital, Vancouver, Canada. Harm Reduct J. 2022. PMID 35120536.
3. Karamouzian M, Dohoo C, Forsting S, et al. Evaluation of a fentanyl drug checking service for clients of a supervised injection facility, Vancouver, Canada. Harm Reduct J. 2018. PMID 30200991.
4. Kennedy MC, Hayashi K, Milloy MJ, et al. Supervised injection facility use and all-cause mortality among people who inject drugs in Vancouver, Canada: A cohort study. PLoS Med. 2019. PMID 31770391.
5. Kerr T, Mitra S, Kennedy MC, et al. Supervised injection facilities in Canada: past, present, and future. Harm Reduct J. 2017. PMID 28521829.
6. Lloyd-Smith E, Wood E, Zhang R, et al. Risk factors for developing a cutaneous injection-related infection among injection drug users: a cohort study. BMC Public Health. 2008. PMID 19068133.
7. Milloy MJ, Kerr T, Tyndall M, et al. Estimated drug overdose deaths averted by North America's first medically-supervised safer injection facility. PLoS One. 2008. PMID 18839040.
8. Panagiotoglou D. Evaluating the population-level effects of overdose prevention sites and supervised consumption sites in British Columbia, Canada: Controlled interrupted time series. PLoS One. 2022. PMID 35316284.
9. Pauly B, Wallace B, Pagan F, et al. Impact of overdose prevention sites during a public health emergency in Victoria, Canada. PLoS One. 2020. PMID 32438390.
10. Speed KA, Gehring ND, Launier K, et al. To what extent do supervised drug consumption services incorporate non-injection routes of administration? A systematic scoping review documenting existing facilities. Harm Reduct J. 2020. PMID 33028363.
11. Stevens A, Keemink JR, Shirley-Beavan S, et al. Overdose prevention centres as spaces of safety, trust and inclusion: A causal pathway based on a realist review. Drug Alcohol Rev. 2024. PMID 39104059.
12. Wood E, Tyndall MW, Lai C, et al. Impact of a medically supervised safer injecting facility on drug dealing and other drug-related crime. Subst Abuse Treat Prev Policy. 2006. PMID 16722550.