Background: Supervised injection sites (SIS) offer a hygienic environment in which people can inject drugs under observation; as such, these harm reduction services have been on the forefront of the overdose epidemic. We sought to understand factors predictive of an overdose requiring an emergency response intervention at SIS in Montréal, Canada.
Methods: We used administrative data from all four Montréal SIS from 1 March 2018 - 31 October 2022 to first calculate the rate of onsite overdose requiring intervention (e.g., naloxone or oxygen administration, nurse or paramedic assessment, etc.) and descriptive statistics. We then used a logistic regression model, with generalized estimating equations to adjust for clients' repeat visits, to test associations between onsite overdose intervention and client gender, age, drug injected (fentanyl vs other opioid vs non-opioid), most frequent injection location, frequency of injecting, duration of injecting, housing stability, attendance at multiple SIS, and time period (before or after 15 March 2020).
Results: During the observation period, Montréal SIS received 122,509 visits from 2,127 unique clients. The rate of overdose requiring intervention was 8.16 (95 % CI 7.66, 8.68) per 1000 visits. While 278 (13 %) of clients experienced an onsite overdose intervention, these clients accounted for 64,267 (52 %) SIS visits. Transgender clients (aOR = 2.28, 95 % CI 1.18, 4.41, compared to men) and clients younger than 25 were at greater risk of experiencing an onsite overdose requiring intervention (e.g., clients 35-44 had an aOR = 0.44, 95 % CI 0.30, 0.64, compared to clients younger than 25). Injecting most often in a public place was associated with greater risk of an onsite overdose requiring intervention (aOR = 3.62, 95 % CI 3.04, 4.30), while reporting unstable housing (aOR = 0.56, 95 % CI 0.47, 0.66) and attending more than one SIS (aOR = 0.13, 95 % CI 0.10, 0.16) predicted lesser risk. Compared to clients who reported injecting daily, greater risk of overdose requiring intervention was observed among those who reported injecting 3-6 days a week (aOR = 1.48, 95 % CI 1.18, 1.85), 1-2 days a week (aOR = 2.43, 95 % CI 1.90, 3.11), and "occasionally" (aOR = 2.60, 95 % CI 2.13, 3.18), but not those who reported not injecting in the past several months (aOR = 0.44, 95 % CI 0.25, 0.79). Compared to clients who had been injecting for five or more years, an injection duration of 1-4 years was associated with reduced risk of an overdose requiring intervention (aOR = 0.79, 95 % CI 0.64, 0.98), while those newer (<1 year) to injecting were at greater risk of such an intervention (aOR = 2.11, 95 % CI 1.50, 2.97). Compared to clients intending to inject fentanyl, we observed a lower risk of an overdose requiring intervention for those injecting other opioids (aOR = 0.22, 95 % CI 0.18, 0.26) or non-opioids (aOR = 0.06, 95 % CI 0.04, 0.08). Injecting after 15 March 2020 was also associated with greater risk (aOR = 1.99, 95 % CI 1.65, 2.41).
Conclusion: Across four Montréal SIS, site visits with an overdose requiring intervention were associated with transgender identity, younger age, stable housing, intending to inject fentanyl, injecting less frequently than daily, injecting for either more than five years or less than a year, single site attendance, and injecting most often in a public place. Montréal SIS serve a diverse clientele who inject drugs, with differing risks of onsite overdose.
Keywords: Overdose; People who inject drugs; People who use drugs; Supervised injection site.
Copyright © 2025 The Authors. Published by Elsevier B.V. All rights reserved.