Amid an escalating drug overdose epidemic [1], Rhode Island became the first state in July 2021 to authorize an overdose prevention site (OPS)—a space for people to consume pre-obtained drugs with sterile supplies. Later, on November 30th, 2021, New York City opened the nation’s first two authorized OPSs with approval from then Mayor De Blasio. International and unsanctioned domestic OPSs have reduced overdose mortality, drug use, and infectious disease risk, and facilitated access to health and social services [2,3,4,5,6,7]. No death has ever been reported in an OPS.

The first sanctioned OPS was established in Switzerland in 1986; today, there are over 120 sites located in 11 countries [8]. International OPCs operate under a range of models (such as stand-alone centralized sites, sites embedded within hospitals or shelters, or mobile sites), though the most common OPSs are single, centralized locations. Indeed, Rhode Island’s regulations are designed for this centralized model, and New York City’s sites were opened in this style as well. But, what if rather than a single site, Rhode Island pursued a model of decentralization, integrating supervised consumption services into the locations already accessed by people who use drugs (PWUD)? A growing body of data supports the need to deliver HIV prevention in low-barrier settings where PWUD already access services [9]. With opioid use disorder (OUD) increasingly recognized as a chronic and relapsing disease, evidence-based treatment (including medications for OUD [MOUD], harm reduction services, and others) should be integrated into primary care, pharmacies, methadone clinics, social services organizations, fire stations, or other settings alongside other chronic conditions. For example, a clinic may designate a clinic room as their OPS, with injection supplies, peer support, and healthcare personnel available to respond in case of an overdose. To patients, this could signal a non-stigmatizing culture, acknowledge that clinicians should support a patient when their substance use disorder is most active, and may facilitate discussions on MOUD during ongoing use. Drawing on successes of syringe exchange and other peer-based approaches to harm reduction, PWUD should be meaningfully included throughout program development and implementation.

A decentralized model may have additional benefits. A majority of clients may travel only 1 mile or less to use an OPS [10], meaning that any single location may be inaccessible to some. Decentralized services would assist regions without geographically concentrated drug use. Additionally, the current epidemics, driven by illicitly manufactured fentanyl and stimulants, require frequent injection events—suitable to a decentralized model allowing multiple access points throughout the day. Finally, as federal approval or funding for OPS operation is unlikely in the near future, a decentralized model using existing healthcare infrastructure may minimize costs and improve feasibility.

Multiple challenges remain. Rhode Island’s proposed regulations make it challenging for an existing facility to be licensed as an OPS. Community members who are concerned about having an OPS in their own neighborhood (including those with not-in-my-backyard [NIMBY] sentiments) may be reassured to know that criminal activity actually decreased in the neighborhood surrounding an unsanctioned OPS in the USA in the 5 years following the OPS opening [11]. There is also no evidence that establishing an OPS leads to an influx of clients from other communities. Rather, the majority harm reduction clients reside close to these programs [10]. Critically, the structure and modality of supervised consumption services can and should change, driven by epidemiologic factors (i.e. rates of overdoses) and by a community’s health needs [8]. Still, with the overdose crisis worsening, a decentralized medical model may improve feasibility and save the most lives.