Prescribed pharmaceutical alternatives vs. a contaminated illicit opioid market — a contested harm-reduction policy tradeoff
The starting condition safe supply programs are designed to address is not "drug use" in the abstract — it is a specific market failure: an unregulated, unlabeled opioid supply where potency varies unpredictably from batch to batch, dealer to dealer, and week to week. Since illicitly manufactured fentanyl and its analogues displaced heroin across much of North America, that variability has become the dominant driver of overdose mortality.
A person who uses opioids regularly typically develops tolerance to a fairly consistent dose. Fatal overdose most often occurs not simply because someone used an opioid, but because the dose they took was far more potent than what their tolerance was calibrated to — frequently because the product was fentanyl, or a fentanyl analogue, sold or believed to be a weaker opioid, or of unexpectedly high concentration within a fentanyl-adulterated batch.
Illicit manufacturing has no quality control: mixing is done by hand or basic equipment, with no guarantee of even distribution through a batch ("hot spots"), and no dosage labeling. Two tablets from the same bag can differ severalfold in potency. This is structurally different from the pharmaceutical supply, where dose is verified and standardized.
Harm reduction researchers frame this as a "poisoning crisis" layered on top of an "addiction crisis" — the two are related but distinct, and interventions that address contamination (like safe supply) target a different mechanism than interventions that address substance use disorder directly (like treatment).
Before considering safe supply, it is worth situating it among more established interventions:
• Naloxone distribution — reverses opioid overdose in progress; widely distributed, strong evidence base, no serious controversy about the intervention itself • Supervised consumption sites (SCS) / overdose prevention sites (OPS) — people use their own (illicit) drugs under supervision so an overdose can be reversed immediately; substantial evidence of reduced on-site mortality • Medication-assisted treatment (MAT) — methadone and buprenorphine substitute for illicit opioids as part of a treatment pathway; decades of RCT evidence for reduced mortality and illicit use • Syringe service programs (SSP) — reduce needle-borne infection transmission; long-established evidence base
Safe supply sits at the newer, less-established end of this spectrum: it shares harm-reduction logic with these interventions (meet people where they are, reduce mortality risk without requiring abstinence) but has a much shorter track record and a thinner evidence base, discussed in later stages.
If the proximate cause of most overdose deaths is not knowing what is actually in the substance, one candidate intervention is to remove that uncertainty for at least some portion of the population: provide people at high risk with a pharmaceutical-grade opioid of known identity and known dose, obtained through a health service rather than an illicit market.
This is the core hypothesis this simulator explores — and, as later stages make clear, it is a hypothesis under active empirical and political dispute, not an established consensus intervention like naloxone or SSPs.
Safe supply programs are not a general prescription-on-demand system. Real-world pilots have used targeted eligibility criteria intended to concentrate a scarce, closely monitored intervention on people facing the highest imminent overdose risk — while leaving the great majority of the illicit-market population, at least initially, outside the program.
Published program models (largely from British Columbia, Canada) have generally targeted people who:
• Have a documented history of opioid use disorder and/or a recent non-fatal overdose • Are unable or unwilling to access, or have not been stabilized by, standard opioid agonist treatment (methadone, buprenorphine) • Use the unregulated supply frequently and are assessed at high risk of overdose death • Are willing to engage with a prescriber and attend a dispensing site with some regularity
The rationale for narrow eligibility is triage: pharmaceutical-grade alternatives, prescriber time, and monitoring capacity are limited, so programs concentrate them on people least likely to be reached or protected by existing services, and most likely to die without an alternative.
Tighter eligibility (this simulator's "high-risk" screening pool) keeps the intervention small and closely monitored, which supporters argue protects program integrity and makes diversion and misuse easier to detect. Critics counter that narrow eligibility limits the population-level impact on overdose deaths, since most illicit-market users never enter the program at all.
Broader eligibility could reach more people — but stretches clinical oversight thinner, and, combined with take-home dispensing (Stage 3), is the configuration most often cited by critics as carrying the highest diversion risk. This is one of the central tuning tensions in real program design, and is reflected in the sliders used later in this simulator.
Because enrollment has generally been small relative to the size of the illicit-opioid-using population in affected regions, safe supply programs to date function more as a targeted clinical intervention for a high-risk subgroup than as a population-wide substitute for the illicit market.
Because safe supply programs generally are not randomized (enrollment follows clinical referral, not lottery), most published evaluations compare enrolled clients to a non-randomized comparison group of similar-risk people who did not enroll, using administrative health records (emergency department visits, hospitalizations, overdose deaths) before and after enrollment.
This observational design is a recurring point of methodological critique, discussed further in Stage 5: differences between enrolled and non-enrolled groups (e.g. enrolled clients may already be more engaged with health services) can bias comparisons in ways a randomized trial would control for.
Once enrolled, clients receive a pharmaceutical-grade opioid — most commonly hydromorphone tablets, sometimes other formulations — of known identity and known dose. How that medication physically reaches the client is one of the most consequential design choices in the entire program, and the one most directly linked to the diversion debate.
Witnessed (supervised) dispensing: the client consumes the medication at a clinic or pharmacy under direct staff observation, similar to supervised methadone dosing. This makes diversion of that specific dose essentially impossible, but requires the client to attend in person, often multiple times a day — a significant burden that can reduce retention, especially for clients with unstable housing, employment, or transportation.
Take-home dispensing: the client is given a supply of tablets to take at their own schedule, away from the clinic. This dramatically improves convenience and access — closer to how a chronic-disease medication is normally managed — but removes direct observation of consumption, opening the possibility that a dispensed dose is sold, traded, or given to someone else instead of consumed by the person it was prescribed for.
Proponents of take-home models argue that requiring multiple daily clinic visits is itself a barrier that pushes marginalized clients back toward the illicit market out of sheer logistical necessity, undermining the program's core purpose. They also note that take-home dosing is standard practice for stable methadone and buprenorphine patients after an initial supervised period, and argue safe supply clients should eventually be treated similarly as trust is established.
Critics respond that hydromorphone tablets are a more readily divertible, more easily sold form factor than liquid methadone, and that take-home models in some jurisdictions have coincided with reports of dispensed tablets appearing in the illicit market — sometimes reportedly at prices low enough to suggest resale by recipients, not exclusively personal use.
This witnessed-vs-take-home choice is exactly what the "Dispensing Model" slider in this simulator represents: it is not a binary policy failure/success switch, but a genuine dial between access/retention on one side and supervision/diversion control on the other.
The therapeutic logic depends on the medication being a known, standardized quantity of a known substance — this is what allows a client to titrate their own use without the batch-to-batch potency roulette of the illicit market described in Stage 1. Hydromorphone is roughly comparable in effect to the opioid many clients are accustomed to, but is not equipotent to fentanyl on a milligram basis, which becomes central in Stage 4: a dose that is "known" is not automatically "adequate" for someone whose tolerance has been shaped by daily fentanyl use.
The entire premise of safe supply rests on a behavioral assumption: that a client given a prescribed alternative will use it in place of the illicit supply, not alongside it. Whether that assumption holds depends heavily on whether the prescribed dose is actually adequate for that individual's tolerance — which is where the sharpest, most evidence-contested part of the debate lives.
A frequently raised critique — from both harm-reduction researchers and program skeptics — is that the potency gap between hydromorphone (a moderate-strength opioid) and the fentanyl-dominated illicit supply is large enough that program doses, especially early in a client's enrollment or under conservative prescribing protocols, may fall well short of what is needed to prevent withdrawal or cravings.
If that happens, a client may take their prescribed dose and still return to the illicit market to "top up" — meaning they remain exposed to contamination risk on top of, not instead of, the program dose. In this simulator, this is modeled directly: at low "Dosing Adequacy," program enrollees continue supplementing from the illicit market at a scaled rate, and their overdose risk does not fall as far as the enrollment narrative alone might suggest.
This is arguably the single most important mechanism to understand in the whole debate: a safe supply program only reduces population overdose risk to the extent that adequate dosing (or an otherwise successful substitution) actually displaces illicit use — it does not automatically do so just by existing.
Where dosing is titrated to the individual — sometimes requiring doses well above initial protocol defaults, achieved through iterative clinical adjustment — proponents point to program evaluations reporting reduced emergency department visits, reduced hospitalization, and reduced self-reported illicit opioid use among enrolled clients relative to before enrollment or relative to comparison groups.
The honest summary is that outcomes appear to depend heavily on individualized, adequately titrated dosing and continuity of engagement — a moderately-run program with rigid, conservative dosing may show much weaker effects than a well-resourced one with flexible, client-responsive titration. This heterogeneity across programs is itself a source of the wider empirical disagreement covered in Stage 5.
Moving the slider toward "underdosed" does not represent a program failing at its stated design — it represents a real, frequently discussed clinical and policy failure mode: doses set too conservatively (whether from prescriber caution, regulatory limits, or cost pressure) relative to an individual client's actual fentanyl-era tolerance.
Moving it toward "adequate" represents individualized titration succeeding — which evidence suggests requires clinical flexibility, follow-up capacity, and willingness to prescribe doses that may look high relative to conventional opioid-prescribing norms.
Zooming out from any single client to population-level policy: does making safe supply programs larger, looser on eligibility, and more take-home-oriented reduce overdose deaths overall — or does it mainly shift risk from overdose toward diversion, without a correspondingly large mortality benefit? This is genuinely disputed, and the evidence base, while growing, is not yet large or randomized enough to settle it.
Several British Columbia program evaluations, using administrative health-record comparisons, have reported associations between safe supply enrollment and reduced all-cause mortality, reduced overdose-related emergency department visits, and reduced acute-care hospitalization among enrolled clients relative to non-enrolled comparison groups or relative to clients' own pre-enrollment periods.
Proponents frame these results — while acknowledging their observational nature — as consistent with the harm-reduction logic laid out in Stage 1: removing contamination-driven risk for a population that existing services (treatment programs, abstinence-based approaches) were failing to reach or retain.
Critics — including some public health researchers, law enforcement voices, and, in Canada, some provincial governments that have since scaled back or restricted take-home hydromorphone dispensing — raise several distinct concerns:
• Diversion — reports and some law-enforcement seizure data suggesting dispensed hydromorphone has entered illicit markets, in some cases reportedly used by others (including youth) to initiate or maintain opioid use • Underdosing — as discussed in Stage 4, concern that program doses often do not match fentanyl-era tolerance, limiting substitution effectiveness • Study design — the absence of randomized controlled trials means favorable observational findings could partly reflect selection effects (clients willing and able to enroll and stay engaged may differ systematically from those who do not) • Scale — enrollment has been small relative to overall illicit-opioid-using populations in affected areas, limiting demonstrated population-level mortality impact even where individual-level benefit is plausible
Both the strongest supportive evidence and the strongest critiques in this debate currently rest substantially on observational data, administrative records, and program-level reporting — not on large randomized trials. Treat specific effect-size claims from either side, including the illustrative numbers used in this simulator, as approximate and contested rather than settled facts.
This simulator's two sliders map onto the real axis of policy disagreement: a "tight eligibility + witnessed dosing" configuration minimizes diversion risk but reaches fewer people and stresses program capacity; a "broad eligibility + take-home dosing" configuration reaches more people with less burden per client but is the configuration most associated with diversion concerns in public debate.
Neither configuration is presented here as correct. The bar comparison in this stage is illustrative of the mechanism the debate is actually about (dosing adequacy driving overdose-risk reduction, dispensing model driving diversion risk) — not a claim about the true magnitude of either effect in any specific jurisdiction.
It is worth closing where Stage 1 began: naloxone distribution, supervised consumption sites, and medication-assisted treatment (methadone/buprenorphine) all have decades of accumulated evidence and are broadly accepted, even by many skeptics of safe supply specifically, as effective harm-reduction or treatment tools.
Safe supply — particularly take-home dispensing of non-agonist-treatment opioids like hydromorphone — is the newest, least-studied, and most politically contested intervention on that spectrum. That does not make it wrong; it makes it an active area of policy experimentation and evaluation, where the honest current answer to "does this work?" is "the mechanism is plausible, some evidence is favorable, some evidence and reporting raises real concerns, and the randomized evidence needed to settle it does not yet exist."
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Tight eligibility + witnessed dosing | High-risk clients only, on-site supervised consumption | Minimizes diversion; maximizes clinical oversight per client | Lowest diversion risk, easiest to evaluate |
| Tight eligibility + take-home dosing | High-risk clients only, unsupervised take-home doses | Improves access/retention for enrolled clients only | Better retention, moderate diversion exposure |
| Broad eligibility + witnessed dosing | Wider client pool, on-site supervised consumption | Reaches more people but strains clinic capacity | Larger reach with controlled diversion |
| Broad eligibility + take-home dosing | Wider client pool, unsupervised take-home doses | Maximizes reach and convenience, minimizes oversight | Largest potential reach; highest diversion-risk exposure |