Date: 2026-08-06 Purpose: close the named Switzerland gap in the drugs whitepaper's honesty box. This is a late, exploratory root trace. It does not revise the A6 scorecard row, rankings, or published site; doing so would require a new, structurally blinded score from an evidence packet that includes this record.
Reader test
The whitepaper used Switzerland as the missing positive counterfactual to the Portugal/Oregon funding-delay lesson. A reader could reasonably take that to mean Switzerland proves a Portugal-style decriminalization package works when it is funded properly. The record below says that is the wrong comparison. It is worth publishing because it narrows a reader-facing claim; it is not a new argument for the existing scorecard.
M3 limitation
The original protocol named Switzerland but did not pre-register Switzerland-specific support/refute/indeterminate conditions, and this pass started before those conditions were written. The pass is therefore a root-trace and scope correction, not a final causal adjudication. Any later comparative search must first specify: (1) the policy bundle to be compared, (2) the population and outcome, (3) the counterfactual, and (4) what result would change A6's evidence, mortality, or durability cells.
M2 search record
Date searched: 2026-08-06. Sources and strings: Swiss Federal Office of Public Health (FOPH): site:bag.admin.ch Switzerland heroin-assisted treatment and four pillar policy; PubMed/Google Scholar: Switzerland heroin-assisted treatment randomized trial, heroin assisted treatment systematic review, and Switzerland drug policy decriminalization law enforcement. Included: federal legal/program documents; systematic reviews and randomized evidence that distinguish supervised diacetylmorphine treatment from a general drug-policy bundle; work on policy adoption. Excluded: advocacy summaries, raw before/after national trends offered as causal evidence, and sources that call every Swiss harm-reduction measure "decriminalization" without identifying the legal change.
Finding 1 — Switzerland is not the positive decriminalization comparison
The Swiss federal description calls the model four pillars: prevention, therapy, harm reduction, and regulation and enforcement. The 2008 Narcotics Act revision put that framework into law. Its stated purposes include preventing unauthorized consumption, protecting public order, and fighting drug-related crime. That is materially different from Portugal's all-drug personal-possession decriminalization or from A6's proposed "decriminalization + dissuasion + prompt treatment funding" package.
Switzerland's relevant exceptional intervention is supervised, diacetylmorphine-assisted treatment (HAT), not a general legalization or decriminalization of heroin. It is federally licensed, delivered in specialist centres with medical and psychosocial care, and limited to adults with severe heroin dependence for at least two years, at least two unsuccessful treatment attempts, and adverse physical, mental, or social consequences. In 2021, FOPH reported about 1,700 recipients in 22 outpatient centres and one prison — about 8% of people with heroin addiction in Switzerland.
Verdict: REFUTES the use of Switzerland as a clean positive counterfactual for A6. It is a differentiated, high-intensity treatment and harm-reduction case inside an enforcement-retaining policy, not an estimate of the effect of decriminalization or of a national treatment appropriation.
Sources: FOPH, four-pillar policy; FOPH, Narcotics Act overview; FOPH, diacetylmorphine-assisted treatment.
Finding 2 — the targeted treatment has an evidence base, with strict scope
The clinical claim is stronger and narrower than the country-case claim. A 2023 systematic review identified nine randomized trials across eight studies (n=2,331) among people with chronic heroin addiction for whom standard opioid-substitution treatment had not worked. Compared with methadone, supervised HAT more consistently improved retention and reduced illicit-drug use; health findings were inconsistent. An earlier Cochrane review likewise found higher retention (four-study RR 1.44, 95% CI 1.19–1.75) and a possible but statistically inconclusive mortality benefit (four-study RR 0.65, 95% CI 0.25–1.69), alongside more medication-related adverse events.
The Swiss Geneva randomized trial fits that population: participants had failed at least two prior treatments. At six months, the HAT arm reported less daily street-heroin use and better selected mental/social-functioning measures, but not better work, housing, somatic-health, or other-drug outcomes. The small, short trial cannot identify a national mortality or public-order effect.
Verdict: SUPPORTED, narrowly. Supervised HAT is evidence-supported for a treatment-refractory, chronic-heroin-use population when delivered with intensive clinical follow-up. It does not establish a treatment for the broader US fentanyl/polysubstance population, a standalone safe-supply program, or the effect of the four-pillar bundle.
Sources: McNair, Monaghan & Montgomery 2023 systematic review; Cochrane review; Perneger et al. 1998 randomized trial.
Finding 3 — “funded properly” is not an established causal finding
FOPH documents a mature, monitored, federally licensed service, but the sources reviewed here do not supply a national counterfactual showing that a particular appropriation level, or speed of appropriation, caused Switzerland's outcomes. The 2009 policy-history case study itself characterizes the original Swiss HAT evaluation as an observational cohort and attributes adoption to documented evaluation, public debate, and adaptations as well as the policy response. That is useful implementation history, not a spending-effect estimate.
Nor can the country trend identify HAT's contribution: four pillars, changing heroin incidence, HIV prevention, policing, methadone and other treatment, and selection into the narrowly eligible HAT program moved together. The available clinical trials estimate a treatment comparison; they do not isolate the national policy bundle or financing mechanism.
Verdict: NOT MEASURED. Retire the phrase “Switzerland funded treatment properly” as evidence for A6 unless a future pass identifies comparable budgets, implementation timing, service capacity, and an appropriate counterfactual.
Sources: Uchtenhagen 2010 policy-change case study; FOPH program and monitoring description.
Implications for the filing
- Replace the open gap, not the scorecard, first. The whitepaper honesty box should say the Switzerland trace is complete: it does not validate A6; it identifies a distinct, tightly controlled HAT model with evidence for a narrower population. Defer the site edit while open PR #39 touches
site/drugs/index.html. - Keep A6's treatment-timing claim provisional. Portugal and Oregon remain two cases with the same rollout problem. Switzerland neither confirms nor refutes that claimed fix because it is not the same legal intervention and this pass found no budget/timing causal evaluation.
- A future architecture may deserve its own row. A US supervised-HAT demonstration for treatment-refractory patients would need separate legal, clinical-capacity, cost, and fentanyl-era transferability analysis. Folding it into A6 would conceal the very design difference this trace found.
Known unknowns
- No comparable national spending-and-capacity series was located in this pass.
- The reviewed RCTs largely predate the current illicit fentanyl supply; their external validity to fentanyl/polysubstance use is unresolved.
- No causal country-level estimate separates the effects of HAT, other opioid treatment, HIV interventions, enforcement, and changing heroin prevalence.