Safety & outcomes in psychedelic medicine: what the data shows
This page explains how regulated programs measure safety, shows the aggregate figures they have published, and describes the anonymized outcomes program we run for Europe. It is not a review system and contains no provider ratings.
Why aggregate transparency, not star ratings
Star ratings work for restaurants because a meal is a low-stakes, repeatable purchase and the reviewer risks nothing by being named. Medical psychedelic care is the opposite: a report about your own treatment reveals sensitive health information, outcomes unfold over weeks rather than minutes, and a handful of self-selected voices — the delighted and the angry — can drown out the typical experience. A five-star average tells a future patient almost nothing about the question that actually matters: how often does this kind of treatment help, and how often does something go wrong?
Regulators reached the same conclusion. Oregon and Colorado, the two US states with licensed psilocybin programs, publish no provider ratings at all. Instead, they require every licensed center to report structured safety and outcome data, then de-identify it and publish it only in aggregate, across all providers. That design has a quiet elegance: because no individual provider is ever scored, there is no incentive to cherry-pick clients, suppress reports, or buy a better number — and patients still get the honest, population-level picture they need.
We think that model — not reviews — is the right foundation for psychedelic medicine in Europe. Below: what the regulated programs have published so far, and how our own program adapts their approach.
What regulated programs report
Oregon Psilocybin Services
Since 1 January 2025, licensed service centers in Oregon must report defined client-level and center-level data to Oregon Psilocybin Services every quarter. OPS then reviews, aggregates, and de-identifies the data — combining it across all licensed service centers before publication, so per-provider attribution is structurally impossible (OPS data dashboard; OHA announcement). The published first-year safety figures:
>99.0%
2.42
2.79
7
Oregon's system also has a known, publicly discussed weakness: structured reporting largely ends about 72 hours after a session, so problems that emerge later are undercounted (ScienceDirect, 2026; Oregon Capital Chronicle). We designed our follow-up windows around exactly this critique — see below.
Colorado Natural Medicine Division
Colorado follows the same pattern: statute directs the state health department (CDPHE), together with the Department of Revenue and DORA, to collect adverse-health-event data, and healing centers must supply demographics and participation outcome data. The Natural Medicine Division publishes an Adverse Health Event Reporting Bulletin for licensees and annual program reports — mandatory licensee reporting in, aggregate public reporting out (Natural Medicine Division; annual reports; SB25-297).
Independent outcome research
Alongside the state dashboards, an OHSU-led, NIDA-funded five-year naturalistic study is tracking standardized outcome measures (PHQ-8, GAD-7, WHO-5) for supervised psilocybin clients — more than 300 enrolled toward a target of 1,600 — with aggregate scientific publication only (OHSU News, 2026; medRxiv preprint). It is further evidence that the field's serious outcome work is done with validated scales and aggregates — not review scores.
Our European program
Europe has no equivalent of the Oregon dashboard yet. Our program is an honest, voluntary adaptation of the regulator model — with its limits stated plainly rather than papered over:
- Anonymous at the source. We collect fully anonymized experience reports: no name, no email, no provider field. Identifiers are never collected, so they can never leak.
- Aggregates only, never per provider. Published figures are combined across providers and therapy types. No figure is ever tied to an individual provider.
- Minimum n = 10 per country. No figure is shown for a country until at least 10 reports have been received there — small cells are suppressed, mirroring the de-identification standards regulators use.
- Adverse-event taxonomy adapted from Oregon's categories. Reports are classified by type (behavioral / medical) and severity, in neutral clinical language.
- Optional validated scales. Reporters can add before/after self-report ratings in the style of validated instruments (PHQ-9 / GAD-7), so the dataset can mature into an outcomes evidence layer.
- Counts and shares, never "rates". Unlike Oregon, we have no mandatory reporting and no session denominator, so we will publish counts and shares of reports received — never a claim like "X% of therapy causes harm".
- Longer follow-up. Learning from the 72-hour critique of Oregon's system, we invite reporters to follow up at 2 and 12 weeks after treatment.
Been through treatment? Share your experience anonymously
Five questions, no account, no names. Your report joins the aggregate dataset and is never published individually.
Share your experience →Frequently asked questions
Is this a reviews or ratings system in disguise?
No. We never publish individual reports, star ratings, or any per-provider score. Reports are combined across many providers and therapy types and published only as aggregate statistics — the same structural choice Oregon makes when it combines data across all licensed service centers before publication. Nothing we publish can be traced to a specific provider, so there is nothing to game and nothing to retaliate against.
Why does it have to be anonymous?
Two reasons. First, an experience report about a mental-health treatment is health data, and the strongest protection is to never collect your identity at all: the form has no name, email, or provider field, and we store no identifiers. Second, anonymity removes the pressure to be polite or to fear consequences — we would rather have honest aggregate data than attributable individual stories.
Will my provider see what I wrote?
No provider ever sees an individual report. Providers, like everyone else, only see the published aggregate figures — which are never broken down by provider and are only shown for a country once at least 10 reports have been received there. Because we do not ask which provider treated you, a link back to your provider is structurally impossible.
Who runs the program?
The program is operated by the PsyStandard team — the same team that runs the registry verification behind this directory. It is kept strictly separate from listings and from any commercial relationship: providers cannot pay to influence, access, or suppress the data, and listing in the directory is free regardless of what the aggregate data shows. The full methodology is public.
Methodology note
PsyAccess keeps outcome data strictly separate from listings and from any commercial relationship with providers. Reports are collected with explicit consent, de-identified immediately, and published only in aggregate with small-cell suppression. The same verification principles that govern our listings — documented sources, public methodology — apply here, and our privacy policy describes how submitted data is handled. Powered by PsyStandard.